Sie sind auf Seite 1von 1130

International Journal of Scientific Research in

Computer Science, Engineering and


Information Technology

ISSN : 2456-3307

Volume 5, Issue 1, January-February-2019

International Peer Reviewed, Open Access Journal


Bimonthly Publication

Published By

Website URL : www.technoscienceacademy.com

i
Advisory/Editorial Board

Associate Editor

• Dr. Gaurav Jangra


Assistant Professor, Synetic Business School, Ludhiana, Punjab, India
Email : gaurav.jangra@sbs.ac.in
• Dr. Sheshang D. Degadwala
Head of Computer Department, Sigma Institute of Engineering, Vadodara, Gujarat,
India
Email : sheshang.cs.engg@sigma.ac.in
• Dr. K. Senthil Kannan
Dean Research, Vice Principal & Research Scientist EDAYATHANGUDY GS PILLAY
Arts and Science College Nagapattinam Tamil Nadu, India
Email : senthilkannan@egspec.org

International Editorial Board

• Dr. Andie Siahaan


Computer Science, Universitas Pembangunan Panca Budi, Indonesia
• AbdulGaniyu Abdu Yusuf
Computer Science, National Biotechnology Development Agency (NABDA), Abuja,
Nigeria
• Dr. M. A. Ashabrawy
Computer Science and Engineering, Prince Sattm bin Abdulaziz University, Kingdom
Saudi Arabia(KSA), Saudi Arabia
• Dr. V. Balaji
Bahir Dar University, Bahir Dar, Ethiopia
• Lusekelo Kibona
Department of Computer Science, Ruaha Catholic University (RUCU) , Iringa,
Tanzania
• Md. Amir Hossain
IBAIS University/Uttara University, Dhaka, Bangladesh
• Mohammed Noaman Murad
Department of Computer Science, Cihan University Erbil, Kurdistan Region, Iraq
• Prof. Dr. H. M. Srivastava
Department of Mathematics and Statistics, University of Victoria, Victoria, British
Columbia, Canada

ii
• Prof. Sundeep Singh
Mississauga, Ontario, Canada
• Dr. Froilan Mobo. MOBO, DPA, MBA, MSCS, LPT
Computer and Social Sciences, Carl Balita Review Center : Olongapo City, Zambales,
Philippin
• Dr. Vaibhav Sundriyal
Department of Computer Engineering, Old Dominion University research
Foundation, USA (Phd in Computer Engineering from Iowa State University, USA)
• Dr. Katarzyna Pietrucha-Urbanik
Faculty of Civil and Environmental Engineering, Rzeszow University of Technology,
Poland
• Indrasen Poola
Data Scientist & Artificial Intelligence - Industry Consultant & Professional
Researcher, Palo Alto Networks, Santa Clara, CA, USA
• Dr. Mohd Muntjir
Information Technology Department, College of Computers and Information
Technology, Taif University, Al-Hawiya, Taif, Kingdom of Saudi Arabia
• Vaibhav Sundriyal
Research Scientist, Old Dominion University research Foundation, USA
• Dr. Elsadig Gamaleldeen
Assistant Professor, Omdurman Ahlia University, Sudan
• SK. Saidhbi
Assistant Professor, Department of Information Systems, University of Gondar,
Ethiopia

Editorial Board - India

• Prof. Shardul Agravat


Information Technology, C U Shah Unuiversity, Surendranagar, Gujarat, India
• Prof. Sarita Dhawale
Ashoka Center for Business & Computer Studies, Ashoka Marg, Ashoka Nagar,
Nashik, Maharashtra, India
• Dr. Syed Umar
Professor & R & D Head, Department of Computer Science and Engineering, MLR
Institute of Technology, Hyderabad, Telangana, India
• Prof. C. K. Raina
HOD, Computer Science and Engineering Department, Adesh Institution of
Technology Chandigarh, Gharuan, Chandigarh, India

iii
• Dr. R. Gayathri
Assistant Professor, Department of ECE, Annamalai University, Chidambaram, Tamil
Nadu, India.
• Dr. R. Reka
Professor & Head, Department of Computer Science & Engineering, Annai
Mathammal Sheela Engineering College, Tamil Nadu, India
• Dr. Nilamadhab Mishra
Post Graduate Teaching & Research Department, School of Computing, Debre Berhan
University, Ethiopia, India
• Dr. S. Ramkumar
Department of Computer Applications, Kalasalingam University, Krishnankoil,
Sirivilliputhur, Palani, Dindugal, Tamil Nadu, India
• G. Naveen Balaji
Department of ECE, SNS College of Technology, SNS Kalvi Nagar, Vazhiyampalyam
Pirivu, Coimbatore
• Dr. G. Rajkumar
Department of Computer Applications, N. M. S. S.Vellaichamy Nadar College,
Madurai, Tamilnadu, India
• Prof. Sagar Bhilaji Shinde
E & TC Engineering Department, JSPM, Narhe Technical Campus, Pune,
Maharashtra, India
• Prof Sarita Dhawale
HOD, Computer Science Department, Ashoka Education Foundation’s, Ashoka
Center for Business and Computer Studies, ACBCS, Nashik, India
• Prof. Umesh Sehgal
Associate Professor, GNA University, Phagwara, Punjab, India
• Prof. Vaishali Kalaria
Information Technology, RKU, Rajkot, Gujarat, India
• Prof. H. B. Jethva
Computer Enginering, L. D. College of Engineering, Ahmedabad, Gujarat, India
• Prof. Bakul Panchal
Information Technology, L. D. College of Engineering, Ahmedabad, Gujarat, India
• Prof. Bhavesh Prajapati
Computer Science, Government MCA College Maninagar, Ahmedabad, Gujarat, India
• Dr. Syed Umar
Computer Science and Engineering, KL University, Guntur, Andhra Pradesh, India
• Prof. S. Jagadeesan
Computer Science, Nandha Engineering College Erode, Tamil Nadu, India
• Prof. Joshi Rahul Prakashchandra
Information Technology, Parul Institute of Engineering & Technology, Vadodara,
Gujarat, India
iv
• Dr. Aftab Alam Tyagi
Department of Mathematics, SRM University NCR Campus, Uttar Pradesh, India
• Prof (Dr.) Umesh Kumar
Department of Science & Technology, Govt. Women’s Polytechnic, Ranchi,
Jharkhand, India
• Dr. N. Pughazendi
Computer Science and Engineering, Panimalar Engineering College Chennai,
Tamilnadu, India
• Sachin Narendra Pardeshi
Department of Computer Engineering, R.C.Patel Institute of Technology, Shirpur,
Maharashtra, India
• Dr. Bangole Narendra Kumar
Department of Computer Science and Systems Engineering, Sree Vidyanikethan
Engineering College, Tirupati, Andhra Pradesh, India
• Dr. Dhananjaya Reddy
M.Sc.(Maths),M.Sc.(Stat), M.Phil.(Maths), M.phil. (Statistics), B.Ed., PhD Department
of Mathematics,Government Degree College, Puttur, Chittoor, Andhra Pradesh, India
• Dr. Ajitesh Singh Baghel
Department of Computer Science, Awadhesh Pratap Singh University, Rewa, Madhya
Pradesh, India
• Dr. Dharmaraj K Veer
Director, UGC-INFLIBNET SOUL Co-Ordinator, Maharashtra & Goa Region,
Knowledge Resource Center, Dr. Babasaheb Ambedkar Marathwada University,
Aurangabad, Maharashtra, India
• Rakesh Oruganti
Assistant Professor Jayamukhi Institite of Technology and Science, Chennaraopet
Mandal, Narsampet, Nekkonda Rd, Arshanapally, Telangana, India
• Dr. M. Ramesh Kumar
Associate Professor/Project Coordinator, Department of Computer Science and
Engineering, VSB College of Engineering Technical Campus, Coimbatore, Tamilnadu,
India
• Dr. Ikvinderpal Singh
Department of Computer Science & Applications, Trai Shatabdi GGS Khalsa
College,Amritsar, Punjab, India
• Dr. Kshitij Shinghal
Associate Professor & Dean Academics, Department of Electronics & Communication,
MIT, Moradabad, Uttar Pradesh, India
• Udit Narayan Bera
Department of Electronics and Communication Engineering, St. Aloysius Institute of
Technology, Jabalpur, Madhya Pradesh, India

v
• Dr. Nageswara Rao Moparthi
Department of Computer Science and Engineering, Velagapudi Ramakrishna
Siddhartha Engineering College, Andhra Pradesh, India
• Dr. Sharad Gangele
Professor and Head, Department of Computer Science & Application, RKDF
University, Bhopal, Madhya Pradesh, India
• Dr. K. Praveen Kumar
Associate Professor and Vice-Principal in Chaitanya Institute of Technology and
Science, Kishanpura, Hanamkonda, Telangana, India
• Dr. B. Bazeer Ahamed
Associate Professor, Department of Computer Science & Engineering, Balaji Institute
of Technology and Science, Warangal, Telagana, India
• Dr. Anil Kumar Dubey
Associate Professor & Deputy HOD (Academics), Computer Science, Poornima
Institute of Engineering & Technology, Jaipur, Rajasthan, India
• Dr. Yogesh M. Rajput
College of Computer Science and Information Technology (COCSIT), Ambajogai
Road, Latur, Maharashtra, India
• Dr. K. Umapathy
Associate Professor, ECE, SCSVMV University, Kanchipuram, Tamil Nadu, India
• K. Gutnadha Gupta
Assistant Professor of Computer Science and Enineering, Sri Indu College of
Engineering and Technology, Hyderabad, Telangana, India
• Namrata Singh
Computer Science and Engineering, Abes Engineering College, Ghaziabad, Uttar
Pradesh, India
• Dr. K. Nehru
Professor, Electronics and Communication Engineering Department, Institute of
Aeronautical Engineering, Hyderabad, Telangana, India
• Dr. Radha Mothukuri
Assistant Professor, Department of Computer Science and Engineering, KL
University, Vaddeswaram, Guntur, Andhra Pradesh, India
• Dr. K. R. Ananth
Department of MCA, Velalar College of Engineering and Technology, Erode,
Tamilnadu, India
• Rydhm Beri
Assistant Professor, PG Department of Computer Science, BBK DAV College for
Women, Amritsar, Punjab, India
• S. Kannadhasan
Department of Electronics and Communication Engineering, Tamilnadu Polytechnic
College, Austinpatti, Madurai, Tamil Nadu, India
vi
• Vishal Kanjariya
Assistant Professor, Information Technology Department, Birla Vishvakarma
Mahavidyalaya (BVM), Engineering College, Vallabh Vidyanagar, Managed by
Charutar Vidyamandal, Anand, Gujarat, India
• Dhairya J. Vyas
Assistant Professor, EC Department, Sigma Institute of Engineering, Vadodara,
Gujarat, India
• Arpana Mahajan
Assistant Professor, Computer Department, Sigma Institute of Engineering, Vadodara,
Gujarat, India
• Dr. M. Anjan Kumar
Associate Professor, CSE Department, Vivekananda Institute of Technology &
Science, Karimnagar, Telangana, India
• Dr. A. Vamshi Krishna Reddy
Assistant Professor, Centre for Environment, Institute of Science & Technology,
Jawaharlal Nehru Technological University, Hyderabad, Telangana, India
• Dr. Ashok Koujalagi
Assistant Professor & Postdoctoral Researcher, P. G. Department of Computer
Science, Basaveshwar Science College, Bagalkot, Karnataka, India
• Dr. C. S. KanimozhiSelvi
Professor, Computer Science and Engineering, Kongu Engineering College,
Perundhurai, Erode, Tamil Nadu, India
• Dr. Mukhdeep Singh Manshahia
Assistant Professor (Computing), Department of Mathematics, Punjabi University,
Patiala, Punjab, India
• Dr. Nageswara Rao Moparthi
Department of Computer Science and Engineering, Velagapudi Ramakrishna
Siddhartha Engineering College, Andhra Pradesh, India
• Dr. Parshuram Singh
Associate Professor, Dean, HOD (Physics), Mahaveer Institute of Technology &
Science, MITS-Jadan (A Unit of Vidhya Jyothi Trust, Chennai & Pali), NH-14, Raniya
Bera, Jadan, Pali, Rajasthan, India
• Dr. S. Sasikala
Associate Professor, Department of Computer Applications, Hindusthan College of
Arts and Science, Coimbatore, Tamil Nadu, India
• Dr. G. Maria Priscilla
Professor & Head, Department of Computer Science, Sri Ramakrishna Colleges of Arts
and Science (formerly SNR Sons College), Coimbatore, Tamil Nadu, India
• Dr. M. Jayakameswaraiah
Assistant Professor, Department of CSA, Reva University, Bangalore, Karnataka, India

vii
• Kishor Kumar Gajula
Research & Development Head, Newzen Infotech, Hyderabad, Telangana, India
• Mallavolu Malleswararao
Associate Professor & HOD, Department of Electrical & Electronics Engineering,
RISE Prakasam Group of institutions, Ongole, Andhra Pradesh, India
• Radha Mothukuri
Assistant Professor, Department of Computer Science and Engineering, QIS Institute
of Technology, Ongole, Andhra Pradesh, India
• Prof. S. Raja
Assistant Professor and Research and Development Coordinator, Department of ECE,
Sri Shakthi Institute of Engineering and Technology, Coimbatore, Tamilnadu, India
• Dr. J. Viji Gripsy
Assistant Professor, Department of Computer Science, PSGR Krishnammal College for
Women, Peelamedu, Coimbatore, Tamil Nadu, India
• Dr. Dickson S
Assistant Professor of Mathematics, Vivekanandha Educational Institutions
Tiruchengode, Tamil Nadu, India
• Dr. Ranjana Rajnish
Assistant Professor, Amity Institute of Information Technology(AIIT), Amity
University, Lucknow, Uttar Pradesh, India
• Dr. Nirajkumar C. Mehta
Associate Professor, Institute of Technology and Management Universe, Vadodara.
Gujarat, India
• Pulla Reddy Mekala
Southern Region Farm Machinery Training & Testing Institute, Tractor Nagar,
Garladinne, Anantapur (Dist.), Andhra Pradesh, India
• Dr. Rahul Saha
Division of Research and Development, Lovely Professional University, Punjab, India
• Deepak Jain
Department of Computer Science and Engineering, University of Kota, Rajasthan,
India
• Dr. S. Srithar
Assistant Professor, Computer Science and Engineering, RVS College of Engineering
and Technology, Coimbatore, Tamil Nadu, India

Associate Editor

• Dr. Gaurav Jangra


Assistant Professor, Synetic Business School, Ludhiana, Punjab, India
Email : gaurav.jangra@sbs.ac.in

viii
• Dr. Sheshang D. Degadwala
Head of Computer Department, Sigma Institute of Engineering, Vadodara, Gujarat,
India
Email : sheshang.cs.engg@sigma.ac.in
• Dr. K. Senthil Kannan
Dean Research, Vice Principal & Research Scientist EDAYATHANGUDY GS PILLAY
Arts and Science College Nagapattinam Tamil Nadu, India
Email : senthilkannan@egspec.org

International Editorial Board

• Dr. Andie Siahaan


Computer Science, Universitas Pembangunan Panca Budi, Indonesia
• AbdulGaniyu Abdu Yusuf
Computer Science, National Biotechnology Development Agency (NABDA), Abuja,
Nigeria
• Dr. M. A. Ashabrawy
Computer Science and Engineering, Prince Sattm bin Abdulaziz University, Kingdom
Saudi Arabia(KSA), Saudi Arabia
• Dr. V. Balaji
Bahir Dar University, Bahir Dar, Ethiopia
• Lusekelo Kibona
Department of Computer Science, Ruaha Catholic University (RUCU) , Iringa,
Tanzania
• Md. Amir Hossain
IBAIS University/Uttara University, Dhaka, Bangladesh
• Mohammed Noaman Murad
Department of Computer Science, Cihan University Erbil, Kurdistan Region, Iraq
• Prof. Dr. H. M. Srivastava
Department of Mathematics and Statistics, University of Victoria, Victoria, British
Columbia, Canada
• Prof. Sundeep Singh
Mississauga, Ontario, Canada
• Dr. Froilan Mobo. MOBO, DPA, MBA, MSCS, LPT
Computer and Social Sciences, Carl Balita Review Center : Olongapo City, Zambales,
Philippin
• Dr. Vaibhav Sundriyal
Department of Computer Engineering, Old Dominion University research
Foundation, USA (Phd in Computer Engineering from Iowa State University, USA)

ix
• Dr. Katarzyna Pietrucha-Urbanik
Faculty of Civil and Environmental Engineering, Rzeszow University of Technology,
Poland
• Indrasen Poola
Data Scientist & Artificial Intelligence - Industry Consultant & Professional
Researcher, Palo Alto Networks, Santa Clara, CA, USA
• Dr. Mohd Muntjir
Information Technology Department, College of Computers and Information
Technology, Taif University, Al-Hawiya, Taif, Kingdom of Saudi Arabia
• Vaibhav Sundriyal
Research Scientist, Old Dominion University research Foundation, USA
• Dr. Elsadig Gamaleldeen
Assistant Professor, Omdurman Ahlia University, Sudan
• SK. Saidhbi
Assistant Professor, Department of Information Systems, University of Gondar,
Ethiopia

Editorial Board - India

• Prof. Shardul Agravat


Information Technology, C U Shah Unuiversity, Surendranagar, Gujarat, India
• Prof. Sarita Dhawale
Ashoka Center for Business & Computer Studies, Ashoka Marg, Ashoka Nagar,
Nashik, Maharashtra, India
• Dr. Syed Umar
Professor & R & D Head, Department of Computer Science and Engineering, MLR
Institute of Technology, Hyderabad, Telangana, India
• Prof. C. K. Raina
HOD, Computer Science and Engineering Department, Adesh Institution of
Technology Chandigarh, Gharuan, Chandigarh, India
• Dr. R. Gayathri
Assistant Professor, Department of ECE, Annamalai University, Chidambaram, Tamil
Nadu, India.
• Dr. R. Reka
Professor & Head, Department of Computer Science & Engineering, Annai
Mathammal Sheela Engineering College, Tamil Nadu, India
• Dr. Nilamadhab Mishra
Post Graduate Teaching & Research Department, School of Computing, Debre Berhan
University, Ethiopia, India

x
• Dr. S. Ramkumar
Department of Computer Applications, Kalasalingam University, Krishnankoil,
Sirivilliputhur, Palani, Dindugal, Tamil Nadu, India
• G. Naveen Balaji
Department of ECE, SNS College of Technology, SNS Kalvi Nagar, Vazhiyampalyam
Pirivu, Coimbatore
• Dr. G. Rajkumar
Department of Computer Applications, N. M. S. S.Vellaichamy Nadar College,
Madurai, Tamilnadu, India
• Prof. Sagar Bhilaji Shinde
E & TC Engineering Department, JSPM, Narhe Technical Campus, Pune,
Maharashtra, India
• Prof Sarita Dhawale
HOD, Computer Science Department, Ashoka Education Foundation’s, Ashoka
Center for Business and Computer Studies, ACBCS, Nashik, India
• Prof. Umesh Sehgal
Associate Professor, GNA University, Phagwara, Punjab, India
• Prof. Vaishali Kalaria
Information Technology, RKU, Rajkot, Gujarat, India
• Prof. H. B. Jethva
Computer Enginering, L. D. College of Engineering, Ahmedabad, Gujarat, India
• Prof. Bakul Panchal
Information Technology, L. D. College of Engineering, Ahmedabad, Gujarat, India
• Prof. Bhavesh Prajapati
Computer Science, Government MCA College Maninagar, Ahmedabad, Gujarat, India
• Dr. Syed Umar
Computer Science and Engineering, KL University, Guntur, Andhra Pradesh, India
• Prof. S. Jagadeesan
Computer Science, Nandha Engineering College Erode, Tamil Nadu, India
• Prof. Joshi Rahul Prakashchandra
Information Technology, Parul Institute of Engineering & Technology, Vadodara,
Gujarat, India
• Dr. Aftab Alam Tyagi
Department of Mathematics, SRM University NCR Campus, Uttar Pradesh, India
• Prof (Dr.) Umesh Kumar
Department of Science & Technology, Govt. Women’s Polytechnic, Ranchi,
Jharkhand, India
• Dr. N. Pughazendi
Computer Science and Engineering, Panimalar Engineering College Chennai,
Tamilnadu, India

xi
• Sachin Narendra Pardeshi
Department of Computer Engineering, R.C.Patel Institute of Technology, Shirpur,
Maharashtra, India
• Dr. Bangole Narendra Kumar
Department of Computer Science and Systems Engineering, Sree Vidyanikethan
Engineering College, Tirupati, Andhra Pradesh, India
• Dr. Dhananjaya Reddy
M.Sc.(Maths),M.Sc.(Stat), M.Phil.(Maths), M.phil. (Statistics), B.Ed., PhD Department
of Mathematics,Government Degree College, Puttur, Chittoor, Andhra Pradesh, India
• Dr. Ajitesh Singh Baghel
Department of Computer Science, Awadhesh Pratap Singh University, Rewa, Madhya
Pradesh, India
• Dr. Dharmaraj K Veer
Director, UGC-INFLIBNET SOUL Co-Ordinator, Maharashtra & Goa Region,
Knowledge Resource Center, Dr. Babasaheb Ambedkar Marathwada University,
Aurangabad, Maharashtra, India
• Rakesh Oruganti
Assistant Professor Jayamukhi Institite of Technology and Science, Chennaraopet
Mandal, Narsampet, Nekkonda Rd, Arshanapally, Telangana, India
• Dr. M. Ramesh Kumar
Associate Professor/Project Coordinator, Department of Computer Science and
Engineering, VSB College of Engineering Technical Campus, Coimbatore, Tamilnadu,
India
• Dr. Ikvinderpal Singh
Department of Computer Science & Applications, Trai Shatabdi GGS Khalsa
College,Amritsar, Punjab, India
• Dr. Kshitij Shinghal
Associate Professor & Dean Academics, Department of Electronics & Communication,
MIT, Moradabad, Uttar Pradesh, India
• Udit Narayan Bera
Department of Electronics and Communication Engineering, St. Aloysius Institute of
Technology, Jabalpur, Madhya Pradesh, India
• Dr. Nageswara Rao Moparthi
Department of Computer Science and Engineering, Velagapudi Ramakrishna
Siddhartha Engineering College, Andhra Pradesh, India
• Dr. Sharad Gangele
Professor and Head, Department of Computer Science & Application, RKDF
University, Bhopal, Madhya Pradesh, India
• Dr. K. Praveen Kumar
Associate Professor and Vice-Principal in Chaitanya Institute of Technology and
Science, Kishanpura, Hanamkonda, Telangana, India
xii
• Dr. B. Bazeer Ahamed
Associate Professor, Department of Computer Science & Engineering, Balaji Institute
of Technology and Science, Warangal, Telagana, India
• Dr. Anil Kumar Dubey
Associate Professor & Deputy HOD (Academics), Computer Science, Poornima
Institute of Engineering & Technology, Jaipur, Rajasthan, India
• Dr. Yogesh M. Rajput
College of Computer Science and Information Technology (COCSIT), Ambajogai
Road, Latur, Maharashtra, India
• Dr. K. Umapathy
Associate Professor, ECE, SCSVMV University, Kanchipuram, Tamil Nadu, India
• K. Gutnadha Gupta
Assistant Professor of Computer Science and Enineering, Sri Indu College of
Engineering and Technology, Hyderabad, Telangana, India
• Namrata Singh
Computer Science and Engineering, Abes Engineering College, Ghaziabad, Uttar
Pradesh, India
• Dr. K. Nehru
Professor, Electronics and Communication Engineering Department, Institute of
Aeronautical Engineering, Hyderabad, Telangana, India
• Dr. Radha Mothukuri
Assistant Professor, Department of Computer Science and Engineering, KL
University, Vaddeswaram, Guntur, Andhra Pradesh, India
• Dr. K. R. Ananth
Department of MCA, Velalar College of Engineering and Technology, Erode,
Tamilnadu, India
• Rydhm Beri
Assistant Professor, PG Department of Computer Science, BBK DAV College for
Women, Amritsar, Punjab, India
• S. Kannadhasan
Department of Electronics and Communication Engineering, Tamilnadu Polytechnic
College, Austinpatti, Madurai, Tamil Nadu, India
• Vishal Kanjariya
Assistant Professor, Information Technology Department, Birla Vishvakarma
Mahavidyalaya (BVM), Engineering College, Vallabh Vidyanagar, Managed by
Charutar Vidyamandal, Anand, Gujarat, India
• Dhairya J. Vyas
Assistant Professor, EC Department, Sigma Institute of Engineering, Vadodara,
Gujarat, India

xiii
• Arpana Mahajan
Assistant Professor, Computer Department, Sigma Institute of Engineering, Vadodara,
Gujarat, India
• Dr. M. Anjan Kumar
Associate Professor, CSE Department, Vivekananda Institute of Technology &
Science, Karimnagar, Telangana, India
• Dr. A. Vamshi Krishna Reddy
Assistant Professor, Centre for Environment, Institute of Science & Technology,
Jawaharlal Nehru Technological University, Hyderabad, Telangana, India
• Dr. Ashok Koujalagi
Assistant Professor & Postdoctoral Researcher, P. G. Department of Computer
Science, Basaveshwar Science College, Bagalkot, Karnataka, India
• Dr. C. S. KanimozhiSelvi
Professor, Computer Science and Engineering, Kongu Engineering College,
Perundhurai, Erode, Tamil Nadu, India
• Dr. Mukhdeep Singh Manshahia
Assistant Professor (Computing), Department of Mathematics, Punjabi University,
Patiala, Punjab, India
• Dr. Nageswara Rao Moparthi
Department of Computer Science and Engineering, Velagapudi Ramakrishna
Siddhartha Engineering College, Andhra Pradesh, India
• Dr. Parshuram Singh
Associate Professor, Dean, HOD (Physics), Mahaveer Institute of Technology &
Science, MITS-Jadan (A Unit of Vidhya Jyothi Trust, Chennai & Pali), NH-14, Raniya
Bera, Jadan, Pali, Rajasthan, India
• Dr. S. Sasikala
Associate Professor, Department of Computer Applications, Hindusthan College of
Arts and Science, Coimbatore, Tamil Nadu, India
• Dr. G. Maria Priscilla
Professor & Head, Department of Computer Science, Sri Ramakrishna Colleges of Arts
and Science (formerly SNR Sons College), Coimbatore, Tamil Nadu, India
• Dr. M. Jayakameswaraiah
Assistant Professor, Department of CSA, Reva University, Bangalore, Karnataka, India
• Kishor Kumar Gajula
Research & Development Head, Newzen Infotech, Hyderabad, Telangana, India
• Mallavolu Malleswararao
Associate Professor & HOD, Department of Electrical & Electronics Engineering,
RISE Prakasam Group of institutions, Ongole, Andhra Pradesh, India
• Radha Mothukuri
Assistant Professor, Department of Computer Science and Engineering, QIS Institute
of Technology, Ongole, Andhra Pradesh, India
xiv
• Prof. S. Raja
Assistant Professor and Research and Development Coordinator, Department of ECE,
Sri Shakthi Institute of Engineering and Technology, Coimbatore, Tamilnadu, India
• Dr. J. Viji Gripsy
Assistant Professor, Department of Computer Science, PSGR Krishnammal College for
Women, Peelamedu, Coimbatore, Tamil Nadu, India
• Dr. Dickson S
Assistant Professor of Mathematics, Vivekanandha Educational Institutions
Tiruchengode, Tamil Nadu, India
• Dr. Ranjana Rajnish
Assistant Professor, Amity Institute of Information Technology(AIIT), Amity
University, Lucknow, Uttar Pradesh, India
• Dr. Nirajkumar C. Mehta
Associate Professor, Institute of Technology and Management Universe, Vadodara.
Gujarat, India
• Pulla Reddy Mekala
Southern Region Farm Machinery Training & Testing Institute, Tractor Nagar,
Garladinne, Anantapur (Dist.), Andhra Pradesh, India
• Dr. Rahul Saha
Division of Research and Development, Lovely Professional University, Punjab, India
• Deepak Jain
Department of Computer Science and Engineering, University of Kota, Rajasthan,
India
• Dr. S. Srithar
Assistant Professor, Computer Science and Engineering, RVS College of Engineering
and Technology, Coimbatore, Tamil Nadu, India

xv
CONTENT
Sr. No Article/Paper Page No
Automatic Tool for Prediction of Type of Cancer Risk and Recommendations
1 01-08
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi
Movie Recommended System by Using Collaborative Filtering
2 09-15
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar
Comparison and an Improved Validation Optimistic Approach for Concurrency
3 Control 16-21
Monika Patel, Dr. Dhiren B Patel
Airline Data Analysis
4 22-29
Navuluri Madhavilatha, Bheema Shireesha, Chunduru Anilkumar
Li-Fi the Next Generation of Wireless Communication through Visible Light
Communication (VLC) Technology
5 30-37
Imran Siddique, Muhammad Zubair Awan, Muhammad Yousaf Khan, Azhar
Mazhar
Demand Supply Oriented Taxi Suggestion System for Vehicular Social
6 Networks with Fuel Charging Mechanism 38-44
Selvi C, Keerthana D
Aspect Ranking Technique for Efficient Opinion Mining using Sentiment
7 Analysis : Review 45-49
Prof. Sonali D. Borase, Prof. Prasad P. Mahale
Big Data Anonymization in Cloud using k-Anonymity Algorithm using Map
8 Reduce Framework 50-56
Anushree Raj, Rio G L D'Souza
Improvement of Privacy and Security in Hybrid Cloud with Attribute Group
9 Based Access Control 57-62
Kayalvili S, Sowmitha V
Survey on Factors Determining the Share Price Movement in Indian Stock
10 Market - A Big Data Approach 63-69
Nitha K. P., Dr. S. Sivakumari
Enhancing Privacy Preserving Query Retrieval Using Locality Sensitive
11 Hashing 70-73
Archana M.S, K.Deepa
Enhanced Security for Dynamic Multi Keyword Ranked Search Using Greedy
12 Best First Search 74-78
Bibin Baby, Sharmila Banu
13 Cyber Security, Challenges, Some Practical Solutions 79-87

xvi
Y. V. Sai Bharadwaj, Sai Bhageerath Y.V, Prof. Y.V.S.S.S.V. Prasada Rao
Prediction of Fish Yields in Lakes and Reservoirs from simple Empirical Models
14 using Artificial Neural Network (ANN) : An Review 88-100
D. Karunakaran, M. Balakrishnan
Evolving Cryptographic Approach for Enhancing Security of Resource
15 Constrained Mobile Device Outsourced Data in Cloud Computing 101-106
K. Kaviya, K. K. Shanthini, Dr. M. Sujithra
Semantic Web Search Engines : A Comparative Survey
16 107-115
Anita Kumari, Dr. Jawahar Thakur
Image Dehazing Technique Based On DWT Decomposition and Intensity
17 Retinex Algorithm 116-122
Sunita Shukla, Prof. Silky Pareyani
Data Extraction in Malignant Environments
18 123-130
Amith Manda, Dr. R. P. Ramkumar
Diagnosis of Various Thyroid Ailments using Data Mining Classification
19 Techniques 131-136
Umar Sidiq, Dr. Syed Mutahar Aaqib, Dr. Rafi Ahmad Khan
Analysis of An Indoor Air Quality Antenna for Asthma Management in
20 Children 137-142
N. Vijayalakshmi, J. G. Rohini
Home Automation using Google Voice Assistance with Raspberry Pi
21 143-146
Undra Narasimha Babu, S. N. Sreekanth
Sentiment Analysis of movie reviews using Microsoft Text Analytics and
Google Cloud Natural Language API
22 147-150
S Jaswanth Sai Chowdary, Sk Waseem Farooq, Sk Abubakar Siddique, S
Dinesh
Blockchain and its Role in the Internet of Things (IoT)
23 151-157
Tanweer Alam
A Review on Various Payment and Billing Techniques
24 Alka D. Aswar, Diksha M. Chhipane, Dipali C. Sontakke, Pallavi P. Onkar, 158-162
Snehal S. Helonde, Prof. S. V. Sonekar
Finite/Infinite Queueing Models Performance Analysis in Optical Switching
25 Network Nodes 163-170
Waleed M. Gaballah
Improving Central and Remote Access Units in Cloud Based Wi-Fi Network for
26 Better Client and Service 171-179
R. Kavitha, G.Sathiyabama
27 Effective Prediction for Rock Burst Dataset Using Classification Algorithms 180-184

xvii
with Particle Swarm
Sadesh S, Banupriya GK
Enhancement in the Lifetime of Wireless Sensor Network
28 185-190
Bhumika Yadav, Ram Neresh Sharma
A Review Paper on IoT Based Smart Irrigation System
29 Anjali Dokhande, Chetna Bomble, Rakshanda Patil, Puja Khandekar, Nayan 191-196
Dhone, Prof. Chandrashekhar Gode
Marg-Sudhar : Pothole Reporting System Using Android Smart Phone
30 Riya Agrawal, Payal Patel, Yash Telang, Ankita Bhandarkar, Irfan Mushtaq, 197-202
Dr. S. V. Sonekar
Smart Meals
31 Shivali Singh, Aishwarya Narnaware, Kirtikumar Jain, Kash Patel, Prof. S 203-206
Dhawale
A Review on Different Face Recognition Techniques
32 Aishwarya Admane, Afrin Sheikh, Sneha Paunikar, Shruti Jawade, Shubhangi 207-213
Wadbude, Prof. M. J. Sawarkar
A Survey on Various Bio-Metric Techniques and its Application for Child Care
33 Shanon Paul, Payal Ramteke, Sugam Sangai, Minakshee Shende, Aafreen 214-220
Qureshi, Prof. Umesh Samarth
Liver Cancer Detection
34 221-228
L. S. Rohith Anand, B. Shannmuka, R. Uday Chowdary, K. Satya Sai Krishna
Emotion Detection using Facial expressions with Transfer Learning
35 Sairama Geetanand, Shaik Nazeer Basha, Sajja Nageswara Rao, Rayudu 229-234
Srinivas
A Review on Various Algorithms for Student Performance Prediction
37 Mayuri S. Dongre, Neha S.Vidya, Vanita D. Telrandhe, Shweta R. Chaudhari, 235-239
Anup A. Umrikar, Prof. Prachiti V. Adghulkar
E-Toll System Using RFID
38 Ankita Bambole, Shrishti Thakur, Suvidha Ramteke, Sneha Borkar, Hitesh 240-243
Kasturi, Prof. M. M. Baig
A Review on Indoor Navigation Systems and Technologies for Guiding the
Users
39 244-250
Meenakshi D. Dhobale, Pooja S. Verma, Pushpa R. Joshi, Jyotsna K. Tekade,
Swati Raut
A Review Study on Online Job Portal
40 Aafreen Khan, Anirudha Wankhade, Priyanka Pakhide, Sagar Meshram, Prof. 251-255
Sonali Zunke
41 Voice Based Navigation System for the Blind People 256-259

xviii
N. Vijayalakshmi, K. Kiruthika
Data Encryption Key Sharing Using Image Pixel Color Value
42 260-265
Vijay Gokul Koli, Raj Kumar Paul
A Review on Prediction of Diabetic Retinopathy Using Data Mining
43 Algorithms 266-272
Kajal Sanjay Kothare, Prof. Kalpana Malpe
A Review on Predicting User Behaviour through Session Using Web Mining
44 Veena Moharle, Sonali Waghade, Divyani Dhoke, Sheetal Kolhe, Snehal 273-277
Narule, Prof. Merajul Haque
Elimination of Impulse Noise using Mean Median filter for Image Enhancement
45 278-288
Trupti Arun Jangale, Raj Kumar Paul
E - Tax Information System
46 289-291
M. Florence Dayana, B. Vinithamani
Children Tracking Classification Control Using Android
47 292-295
M. Juno Isabel Susinthra, S. Praveena
Voice Based Information Sharing System for Visually Impaired People
48 296-298
R. Gomathijayam, A. Jenifer
A Review on Big Data Analytics Via Social Media
49 299-302
T. K. Anusuya, A. Preethi
Automatic Railway Gate Control System Using RFID with High Speed Alerting
50 System 303-306
M. Florence Dayana, R. Bhuvaneshwari
Design of Robot for Scrambling on Tree and Clutching Fruit Using Foundary
51 Design 307-311
V. Manochithra, S. Karthika
Spinal Muscular Atrophy Type B Awareness And Symptoms Prediction Using
52 3d Segmentation Process In Mri Images 312-318
V. Manochithra, G. Sumithra
A Survey of Telemedicine Services using Datamining
53 319-321
T. K. Anusuya, P. Maharajothi
Medical Product Verify and Complaint Management System for Ngo's
54 322-326
R. Gomathijayam, J. Ayeesha Parveen
A Survey on Child Safety Wearable Device to Prevent Child Trafficking Using
55 Arduino 327-333
Elakiya M., S. Radhika
Bus Notification System Using Crowdsourcing
56 334-339
B. Harshini Reddy, K. Jhansi Rani, M.Vaishnavi
57 An Extension to Traditional Qwerty Cryptographic Method 340-342

xix
V. Vennila, S. Umamaheswari
A Review on Hadoop Eco System for Big Data
58 343-348
Anushree Raj, Rio D'Souza
A Novel Approach on Prevention of a Mobile Exploration in Machine Using AI
59 349-356
J. Antony Daniel Rex, S. Dinakaran, S. Muthukumaran
Tuition Plus+ For Managing Tuition Class Data
60 Bharat Dhak, Prachi P. Paunikar, Simran S. Khan, Snehal S. Ramteke, 357-362
Yashkumar R. Jais, Priya R. Gupta
Heikin-Ashi Transformation and Vix Index data for Stock Market Index
61 Prediction and It's Effects 363-365
Nishchal Sharma, Dr. Chaman Singh Chauhan
Factors Affecting the Successful Implementation of Cloud and IoT - A Study on
62 IT and Business Leaders 366-396
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh
Development of Remote Operated Agri Vehicle
63 397-403
Juno Isabel Susinthra M, K. Gayadri
Adblock Usage in Web Advertisement
64 404-408
V. Santhi, S. Abirami
Smart-Home Automation Using IoT-Based Sensing and Monitoring Platform
65 409-413
R. Sugantha Lakshmi, P. Karthika, A. Rajalakshmi, M. Sathya
Anticipated Security Model for Session Transfer and Services Using OTP
66 414-418
J. Jose Merlin, A. Prathipa, G. Ramyadevi, P. Radhika
Managing university defined Study material by using Online Portal
67 419-421
Neha Bitodkar, Prof. Aniket Bhoyar, Suraj Chavhan, Priyanka Wahane
A Review on Car Collision Avoidance System based on Interpretation of
Vehicle Event Data Recorder
68 422-428
Prof. Tejal Irkhede, Rohit Kashayp, Nikhil Kumar, Vivek Ranjan, Manish
Kumar
A Review on Online Medical Drug Tracking
69 Swati Hedau, Ankita Shegaonkar, Amruta Aknurwar, Devyani Vaidya, Mehul 429-432
Sahu, Prof. Jogi John
A Survey on Secure User Data Transaction using Attribute - Based Encryption
Scheme over Cloud Data
70 433-437
Navin Sethiya, Hrishikesh Patel, Akshay Harshe, Alekh Gaigole,
Harshvardhan Donadkar, Prof. Priya Karemore
Accessing Network Using Secure Android Applications
71 438-442
R. Gomathyjayam, R. Thilakavathi
72 Web Site Ranking Feedback System 443-447

xx
Shaik Muzammil, Sai Kiran Yerramaneni
IoT Based Heart Attack Early Prediction
73 Shweta Gajbhiye, Bharati Vyas, Shrushti Shrikhande, Anshika Janbandhu, 448-451
Komal Nagpure, Mrunali Agashe
Shakti : An Android App for Mapping of Crime against Women
74 Diksha Ingle, Pritee Dhale, Priyanka Palsapure, Sagar Pazare, Dhawal 452-455
Chimurkar, Niket Farfad, Prof. Prasanna Lohe
Robust Video Data Hiding in Forbidden Zone
75 Ruchika S. Bhambore, Ashwini B. Gurudeo, Shital B. Borkar, Pallavi G. Thul, 456-461
Chanchala V. Udepurkar, Prof. Anup Bhange
The Internet-of-Things (IoT) Security : A Technological Perspective and
Review
76 462-482
Dr. Yusuf Perwej, Firoj Parwej, Mumdouh Mirghani Mohamed Hassan,
Nikhat Akhtar
Secure Micro Payment Using Physical Unclonable Function
77 483-486
M. Jalasri, S. Nalini, N. Magesh Kumar, J. Elumalai
Reversible Data Hiding in Homomorphic Encrypted Domain by Monitoring
78 Ciphertext Group 487-490
V. Santhi, M. Abinaya
A Review : Venue Booking System
79 Shruti N. Ghule, Harish M. Mohurle, Akshay P. Shegaonkar, Zeba Parvin 491-494
Abdual Saleem, Pradhnya S. Borkar
A Review on IOT Based Garbage Monitoring System
80 Himanshu Gupta, Amit Chaturkar, Abhay Bagde, Hritik Raj, Ankur Hemane, 495-500
Prof. Prerana Deshmukh
A Survey on Rice Crop Yield Prediction in India Using Improved Classification
81 Technique 501-507
Kolin Sukhadia, M. B. Chaudhari
A Review on Various Road Safety Systems
82 Chandrashekhar Rathod, Sneha Wakde, Roshan Punwatkar, Rupesh Sharma, 508-513
Rupal Patil, Prof. Rohan Kokate
Review on Various Tools and Techniques for Plagiarism Detection
83 Kedar Bhujade, Tushar Bhandirge, Rohan Khobragade, Saurabh Shelare, 514-518
Jayprakash Shukla, Prof. R. S. Thakur
An Elementary Study on Various Techniques Involved in Face Recognition
84 Systems : A Review 519-525
Payal Maken
85 E Return Tool 526-530

xxi
Pooja Aher, Manish Kirnapure, Shubhangi Nemade, Tejaswita Shrivastava,
Lulkaran Chohale
IoT Based LPG Gas Leakage Detector
86 Prof. Pranay Meshram, Nancy Shukla, Stuti Mendhekar, Renuka Gadge, 531-534
Shivani Kanaskar
Machine learning algorithm for Cyber Security - A Review
87 535-545
Mohammad Asif, Pratap M. Mohite, Prof. P. D. Satya
Analysis and Importance of Deep Learning for Video Aesthetic Assessments
88 546-554
Manish Bendale, Madhura V. Phatak, Dr. Nitin N. Pise

xxii
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1838116

Automatic Tool for Prediction of Type of Cancer Risk and


Recommendations
Pallavi Mirajkar1, Dr. G. Prasanna Lakshmi2
1Research Scholar, Faculty of Computer Science, Pacific Academy of Higher Education and Research University,
Udaipur, Rajasthan, India
2(WOS-A) Andhra University, AU North Campus, Visakhapatnam, Andhra Pradesh, India

ABSTRACT

Cancer can begin in any part of the body and can spread to other parts also. It is uncontrollable and it has many
types. In the proposed thesis research paper, a tool for prediction of type of cancer risk with five different
cancer diagnosis and recommendations is presented. For recognizing cancer disease number of tests ought to be
required from the patient. But using data mining techniques these test can be diminished. Indeed, an accurate
prediction of cancer is very difficult task for medical practitioner and it is also high concern to the patients so
that better treatment can be given and it will also increase the survival time of the patients. Our findings
suggested that suitable prediction tool can effectively reduce the several tests for diagnosing cancer and
prediction accuracy thereby increasing the technical possibility of early detection of cancer. The main features
of the tool comprise a balance between the number of necessary inputs and prediction performance, being
portable, and it empowers the automatic development of the cancer risk prediction tool in cancer disease.
Keywords : Prediction Tool, Cancer, Data Mining, Automation, Integration.

I. INTRODUCTION able to consult with proposed prediction tool, we


chose to develop cancer prediction tool as an online
The rising high-performance computing has system.
benefited numerous disciplines in finding realistic
solutions to their issues. Our health services are no Here the scope of the prediction tool is that
special case to this. Data mining tools have been integration of various risk factors that causes cancer,
created for useful investigation of medical with computer-based patient records could reduce
information, to help oncologist in improving medical errors, enhance patient safety, improves the
determination for treatment purposes. In cancer prediction of cancer risk in practice variation, and
disease research, data mining technique have played improve patient survival rate. The application is fed
out a noteworthy role. Cancer disease categorization with different details to help medical practitioner to
contributes the unsafe reason for the treatment of predict risk of type of cancer. The application permits
patients. user to share their health connected issues. It then
The aim of thesis work is to present an easy to use processes user’s particular details to determine for
tool that provides predictions of cancer risk in varied cancer disease that might be related to it. Here
patients or individual. Due to the internet facility we tend to utilize some intelligent data mining
available everywhere and the ease with which one is techniques to figure the correct risk level of cancer

CSEIT1838116 | Received : 28 Dec 2018 | Accepted : 03 Jan 2019 | January-February -2019 [ 5 (1) : 01-08 ]
01
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

that may be identified with patient’s details. Based on paper work is very helped to the cancer research
result, system automatically demonstrates the result center and hospitals to prevent the Non Small Cell
to specific doctors for specific test/ treatment. Lung Cancer.

II. LITERATURE REVIEW Arpit Bansal et, al.(2017) have proposed technique for
a modification in K-Means Clustering Algorithm. In
Fan Wang et.al (2018) worked on predictor to proposed modification, the K-Means clustering will
precisely predict six breast cancer cell types using vanish off the two most important drawbacks of K-
gene expression. They identified 308 predictors of Means clustering that are accuracy level and
breast cancer cell types using machine learning. calculation time consumed in clustering the dataset.
They found that mixture of protein coding genes and Although when they have also used small datasets
long non-coding RNAs are superior predictors these two factors accuracy level and calculation time
compared to when the two sets of transcripts are may not matter much but when they used large
treated independently. They have also shown that datasets that have trillions of records, then little
predictors constrained to a particular cell type dispersion in accuracy level will matter a lot and can
provide as better prognostic markers for the relevant lead to a disastrous situation, if not handled properly,
patient subtype. Author has stated that the breast it can be stated that proposed modification can be
cancer cell type predictors are also related with more extended to achieve the full accuracy level up
patient survivability and therefore have clinical to 100%, with very little time and with more quality
significance. clusters.

Subrata Kumar Mandal (2017) have applied different R.Senkamalavalli and Dr.T.Bhuvaneswari (2017) have
techniques such as data cleaning, feature selection, proposed novel algorithm was experimented on the
feature extraction, data discretization and Breast cancer database. It has been proved that
classification for predicting breast cancer as perfectly. proposed approach achieved a very high accuracy
They stated that Logistic Regression Classifier gives rate than the existing methods. They also
the maximum accuracy with reduced subset of demonstrated a certain level of accuracy in the
features and time complexity of proposed algorithm is classifier, and for finding accurate results there must
least compared to other two classifiers. be sufficient preprocessing of data done. They have
also demonstrated accuracy in diagnosing breast
R.Kaviarasi, Dr.A.Valarmathi (2017) has applied two cancer disease using the K-means classifier, adaboost
types of clustering. The hierarchical clustering is and Support Vector Machines.
produced dendrogram results are produced using
Euclidean distance and Ward.D linkage. The K- Sumalatha.G, Archana.S (2017) has analyzed cancer
Means clustering are produced WSS values against patient data using ZeroR method and J48 algorithm
number of cluster K values. The proposed method of data mining techniques. They stated that proposed
finally validate to the two type of clustering fit values. cancer risk prediction system may give easy and a
The validation measurement result is helped to the cost effective approach for screening cancer and may
distance are measured in two clustering values. The play a significant role in earlier diagnosis process for
results are helped to at the beginning of Non Small different types of cancer. Proposed system can also be
Cell Lung Cancer prevention through human way of used as a source of record with detailed patient
life handled risk factors characteristics. The proposed

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


11
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

history in hospitals as well as help doctors to III. PROPOSED WORK


premeditate on particular therapy for any patient.
Diagnosis of cancer is still challenging for the
Megha Rathi et.al. (2016) have developed a software specialists in the field of medicine. Even now the
tool for the prediction of disease which helps in genuine reason and complete treatment of cancer is
decision making for the treatment method. Proposed not invented. Different tests are available for
tool will be helpful in diagnosing the disease type and predicting cancer, however detecting cancer in
to help out for decision support in medical system. earlier stage is troublesome, yet prior discovery of
They studied hybrid classification technique is used cancer disease is treatable. With the help of data
for classification of medical data sets and is applicable mining we endeavor to predict the risk of cancer in
in healthcare domain. SVM and bootstrap is earlier stage. We build up a system called the cancer
integrated to improve classification accuracy. Author prediction tool which predicts specific cancer risks.
expresses that prediction tool will help doctors or Specifically, Cancer prediction tool estimates the risk
patient to decide in a short time whether the person of the breast, ovarian, oral, stomach, and lung cancers
is suffering from disease and is generic to all types of by analyzing a number of user-provided genetic and
disease. non-genetic factors. The major objective of
prediction tool to provide the earlier warning to the
Prabhakar Chalise et,al.,(2014) have proposed Cluster users, to make a precaution based on their risk status.
analysis aims to highlight meaningful patterns or The proposed prediction tool assists the people to
groups inherent in the data that will be useful in know their chance of cancer risk with ease and
identifying the subtypes of the diseases. Different additionally, it helps the people to take the proper
types of clustering algorithms have been proposed decision based on their cancer risk status. Once the
that use various assays of molecular variation of cells user enters into the cancer prediction system, they
most of which are designed for one type of data at a need to answer the questionnaire, related to cancer
time. These types of methods have been successfully causing factors and symptoms. After that the
implemented in several disease classification studies. prediction system assigns the risk value based on the
user responses. Once the risk value is predicted, the
Rahul Patil et.al (2016) has used various clustering range of the risk can be computed by the prediction
techniques of data mining such as partition based system. We have three levels of risk low level,
method, hierarchical based clustering. They stated medium level, and high level. The result can be given
that hybrid method will help society to analyze and away to the user through data base. The above
understand disease and their health status. mentioned technique can be successfully applied to
the data sets for any cancer (such as breast cancer,
Tanu Minhas and Nancy Sehgal (2017) concluded lung cancer, oral, stomach cancer), as it is effectively
that prediction analysis is an efficient technique for checked on the different types of cancer. Once the
the complex data analysis. They have applied k- risk is calculated the data given by the user is stored
means clustering algorithm with Boltzmann in the data base. The result will be displayed to the
algorithm to increase accuracy of data clustering. user through the database.
They stated that the performance of proposed
algorithm is tested in MATLAB and it has been
analyzed that accuracy is increased up to 20 percent.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


12
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

check their risk of cancer. Collected data also


User Interface contains both patient and non patient data. From the
previous studies total 18 risk factors were considered
for risk level cancer assessment, which includes- age,
User Enter Personal gender, hereditary, occupational hazard, smoking,
Information
chewing tobacco, alcohol, daily exercise, intake of
User Fill Questionnaire vegetables, weight loss, weight gain, soft drink,
contraceptive pills and various symptoms of different
types of cancer.
User Select Organ of
Symptoms 4.3 Data Preprocessing
Sometimes it happens that collected data consists of
repetition of information or more information of
Compare the user input with same person or missing any values of information.
data stored in database Therefore, data preprocessing is vital task which is
necessary in data mining. The main objective of data
pre-processing is making an appropriate analysis of
Match the input value of user collected data. Data preprocessing allows avoiding
with stored data in database the double data and adds the missing values
according to the need of tool. It also helps to reduce
the memory and normalizes the values that are stored
into the database.

Store Computed
Score 4.4 System Architecture
The proposed cancer risk prediction tool is developed
Display Results using various web technologies such as HTML, CSS
and Java Script. It is suitable for several platforms and
Fig. 1 Working of Cancer prediction system is distributable through an application store, just like
other applications. The proposed tool is developed
IV. METHODS AND MATERIAL using Angular JS which wraps the HTML/JavaScript
app into a native container which can access the
4.1 Methodology device functions of multiple platforms. When the
Collected data is prepared to fit into the developed medical practitioner submits the values for multiple
tool appropriately. Here converse about the entire queries, the data is sent over the HTML POST
procedure (collection of data to significant pattern) of method. The data is inserted into a Postgre SQL
prediction of different types of cancer risk level. innodb database and, for each new registration. The
proposed tool generates new XML file for each type
4.2 Data Collection of cancer. For the health care industry, cancer
Collection of data for the developed prediction tool prediction applications yielded new boundaries in
has been done from different diagnostic centers providing better care and treatments to patients.
which contain both male female data of different ages Furthermore, it is building a revolution in the way
and from the various non-patients those who want to data is managed and made available [21]. The

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


13
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

portability of proposed prediction applications can


enhance the productivity of health care professionals.

Sends

Queries

Web Based Cancer


Risk Prediction Tool

Fig.4 Personal Information Tab


Updates

Updates
Queries

Uses

Fig. 2 Architecture of Cancer Risk Prediction


4.5 Use Case Tool

Fig. 5 Lifestyle Tab


Figure 2(a) shows the first screen that appears when
the prediction tool is initiated. User has to login and
then clicking on the Test Fig. 2, all tabs containing
questionnaire available in the application become
visible. User has to answer each question as shown in
fig. 4, fig. 5 fig. 6, fig 7and fig. 8.

Fig.6 Eating Habit Tab

Fig. 3 Home Page

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


14
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

Gender Responses
Male 76
Female 93
Total 169
Table 1

VI. CONCLUSION

The thesis paper in which I developed cancer


prediction system based on the online discourse of
patient’s information. It was shown that relying on
Fig. 7 Symptoms of different types of cancers cancer prediction system will give warning to the
user about cancer risk level. Combining the different
approaches boosts results, as content and interaction
seem to model complementary aspects of patient
discourse. The proposed prediction tool can be
helpful to health industry who wants to diagnose
patient online. The proposed system is web based,
easy to use, versatile and reliable that can be
implemented in remote areas like rural regions in
Primary Health center, to mimic like human
diagnostic expertise for finding the chances of having
cancer. The main aim of the proposed prediction tool
Fig. 8 Display of Result
is to provide the earlier warning to the users, and it
also saves cost and time of the user. It predicts
V. RESULTS AND DISCUSSION
specific cancer risks based on the symptoms found in
the patient. Specifically, Cancer prediction system
CANCER PREDICTION SYSTEM –
POPULATION AND SAMPLE RESULTS estimates the risk of the breast, Oral, Stomach,
Ovarian, and Lung cancers by examining a numerous
▪ To find out the performance of the Data Mining user-provided genetic and non-genetic factors. The
Based Cancer Prediction System among the users proposed prediction system is easily accessible;
based on cancer prediction. people can easily check their risk and take
▪ To identify cancer risk level of the user using appropriate action based on their risk status.
newly developed Data Mining Based Cancer
Prediction System based on various risk factors. Future Scope on the prediction tool can include the
development of different models with high end
To find the effectiveness of data mining based cancer technologies to prevent cancer.
prediction system, the proposed system was
implemented on web based technology. Around 169
responses have been collected during Jun to Dec 2018.
Details of the responses given in the table No. 1

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


15
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

VII. REFERENCES [9]. Ahmad LG, Eshlaghy AT, Poorebrahimi A,


Ebrahimi M, Razavi AR (2013) “Using Three
Machine Learning Techniques for Predicting
[1]. Fan Wang, Zachariah Dohogne , Jin Yang , Yu Breast Cancer Recurrence”. J Health Med
Liu and Benjamin Soibam “Predictors of breast Inform 4: 124. doi:10.4172/2157-7420.1000124
cancer cell types and their prognostic power in [10]. V.Kirubha , S.Manju Priya “Survey on Data
breast cancer patients” BMC Genomics (2018) Mining Algorithms in Disease Prediction”
19:137 https://doi.org/10.1186/s12864-018- International Journal of Computer Trends and
4527-y. Technology (IJCTT) – Volume 38 Number 3 -
[2]. Dave Smith, “Data Mining in the Clinical August 2016.
Research Environment”. Available at. [11]. P. Saranya , B. Satheeskumar “A Survey on
http://www.sas.com/ Feature Selection of Cancer Disease Using Data
[3]. Damtew A., “Designing a predictive model for Mining Techniques” International Journal of
heart disease detection using data mining Computer Science and Mobile Computing,
Techniques” A Thesis Submitted to the School Vol.5 Issue.5, May- 2016, pg. 713-719.
of Graduate Studies of Addis Ababa University, [12]. R.Senkamalavalli and Dr.T.Bhuvaneswari
2011. “Improved Classification Of Breast Cancer Data
[4]. https://www.cancer.gov/ Using Hybrid Techniques” International
[5]. Subrata Kumar Mandal “Performance Analysis Journal of Advanced Research in Computer
Of Data Mining Algorithms For Breast Cancer Science Volume 8, No. 8, September-October
Cell Detection Using Naïve Bayes, Logistic 2017 ISSN No. 0976-5697.
Regression and Decision Tree” International [13]. Sumalatha.G, Archana.S “A Study on Early
Journal Of Engineering And Computer Science Prevention and Detection of Breast Cancer
ISSN: 2319-7242 Volume 6 Issue 2 Feb. 2017, using Data Mining Techniques” International
Page No. 20388-20391. Journal of Innovative Research in Computer
[6]. R. Kaviarasi, Dr. A.Valarmathi “Near The and Communication Engineering (An ISO
Beginning of Non Small Cell Lung Cancer 3297: 2007 Certified Organization) Website:
Avoidance in Human Way of Life Risk Factors www.ijircce.com Vol. 5, Issue 6, June 2017.
Classification Using Clustering Algorithm in [14]. Megha Rathi , Vikas Pareek “Disease prediction
the R Environment” International Journal of tool: an integrated hybrid data mining approach
Advanced Research in Computer Science, 8 (5), for healthcare” IRACST - International Journal
May-June 2017,1023-1026. of Computer Science and Information
[7]. Arpit Bansal, Mayur Sharma “Improved k-mean Technology & Security (IJCSITS), ISSN: 2249-
clustering algorithm for prediction analysis 9555 Vol.6, No.6, Nov-Dec 2016.
using classification technique in data mining”, [15]. Rahul Patil, Pavan Chopade, Abhishek Mishra,
IJCA (0975 – 8887) Vol. 157 – No 6, January Bhushan Sane, Yuvraj Sargar “Disease
2017. Prediction System using Data Mining Hybrid
[8]. Prabhakar Chalise, Devin C.Koestler, Milan Approach” Communications on Applied
Bimali, Qing Yu, Brooke L.Fridley to Electronics (CAE) – ISSN : 2394-4714
“Integrative clustering methods for high – Foundation of Computer Science FCS, New
dimensional molecular data” , Transl Cancer York, USA Volume 4 – No.9, April 2016.
Res 2014.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


16
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

[16]. Tanu Minhas , Nancy Sehgal “Prediction Author’s :


Analysis Technique using Hybrid Clustering
and SVM Classification” International Journal Mrs. Pallavi Mirajkar passed M.Sc.
(Computer Science) from North
of Innovative Research in Science, Engineering
Maharashtra University, Jalgaon.
and Technology (An ISO 3297: 2007 Certified
Currently she is pursuing her Ph.D.
Organization) Website: www.ijirset.com Vol. 6, from Pacific University, Udaipur,
Issue 7, July 2017. Rajasthan, India
[17]. K. Suneetha “Early Prediction and Detection of
Lung Cancer using Data Mining” International Dr. G.Prasanna Lakshmi, Guide.
Journal of Advanced in Management,
Women Scientist (WOS-A) Andhra
Technology and Engineering Sciences Volume
University, Andhra Pradesh, India
7, Issue 12, 2017 ISSN NO : 2249-7455.
[18]. Chih-Jen Tseng, Chi-Jie Lu, Chi-Chang Chang,
Gin-Den Chen, Chalong Cheewakriangkrai
“Integration of data mining classification
techniques and ensemble learning to identify
risk factors and diagnose ovarian cancer Cite this article as :
recurrence” Contents lists available at Science
Direct Artificial Intelligence in Medicine Pallavi Mirajkar, Dr. G. Prasanna Lakshmi,
"Automatic Tool for Prediction of Type of Cancer
journal homepage:
Risk and Recommendations", International Journal of
www.elsevier.com/locate/aiim/ Artificial
Intelligence in Medicine 78 (2017) 47–54. Scientific Research in Computer Science, Engineering
and Information Technology (IJSRCSEIT), ISSN :
[19]. Huang M-W, Chen C-W, Lin W-C, Ke S-W,
2456-3307, Volume 5 Issue 1, pp. 01-08, January-
Tsai C-F (2017) “ SVM and SVM Ensembles in
February 2019.
Breast Cancer Prediction” PLoS ONE
Available at doi
12(1):e0161501.
: https://doi.org/10.32628/CSEIT1838116
doi:10.1371/journal.pone.0161501.
Journal URL : http://ijsrcseit.com/CSEIT1838116
[20]. Florije Ismaili , Luzana Shabani , Bujar Raufi ,
Jaumin Ajdari , Xhemal Zenuni “Enhancing
breast cancer detection using data mining
classification techniques” 2nd World
Conference on Technology, Innovation and
Entrepreneurship May 12- 14, 2017.
[21]. Ana Silva, Tiago Oliveira, Vicente Julian, Jose
Neves, and Paulo Novais “A Mobile and
Evolving Tool to Predict Colorectal Cancer
Survivability” Published by Springer
International Publishing Switzerland 2016. All
Rights Reserved L. Iliadis and I. Maglogiannis
(Eds.): AIAI 2016, IFIP AICT 475, pp. 14–26,
2016.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


17
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19511

Movie Recommended System by Using Collaborative Filtering


Bheema Shireesha1, Navuluri Madhavilatha2, Chunduru Anilkumar3
1,3 Assistant Professor, Department of Computer Science and Engineering, Dr. APJ Abdul Kalam IIIT Ongole,
Andhra Pradesh, India
2Guest Faculty, Department of Computer Science and Engineering, Dr. APJ Abdul Kalam IIIT Ongole, Andhra

Pradesh, India

ABSTRACT

Recommendation system helps people in decision making an item/person. Recommender systems are now
pervasive and seek to make profit out of customers or successfully meet their needs. Companies like Amazon
use their huge amounts of data to give recommendations for users. Based on similarities among items, systems
can give predictions for a new item’s rating. Recommender systems use the user, item, and ratings information
to predict how other users will like a particular item. In this project, we attempt to under- stand the different
kinds of recommendation systems and compare their performance on the Movie Lens dataset. Due to large size
of data, recommendation system suffers from scalability problem. Hadoop is one of the solutions for this
problem.
Keywords : Hadoop, Collaborative filtering, Machine Learning, HDFS, RSSI

I. INTRODUCTION then the user will simply stop using it. This has led to
a high emphasis by tech companies on improving
Recommendations system is a type of information their recommendation systems. However, the
filtering system which attempts to predict the problem is more complex than it seems. For
preferences of a user, and make suggests based on recommendation, our proposed system uses
these preferences. There are a wide variety of collaborative filtering machine learning algorithm.
applications for recommendation systems. These have Collaborative filtering (CF) is a machine learning
become increasingly popular over the last few years algorithm which is widely used for recommendation
and are now utilized in most online platforms that we purpose. Collaborative filtering finds nearest neighbor
use. The content of such platforms varies from movies, based on the similarities. The metric of collaborative
music, books and videos, to friends and stories on filtering is the rating given by the user on a particular
social media platforms, to products on e-commerce item. Different users give different ratings to items.
websites, to people on professional and dating Users, who give almost same rating to items, are the
websites, to search results returned on Google [1]. nearest neighbors. In case of User based collaborative
Due to the advances in recommender systems, users filtering, based on the ratings given by the users,
constantly expect good recommendations. They have nearest neighbors has been find. Item based
a low threshold for services that are not able to make collaborative filtering predicts the similarity among
appropriate suggestions. If a music streaming app is items. To recommend an item, items which are liked
not able to predict and play music that the user likes, by the user in his past have been found. Item which is

CSEIT19511| Received : 01 Jan 2018 | Accepted : 03 Jan 2019 | January-February -2019 [ 5 (1) : 09-15 ]
09
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar Int J S Res CSE & IT. January-February-2019 ; 5(1) : 09-15

similar to those items has been recommended [2]. measurements and analysis were made. The
Internet contains a huge volume of data for remaining part of this report is as follows: Related
recommendation purpose. Due to size of data, if Works, Measurement Setup, Proposed Work,
recommendation computation has been done in single Measurements Analysis, Results, Discussions, Future
system, then performance may degrade, and we Work.
cannot find an efficient solution. Hence we require
distributed environment so that computation can be II. RELATED WORK
increased and performance of recommendation
system gets improve .My goal is to apply a 2.1 Collaborative Filtering
collaborative filtering algorithm in a rating website
that collects users’ information, such as location and In the collaborative filtering algorithm, the system
gender, item’s information, such as category and has a recorded set of items and users and how the
description, as well as ratings for items by users. There users rated those items. Then algorithm is used to
are many algorithms that could be applied on data to predict the rating for a user who has not rated the
predict a user preference. User-based, Item-based, and item yet. A rating for an item can be predicted from
Model-based methods are ways of predicting a user the ratings given to the item by users who are similar
preference. The number of users, items, or clusters in in taste to the given user. The traditional collaborative
each one respectively will determine the function filtering algorithms include User-based, Item-based,
performance [3]. However, the most well-known and and Model based methods. To explain how these
common one is User-based Collaborative Filtering. In methods works we are going to use the following
this algorithm, we predict an item’s rate for a user by notations [6]. “Let U be a set of N users and I a set of
collecting information about this user and similar M items. The traditional collaborative filtering
users. algorithms include User-based, Item-based, and
Model based methods. To explain how these methods
Machine Learning: works we are going to use the following notations.
Applying machine learning in real-time using “Let U be a set of N users and I a set of M items. Vui
Collaborative Filtering. Parsing data retrieved from a denotes the rating of user u U on item i I, and SI
database and predicting user preference. Evaluating stands for the set of items that user u has rated”.
different approaches of recommender systems. What I
was trying to do was to build a system that collects 2.2 User Based Collaborative Filtering
information and then uses the stored data in a In this method, we predict the user behavior against a
machine learning algorithm [4]. Predicting users’ certain item using the weighted sum of deviations
preferences using data may give more accurate results from mean ratings of users that previously rated this
than any algorithm that does not use previous data. item and the user mean rate. The weight that we
Most systems like Amazon, eBay, and others suggest previously mentioned can be calculated using Pearson
things to users based on similarities among users, Correlation according to the following formula:
items, or both. This will make those systems more
personalized and efficient from a user’s perspective.
Commercial and trading systems gain trust and profits
The prediction formula is stated as below:
using such systems if they successfully predict what
users want at what time and where. The datasets were
created and averaged [5]. The corresponding

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


10
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar Int J S Res CSE & IT. January-February-2019 ; 5(1) : 09-15

Often, these systems are able to collect information


about a user’s.

3.2 Definition 2
2.3 Model-Based Collaborative Filtering
In this method, the system will use an unsupervised Collaborative Filtering: Collaborative Filtering
learning method to partition the space and then techniques make recommendations for a user based
classify the users using a similarity metric to a on ratings and preferences data of many users. The
segment or a cluster. To avoid having large clusters, main underlying idea is that if two users have both
systems use different methods to generate more small liked certain common items, then the items that one
practical clusters. This process starts with an initial set user has liked that the other user Search Engine
of clusters that contains only one user, and then Architecture.
repeatedly assigns users to these clusters based on a
similarity metric. Limiting features may become 3.3 Definition 3
necessary to reduce clustering complexity. The system
will create vectors for each segment and match the Content Based Recommendation: Content Based
user to the vector. Nonetheless, the user may be Recommendation algorithm takes into account the
classified as belonging to more than one cluster with a likes and dislikes of the user and generates a User
measure of similarity strength . Profile. For generating a user profile, we take into
account the item profiles (vector describing an item)
and their corresponding user rating. The user profile
is the weighted sum of the item profiles with weights
being the ratings user rated. Once the user profile is
generated, we calculate the similarity of the user
profile with all the items in the dataset, which is
III. PRELIMINARIES
calculated using cosine similarity between the user
profile and item profile. Advantages of Content Based
3.1 Defination1
approach is that data of other users is not required and
the recommender engine can recommend new items
Recommended system: A recommendation system is a
which are not rated currently, but the recommender
type of information filtering system which attempts to
algorithm doesn’t recommend the items outside the
predict the preferences of a user, and make suggests
category of items the user has rated.
based on these preferences. There are a wide variety
of applications for recommendation systems. These
IV. PROPOSED APPROACH
have become increasingly popular over the last few
years and are now utilized in most online platforms
4.1 Proposed System
that we use [7]. The content of such platforms varies
from movies, music, books and videos, to friends and
For recommendation, our proposed system uses
stories on social media platforms, to products on
collaborative filtering machine learning algorithm.
ecommerce websites, to people on professional and
Collaborative filtering (CF) is a machine learning
dating websites, to search results returned on Google.
algorithm which is widely used for recommendation
purpose. Collaborative filtering finds nearest neighbor

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


11
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar Int J S Res CSE & IT. January-February-2019 ; 5(1) : 09-15

based on the similarities. The metric of collaborative implementations, consult the Breese etc. All paper
filtering is the rating given by the user on a particular and/or our code.
item. Different users give different ratings to items.
Users, who give almost same rating to items, are the Item-based collaborative filtering is a model-based
nearest neighbors. In case of User based collaborative algorithm for making recommendations. In the
filtering, based on the ratings given by the users, algorithm, the similarities between different items in
nearest neighbors has been find. Item based the dataset are calculated by using one of a number of
collaborative filtering predicts the similarity among similarity measures, and then these similarity values
items. To recommend an item, items which are liked are used to predict ratings for user-item pairs not
by the user in his past have been found. Item which is present in the dataset. The system will search group
similar to those items has been recommended. for users by using k-means to find the distance
Internet contains a huge volume of data for between users, the group of users and clustering of
recommendation purpose. Due to size of data, if users. The system clustering with the k- means
recommendation computation has been done in single algorithms by measuring the distance of each data
system, then performance may degrade, and we point from the center of the 10 groups by using
cannot find an efficient solution. Euclidean distance and calculated information will be
stored in the database.
Hence we require distributed environment so that
computation can be increased and performance of 4.2 User Based CF
recommendation system gets improve. My goal is to
apply a collaborative filtering algorithm in a rating The dataset is firstly loaded into Hadoop distributed
website that collects users’ information, such as file system (HDFS). Then we perform User-based CF
location and gender, item’s information, such as using Mahout [9]. We take rating matrix, in which
category and description, as well as ratings for items each row represents user and column represents item,
by users. There are many algorithms that could be corresponding row-column value represents rating
applied on data to predict a user preference. User- which is given by a user to an item. Absence of rating
based, Item-based, and Model-based methods are value indicates that user has not rated the item yet.
ways of predicting a user preference [8]. The number There are many similarity measurement methods to
of users, items, or clusters in each one respectively compute nearest neighbors. We have used Pearson
will determine the function performance. However, correlation coefficient to find similarity between two
the most well-known and common one is User-based users [10]. Hadoop is used to calculate the similarity.
Collaborative Filtering. In this algorithm, we predict The output of the Hadoop Map phase i.e. userid and
an item’s rate for a user by collecting information corresponding itemid are passed to reduce phase. In
about this user and similar users. reduce phase, output has been generated and sorted
according to userid. Output again has been stored in
Memory-based algorithms approach the collaborative HDFS.
filtering problem by using the entire database. As
described by Breese et. al , it tries to find users that 4.3 Item Based CF
are similar to the active user (i.e. the users we want to
make predictions for), and uses their preferences to 3.2 Item Based CF Dataset is loaded into HDFS, then
predict ratings for the active user. This page will talk using Mahout we performs Item based CF. Past
about the general ideas; for specific equations and information of the user, i.e. the ratings they gave to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


12
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar Int J S Res CSE & IT. January-February-2019 ; 5(1) : 09-15

items are collected. With the help of this information


the similarities between items are builded and
inserted into item to item matrix. Algorithm selects
items which are most similar to the items rated by the
user in past. In next step, based on top-N
recommendation, target items are selected.

5.1.1 Global average:


V. EXPERIMENTAL ANALYSIS
The global average technique serves as a simple
baseline technique. The average rating for all users
5.1 Baseline methods
across all movies is computed. This global average
serves as a prediction for all the missing entries in the
We try out the following simple baseline methods to
ratings matrix.
give us an idea of the performance to expect from.
The reason behind these inappropriate results is, RSSI
5.1.2 User average:
provides advantages but further research is necessary
All users exhibit varying rating behaviors. Some users
for its sensitive reaction due to environment. RSSI
are lenient in their ratings, whereas some are very
ranging can be inaccurate and inconsistent especially
stringent giving lower ratings to almost all movies.
in indoor environment
This user bias needs to be incorporated into the rating
predictions. We compute the average rating for each
[8]. More over the received signal strength can vary
user. The average rating of the user is then used as the
considerably over small distances and small time
prediction for each missing rating entry for that
scales, due to multipath fading received signals or
particular user. This method can be expected to
path loss can exhibit wide variations even when d
perform slightly better than the global average since it
changes by as little as a few centimeters in the case of
takes into account the rating behavior of the users
802.15.4 radios[8].
into account.

5.1.3 Movie average:


Some movies are rated highly by almost all users
whereas some movies receive poor ratings from
everyone. Another simple baseline which can be
expected to perform slightly better than the global
average is the movie average method. In this
technique, each missing rating entry for a movie j is
assigned the average rating for the movie j.

5.1.4 Adjusted average:


This simple method tries to incorporate some
information about the user i and the movie j when
making a prediction for the entry ri j. We predict a
missing entry for user I and movie j, by assigning it

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


13
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar Int J S Res CSE & IT. January-February-2019 ; 5(1) : 09-15

the global average value adjusted for the user bias and Collaborative filtering is the most successful and
movie bias. popular algorithm in the recommender system’s field.
It helps customers to make a better decision by
The adjusted average rating is given by recommending interesting items. Even though this
ri j = lobalav + (uav(i)ij) + (mav(j)ij) algorithm is the best, it suffers from poor accuracy
and high running time. To solve these problems, this
The user bias is given by the difference between the paper proposed a recommendation approach based on
average user rating and the global average rating. The user clustering by using the Euclidian distance to
movie bias is given by the difference between the calculate two users to cluster dataset. This method
average movie rating and the global average rating. combines clustering and neighbors’ vote to generate
Consider the following example which demonstrates predictions. In the future there may be techniques,
the working of the adjusted average method. fuzzy c-means in the group stages of the first system
to provide a more effective segmentation.
5.2 Dateset:
For experiment, we have used MovieLens dataset of VII. REFERENCES
size 1M. The dataset contains 10000054 ratings and
95580 tags applied to 10681 movies by 71567 users. [1]. A Survey of Collaborative Filtering
There are three files, movies.dat, ratings.dat and Techniques;https://www.hindawi.com/journa
tags.dat. Ratings data file has at least three columns; ls/aai/2009/421425/.,2017, Vol. 13(7).
those are UserId, given by user to movie. [2]. Google News Personalization: Scalable
Online Collaborative Filtering; Das et al;
5.3 Result Analysis: https://www2007.org/papers/paper570.pdf
For movie recommendation, important factor is the [3]. Intro to Recommender Systems:
list of recommended items as soon as possible. Since Collaborative
we are using Hadoop, speedup and efficiency varies as Filtering;http://blog.ethanrosenthal.com/2015
number of nodes varies. To analyze this we have /11/02/intro-to-collaborative -filtering/"
obtained the number of movies which are [4]. Zhan J, Hsieh CL, Wang IC, Hsu TS, Liau CJ,
recommended as threshold changes, Speedup and Wang DW. Privacy-preserving collaborative
efficiency according to number of nodes. recommender systems. IEEE Transactions on
Systems, Man, and Cybernetics, Part C
VI. FUTURE WORK AND CONCLUSION (Applications and Reviews).
[5]. Gong S. A collaborative filtering
Most obvious ideas is to add features to suggest movies recommendation algorithm based on user
with common actors, directors or writers. In addition, clustering and item clustering. Journal of
movies released within the same time period could Software. 2010; 5(7):745-52.
also receive a boost in likelihood for recommendation. [6]. Manvi SS, Nalini N, Bhajantri LB.
Similarly, the movies total gross could be used to Recommender system in ubiquitous
identify a user’s taste in terms of whether he/she commerce. In international conference on
prefers large release blockbusters, or smaller indie electronics computer technology 2011 (pp.
films. However, the above ideas may lead to over 434-8). IEEE.
fitting, given that a user’s taste can be highly varied, [7]. Pu P, Chen L, Hu R. A user-centric
and we only have a guarantee that 20 movies evaluation framework for recommender

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


14
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar Int J S Res CSE & IT. January-February-2019 ; 5(1) : 09-15

systems. In proceedings of the fifth ACM


conference on recommender systems 2011
(pp. 157-64). ACM.
[8]. Hu R, Pu P. Acceptance issues of personality-
based recommender systems. In proceedings
of the third ACM conference on
recommender systems 2009 (pp. 221-
4).ACM."
[9]. Pathak B, Garfinkel R, Gopal RD, Venkatesan
R, Yin F. Empirical analysis of the impact of
recommender systems on sales. Journal of
Management Information Systems. 2010;
27(2):159-88. " ,
[10]. Witten IH, Frank E, Hall MA. Data mining:
practical machine leaning toots and
techniques. Morgan Kaufmann Publishers,
Elsevier; 2011."

Cite this article as :

Bheema Shireesha, Navuluri Madhavilatha,


Chunduru Anilkumar, "Movie Recommended System
by Using Collaborative Filtering", International
Journal of Scientific Research in Computer Science,
Engineering and Information Technology
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
09-15, January-February 2019.
Available at
doi : https://doi.org/10.32628/CSEIT19511
Journal URL : http://ijsrcseit.com/CSEIT19511

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


15
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19517

Comparison and an Improved Validation Optimistic Approach


for Concurrency Control
Monika Patel1, Prof. Dr. Dhiren B. Patel2
1MCA Department, S. K. Patel Institute Computer Studies, Gandhinagar, Gujarat, India
2 Department of Computer Science, Gujarat Vidhyapith, Ahmedabad, Gujarat, India

ABSTRACT

Concurrency Control entails the synchronization of accesses to the distributed database, such that the integrity
of the database is maintained. Devising an efficient concurrency control technique is challenging. There is a
need for improvised concurrency control technique to coordinate the simultaneous execution of transactions in
a multi-processing database system. Traditionally, concurrency control techniques are locking, timestamp
ordering and optimistic. These techniques have been evaluated by analytical modeling. In this undertaken work,
the analytical modeling has been investigated and evaluated. The concurrency control in order to improve
optimistic validation method with the traditional techniques, a new technique has been proposed.
Keywords : DBMS, DDBMS, Concurrency, Lock Maintenance, Restart Overhead, Abort Ratio, Deadlock
Possibility

I. INTRODUCTION computer network. The distributed database


management system is defined as the software system
Concurrency Control is a feature that is used to that permits the management of the DDBS to make
coordinate the simultaneous execution of transactions the distribution transparent to the users.
in a multi-processing database system while
preserving data integrity. It is the activity to manage In Concurrency Distributed Transactions, each server
the actions of transactions that operate in parallel manages a set of objects and is responsible for
access, shared data and potentially interfere with one ensuring that they remain consistent when accessed
another. The concurrency control problem is by concurrent transactions. Therefore, each server is
exacerbated in a distributed DBMS (DDBMS) because responsible for applying concurrency control to its
users may access data stored in many different own objects. The members of distributed
computers in a distributed system, and a concurrency transactions' servers are jointly responsible for
control mechanism at one computer cannot ensuring that they are performed in a serially
instantaneously know about interactions at other equivalent manner.
computers.

A distributed database is a collection of multiple,


logically interrelated databases distributed over a

CSEIT19517 | Received : 01 Jan 2019 | Accepted : 10 Jan 2019 | January-February -2019 [ 5 (1) : 16-21 ]
16
Monika Patel, Prof. Dr. Dhiren B. Patel et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 16-21

II. RELETED WORK transaction releases its locks after phase 1 of the
protocol.
Traditionally, concurrency control techniques have
been divided into three categories – locking, A lock manger in different servers set their locks
timestamp ordering and optimistic. Concurrency independently of one another. It is possible that
control techniques have been evaluated by analytical different servers may impose different ordering on
modeling. Kung and Robinson have the first to transactions. Consider the following interleaving of
develop the Optimistic Concurrency Control. transactions T1 and T2 at servers X and Y [8].

a. Problem Analysis and Analytic Modeling c. Timestamp ordering concurrency control


A distributed database means multiple servers or In a single site transaction, the lock manger issues a
multiple site from where data can be read or written. unique timestamp to each transaction when it starts.
When same data accessed and updated by many users, Serial equivalence is enforced by committing the
in the case where the multiple copy of same data can versions of the timestamps of transactions that
exist. In such situation there is a possibility that data accessed them. In distributed transactions, we require
read or written wrong which creates inconsistency in that each manager issue global and unique
the database. timestamps. A global a unique timestamp is issued to
the transaction by the first manager accessed by the
Problems occur when multiple users access the same transaction. The transaction timestamp is passed to
data item. It may happen - data item X can be read by the manager at each site whose data perform an
Transaction T1 and written by Transaction T2 at the operation in the transaction.
same time, creates conflicts in the database. When
same data accessed from many terminals there may When timestamp ordering is used for concurrency
be a chance of concurrency. Concurrency may result control, conflicts are resolved as each operation
in an inconsistent data in the database. To overcome performed. If the decision of a conflict requires the
this problem concurrency can be controlled by transaction to be aborted, the manger will be
concurrency control – means coordination between informed and it will abort the transaction from all
two or more transaction to access the data. the sites.
Commonly known methods of concurrency control
are 2-phase locking, optimistic, and ordering. d. Optimistic Method
The optimistic concurrency control method allows all
b. 2-phase Locking transaction to execute and validate before it will be
In a distributed transaction, the locks on an object are committed. Transaction numbers are assigned at the
held locally. The local lock manager can decide start of the transaction which is executed in the serial
whether to grant a lock or make the requesting manner. A distributed transaction is validated by the
transaction wait. However, it cannot release any several independent sites, each site validate for the
locks until it knows that the transaction has been data used from that site.
committed or aborted at all the servers involved in
the transaction. When locking is used for There are four phases in the transaction execution in
concurrency control, the objects remain locked and the optimistic concurrency control method:
are unavailable for other transactions during the 1. Read: Since reading a value of an object cannot
atomic commit protocol, although an aborted cause a loss of integrity, reads are completely

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


17
Monika Patel, Prof. Dr. Dhiren B. Patel et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 16-21

unrestricted. A transaction reads the values of a set of value updated only when it does not affect the
objects and assigns them to a set of local variables. consistency of the database. If updating of transaction
2. Compute: The transaction calculates a set of values not making database inconsistent, then transaction is
for data objects called the write set. These values are updated to the database and it is recorded in the
assigned to a set of matching local variables. Thus, all successfully completed transaction table, and if
writes after calculation take place on a transaction’s updating of transaction makes database inconsistent
copy of the objects of the database. then updated values removed from the temporary
3. Validate: The transaction’s local read set and write memory and it is recorded in the roll backed
set are validated against a set of committed transaction table. Finally, all the transactions are
transactions. recorded in the transaction log table.
4. Commit and Write: If the transaction succeeds in
validation, it is considered committed in the system In this work, an Improved Optimistic Validation
and is assigned a timestamp, otherwise the Method has been proposed, to improve the optimistic
transaction is rolled back or restarted at either the method. In this method, all the transactions are
compute phase or the read phase. If a transaction executed without restriction and when they need to
succeeds in the validation phase, its write set is made update the database, they have to pass the validation
global and the values of the write set become values test but in the improved method read transaction
of entities in the database at each node. does not go for validation check, so it decreases the
possibilities of restart or abort the transaction.
III. PROPOSED WORK
The major problem with optimistic method was
Model for Concurrent Execution in Distributed when one transaction go for validation, due to some
Databases reason it gets aborted because the resources used by
this transaction is also used by any other transaction
The below Fig 1. Shows the basic model for so there is the possibility of conflict. Once the
concurrent execution in distributed databases. transaction is aborted it gets the same timestamp and
Distributed databases mean there are multiple users starts execution, but it can also happen that the same
who are accessing data from various sites and it may resources used by the other transaction and so that
happen that two or more than two users accessing the transaction again gets aborted.
same data at the same time, at this situation problem
comes. When transaction gets aborted several times and does
not get finished or does not get the chance to execute,
The model shows the basic multiuser environment which means the transaction is in the condition of
with concurrent transactions and it’s after results. starvation. Starvation is the situation where
Assume n numbers of transactions are simultaneously transaction doesn’t get the fair chance to execute.
accessing the same database and executed the
numbers of operations in the one transaction. That
means several updates, read, delete operations are
going to be performed on the single database at time.
As soon as transaction starts its execution database
program identify the table and column in which data
is going to be read from or written into. Transaction

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


18
Monika Patel, Prof. Dr. Dhiren B. Patel et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 16-21

Improved Optimistic Validation Method

In Improved Optimistic Validation Method, it was


observed that, in most applications, the possibilities
of two clients’ transactions accessing the same data
item are low. Transactions are allowed to proceed as
though there were no possibilities of conflict with
other transactions until the client completes its task
and issues a terminatetransaction request. When a
conflict arises, some transaction is generally aborted
and will need to be restarted by the client.

This method was developed as an extension of the


optimistic method proposed by Kung and Robinson
[3]. It was assumed that, the conflicts are very rare in
the database system, and the number of read
transactions are very high than the write transaction.

The Improved Optimistic Validation Method adds


new concept to the previous optimistic method. In
Fig-1 : Model for Concurrent Execution in distributed
this method, there are three phases namely
database
Operational Phase, Validation Phase, Update Phase.
In this method, read transaction does not need to
Comparison of ORDER and Optimistic Method with
perform validation phase, which minimizes the
common criteria
execution time of the transaction. This improved
method has another feature of starvation, which
The below table shows the comparison between the
checks for the aborted transaction, and give it a right
two methods ORDER and Optimistic, in which we
chance to execute.
can analyse that the ORDER method uses locks and
Optimistic method does not, uses locks. But it checks
Validation Algorithm
the validation at time of committing the transaction.
Table shows the basic comparison using the basic
Tn: Transaction
assumptions.
StartT: The highest sequence number
at the start of Tn
Table -1 : Comparison of ORDER and Optimistic
ValidateT: The highest sequence
Method
number at the beginning of its
validation phase
declare valid=true
declare op=read/write
if(op==read)then valid=true
else

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


19
Monika Patel, Prof. Dr. Dhiren B. Patel et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 16-21

for t=StartT+1 to ValidateT do Table -2 : Comparison of ORDER, Optimistic and


if(writeset(t)&readset(Tn)=={})then Improved Optimistic Validation Method
valid=true
else
valid=false
if valid=true then go to update phase

Above algorithm will provide efficient results in


situation where the possibility for read transaction is
more.

IV. RESULTS
The above table shows the comparison between three
The Improved Optimistic Validation Method gives methods with some common criteria. Method 3 is
optimum performance when the system has improved method, developed during this research
minimum conflict ratio and the transactions are read work. If we look into the table, method 3 has been
in most cases. The proposed method does not disturb improved in two parameters – Validation for read
read transaction, increases the execution of transactions and starvation for execution.
transaction. The additional feature has been added of
the starvation. Due to the abort ratio, sometime one V. CONCLUSION
of the transactions is not able to perform and gets
aborted many times, in that case proposed method The Improved Optimistic Validation Method solves
helps the transaction to execute with different above discussed problems in the field of concurrency
conditions. control. The proposed method provides non-locking
algorithm for the transaction execution. It reduces
Comparison of ORDER, Optimistic and Improved the overhead of lock acquisition and lock releases.
Optimistic Validation Method The improved method uses no locks, meets no
deadlock leads to minimum abort ratio.
In this section of chapter, we will compare the
methods which are previously developed and new The assumption was made that, proposed method
one. The comparison criteria have been selected as suits to the environment having minimum conflict.
per the need of the database environment. We will And the most of the transaction has read operation to
compare three methods listed below in the Table 2. perform. The main aim to develop this method is to
improve the efficiency and transaction execution
Method 1: ORDER speed.
Method 2: Optimistic Method
Method 3: Improved Optimistic Validation Method The Improved Optimistic Validation Method
provides the direct access of the data without any
locking. In the method each transaction gets the
timestamp which is the real arrival time of the
transaction. There are three phases of the method:

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


20
Monika Patel, Prof. Dr. Dhiren B. Patel et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 16-21

Operational phase, Validation Phase, and Update [8]. G. Coulouris, J. Dollimore and T. Kindberg T.
phase. Distributed Systems Concepts and Design,
published by Pearson Education Limited 2005.
Limitations and Future Enhancements [9]. Morris R., Wong W., Performance analysis of
locking and optimistic concurrency control
The Improved Optimistic Validation Method is algorithms, May 1985.
efficient to work with the system having minimum [10]. M. Mohamed, M. Badawy, A. El-Sayed., Survey
conflicts, at the same time the proposed method may on Concurrency Control Techniques,
not sufficient in the case where possibility of conflict Communications on Applied Electronics (CAE)
is very high and number of write transactions is high. – ISSN : 2394-4714 Foundation of Computer
A method can be developed, which can control over Science FCS, New York, USA,2016
write transactions using locking concept. A new
method can be derived to enhance proposed method Cite this article as :
the way, read transactions should not be disturbed.
Monika Patel, Dr. Dhiren B Patel, "Comparison and
VI. REFERENCES an Improved Validation Optimistic Approach for
Concurrency Control", International Journal of
[1]. B. Bhargava, Concurrency Control in Database Scientific Research in Computer Science, Engineering
Systems, IEEE Transaction on Knowledge and and Information Technology (IJSRCSEIT), ISSN :
Data Engineering, JANUARY/FEBRUARY 2456-3307, Volume 5 Issue 1, pp. 16-21, January-
1999. February 2019.
[2]. A. Thomasian, Concurrency Control: Methods, Available at doi :
Performance, and Analysis, ACM Computing https://doi.org/10.32628/CSEIT19517
Surveys, March 1998. Journal URL : http://ijsrcseit.com/CSEIT19517
[3]. H. T. Kung, J. T. Robinson, On Optimistic
Methods for Concurrency Control, ACM
Transactions on Database Systems, Vol 6, No. 2,
June 1981.
[4]. P. L. Lehman, AND S. B. Yao, Efficient locking
for concurrent operations on B-trees. Submitted
for publication.
[5]. R. Shrinivasan, Network-Aided Concurrency
Control in Distributed Databases, January 2002.
[6]. S. Ceri, G. Pelagatti, Distributed Databases
Principals and Systems, McGraw-Hill
International Edition, 1985.
[7]. A. Thomasian, Distributed Optimistic
Concurrency Control Methods for High
Performance Transaction Processing, IEEE
Transaction on Knowledge and Data
Engineering, VOL. 10, NO. 1, January/February
1998.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


21
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19514

Airline Data Analysis


Navuluri Madhavilatha1, Bheema Shireesha2, Chunduru Anilkumar3
1Guest Faculty, Department of Computer Science and Engineering, Dr. APJ Abdul Kalam IIIT Ongole, Andhra
Pradesh, India
2, 3 Assistant Professor, Department of Computer Science and Engineering, Dr. APJ Abdul Kalam IIIT Ongole,
Andhra Pradesh, India
ABSTRACT

In the contemporary world, Data analysis is a challenge in the era of varied inters disciplines through there is a
specialization in the respective disciplines. In other words, effective data analytics helps in analyzing the data of
any business system. Flight delays hurt airlines, airports, and passengers. Their prediction is crucial during the
decision-making process for all players of commercial aviation. The goal of our project is to get monthly wise
statistics of airline data and taking particular airport as target we are further analyzing the data to get the
hourly statistics. And also we are finding out the most popular source-destination pairs and calculating the
average delays at every airport. The data for this project comes from the stat-computing.org website. In
particular, in the year 2008 data 70,09,728 titles recorded there which includes information on the Origin,
Destination, Month, Year, DayofWeek, DayofMonth, DepDelay, ArvDelay, DepTime, ArvTime and a few other
less interesting variables. Conveniently, you can export the data directly as a csv file.
Keywords : Machine Learning, Data Mining, Big Data, Statistics, Data Visualization, Data Analytics, ASCII, SRS,
GNU

I. INTRODUCTION Data:
Data is a set of values of qualitative or quantitative
The goal of our project is to get monthly wise variables. It can representation of facts as text,
statistics of airline data and taking particular airport numbers, graphics, images, sound or video. Raw data,
as target we are further analyzing the data to get the also known as primary data, is collected from a source.
hourly statistics. And also we are finding out the If a scientist sets up a computerized thermometer
most popular source-destination pairs and calculating which records the temperature of a chemical mixture
the average delays at every airport [1]. The data for in a test tube every minute, the list of temperature
this project comes from the statcomputing.org readings for every minute, as printed out on a
website in particular, in the year 2008 data 70,09,728 spreadsheet or viewed on a computer screen is "raw
titles recorded there which includes information on data" [2]. Raw data can be input to a computer
the Origin, Destination, Month, Year, DayofWeek, program or used in manual procedures such as
DayofMonth, DepDelay, ArvDelay, DepTime, analyzing statistics from a survey. The term "raw
ArvTime and a few other less interesting variables. data" can refer to the binary data on electronic
Conveniently, you can export the data directly as a storage devices, such as hard disk drives (also referred
csv file. to as "low-level data"). All software is divided into
two general categories: data and programs. Programs

CSEIT19514 | Received : 01 Jan 2019 | Accepted : 10 Jan 2019 | January-February -2019 [ 5 (1) : 22-29 ]
22
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

are collections of instructions for manipulating data. between variables. Once you find these insights, you
Strictly speaking, data is the plural of datum, a single validate the findings by applying the detected
piece of information. In practice, however, people patterns to new subsets of data. The ultimate goal of
use data as both the singular and plural form of the data mining is prediction [5]. Data Science is an
word [3]. The term often used to distinguish binary umbrella that contain many other fields like Machine
machine-readable information from textual human- learning, Data Mining, big Data, statistics, Data
readable information. For example, some applications visualization, data analytics etc.
make a distinction
II. METHODS AND MATERIAL

System Requirement Specification

2.1 Software requirement specifications


It specifies the hardware and software requirements
that are required in order to run the application
properly. The Software Requirement Specification
(SRS) is explained in detail, which includes overview
Between data files (files that contain binary data) and
of this dissertation as well as the functional and non-
text files (files that contain ASCII (American
functional requirement of this dissertation.
Standard Code for Information Interchange) data.
Functional Requirements: R Studio
In database management systems, data files are the
Non-Functional Requirements:Dataset (.csv file)
files that store the database information, whereas
other files, such as index files and data dictionaries,
2.2 Hardware requirements:
store administrative information, known as metadata.
Processor : Any Processor above 500 MHz
The seven V’s sum it up pretty well – Volume,
Ram : 2 GB
Velocity, Variety, Variability, Veracity, Visualization,
Hard Disk : 10 GB
and Value [4]. Science is the intellectual and practical
Input device : Standard Keyboard and Mouse
activity encompassing the systematic study of the
Output device : VGA and High Resolution Monitor.
structure and behavior of the physical and natural
world through observation and experiment Data
III. Technology Description
science is an interdisciplinary field about processes
and systems to extract knowledge or insights from
3.1 Introduction to R
data in various forms, either structured or
unstructured, which is a continuation of some of the
R is a language and environment for statistical
data analysis fields such as statistics, machine
computing and graphics. It is a GNU project which
learning, data mining, and predictive analytics.
is similar to the S language and environment which
was developed at Bell Laboratories (formerly AT&T,
Difference between data mining and data science
now Lucent Technologies) by John Chambers and
colleagues. R can be considered as a different
Data mining refers to the science of collecting all the
implementation of S. There are some important
past data and then searching for patterns in this data.
differences, but much code written for S runs
You look for consistent patterns and / or relationships
unaltered under R.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


23
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

R provides a wide variety of statistical (linear and R is designed around a true computer language, and it
nonlinear modelling, classical statistical tests, time- allows users to add additional functionality by
series analysis, classification, clustering etc) and defining new functions. For computationally-
graphical techniques, and is highly extensible [6]. intensive tasks, C, C++ and FORTRAN code can be
The S language is often the vehicle of choice for linked and called at run time. Advanced users can
research in statistical methodology, and R provides an write C code to manipulate R objects directly. Many
Open Source route to participation in that activity. users think of R as a statistics system. We prefer to
think of it of an environment within which
One of R’s strengths is the ease with which well- statistical techniques are implemented. R can be
designed publication-quality plots can be produced, extended (easily) via packages. There are about eight
including mathematical symbols and formulae packages supplied with the R distribution and many
where needed. Great care has been taken over the more are available through the CRAN family of
defaults for the minor design choices in graphics, but Internet sites covering a very wide range of modern
the user retains full control. statistics. R has its own Latex-like documentation
format, which is used to supply comprehensive
R is available as Free Software under the terms of documentation, both on-line in a number of formats
the Free Software Foundation’s GNU General Public and in hardcopy.
License in source code form. It compiles and runs on
a wide variety of UNIX platforms and similar systems 3.3 R Studio
(including FreeBSD and Linux), Windows and MacOS.
R Studio is an integrated development environment
3.2 The R Environment (IDE) for the R programming language. Some of its
features include: Customizable workbench with all of
R is an integrated suite of software facilities for data the tools required to work with R in one place
manipulation, calculation and graphical display. It (Console, source, plots, workspace, help, history, etc.).
includes an effective data handling and storage Syntax highlighting editor with code completion.
facility, a suite of operators for calculations on arrays,
in particular matrices, a large, coherent, integrated Execute code directly from the source editor (line,
collection of intermediate tools for data analysis, selection, or file). Full support for authoring Sweave
graphical facilities for data analysis and display either and TeX documents.
on-screen or on hardcopy, and a well-developed,
simple and effective programming language which Runs on all major platforms (Windows, Mac, and
includes conditionals, loops, user-defined recursive Linux) and can also be run as a server, enabling
functions and input and output facilities[7]. The term multiple users to access the R Studio IDE using a web
“environment” is intended to characterize it as a fully browser.
planned and coherent system, rather than an
incremental accretion of very specific and inflexible
tools, as is frequently the case with other data
analysis software.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


24
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

IV. System Design graph we can conclude that most number of flights
(627931) are running in July month. And second
most number of flights (616090) are running in
March month.

5.2 Flight statistics of Indy airport


Here we are going to take Indy airport as an Origin
as well as destination calculating monthly wise,
weekly wise, day wise and hourly wise flight
statistics.

5.2.1 Monthly wise statistics of Indy airport

Indy as Origin:-

Fig. 4.1 Analysis of airline data

Data set (.csv file) contains 29 columns as “origin,


destination, month, day of week departure delay,
departure time, arrival time, arrival delay etc”. By
taking data set as an input we generate the graphs
which can be easily understood by the client.

V. Experimental Results

5.1 Monthly wise statistics of entire Airline


data

From the above graph we can conclude that more


number of flights are leaving Indy airport in the
month of July and the less number of flights in the
month of November.

Indy as Destination:
From the below graph we can conclude that more
number of flights are arriving Indy airport in the
month of July and the less number of flights in the
month of November.

The above graph is drawn by taking Months on X-


axis and Number of Flights on Y-axis. From the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


25
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

5.2.3 Day wise statistics of Indy airport

5.2.2 Weekly wise statistics of Indy airport Indy as Origin:-


Indy as Origin:-
From the below graph we can conclude that more
From the below graph we can conclude that more number of flights are leaving the Indy airport in the
number of flights are leaving Indy airport on duration of 6am – 7am and the less number of flights
Wednesday and the less number of flights on in the duration of 12am – 4 am in a day.
Saturday in a week.

Indy as Destination:- Indy as Destination:-


From the below graph we can conclude that more From the below graph we can conclude that more
number of flights are arriving Indy airport on number of flights are arriving the Indy airport in the
Wednesday and the less number of flights on duration of 10am – 11am and the less number of
Saturday in a week. flights in the duration of 2am – 6 am in a day.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


26
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

5.2.4 Hourly wise statistics of Indy airport 5.3 Source to Destination airport pairs
Indy as Origin:- Here we are going to take a look at the most and
From the below graph we can conclude that more least popular airport-pairs based on the flights
number of flights are leaving the Indy airport in the operating in between them (Means one as Origin
duration of 6am – 7am and the less number of flights and another as Destination).
in the duration of 12am –
4 am in a day. 5.3.1 Most Popular

Indy as Destination:- These are the top 6 airport pairs with max number
From the below graph we can conclude that more of flights running between them. In these SFO to
number of flights are arriving the Indy airport in the LAX stands first with 13788 running.
duration of 10am – 11am and the less number of
flights in the duration of 2am. 5.3.2 Least Popular
These are the least 6 airport pairs with minimum
number of flights running between them.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


27
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

5.5.1 Departure delays for all airports

5.4 Average delays of most popular airports


Here we are comparing the average flight delays
among the top 10 most popular airports.

In the above graphs we are representing the


largest average and smallest average departure
Based on the graph we can conclude that OGG delays of the first six (head) airports.
airport running flights with least amount of arrival
and departure delays. And ORD airport running 5.5.2 Arrival delays for all airports
flights with high Amount of departure delay. LGA In the below graphs we are representing the
airport running flights with high Amount of arrival largest average and smallest average arrival delays
delay. of the first six (head) airports.

5.5 Average delays for all airports


Here, we are calculating the average delays (both
arrival and departure) of all the airports.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


28
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

airports we come to know the busy hours and days


in which the passenger flow will be high. Those
days we can arrange the alternatives in order to
reduce the problems faced by the passengers as
well as by the authorities like increasing the
security, food and water facilities, transport
facilities outside the airport ,runway clearance etc.

VIII. REFERENCES

[1]. Abdelghany, K. F., Abdelghany, A. F., and Raina S.,


(2004) A model for projecting flight delays during
irregular operation conditions, Journal of Air
Transport Management, Volume 10, Issue 6, Pages
385-394.
[2]. American Statistical Association(stat-computing.org)
[3]. Thearling, K., Data Mining and Analytic
Technologies, www.thearling .com, 2004.
[4]. Tutorials point (http://www.tutorilaspoint.com/r/)
[5]. Aisling, R., and J.B. Kenneth, (1999) An assessment of
the capacity and congestion levels at European
airports, ERSA conference papers ersa 99, pages 241,
European Regional Science Association
[6]. Bureau of Transportation Statistics, Airline On-Time
Statistic. U.S. Department of Transportation.
VI. CONCLUSION Washington, D.C.
http://www.bts.gov/programs/airline information
Not all arrivals can occur when they are scheduled [7]. Hansen, M., and C. Y. Hsiao (2005), Going South? An
to, then airport congestion happens. The delay Econometric Analysis of US Airline Flight Delays
distribution of the airport can make it easier to from 2000 to 2004, Presented at the 84rd Annual
understand the airport delay. We also found the Meeting of the Transportation Research Board (TRB),
Washington D.C., 2005.
most popular origin to destination pairs according
to the number of flights running. We compared
the average arrival and departure delays among Cite this article as :
Navuluri Madhavilatha, Bheema Shireesha, Chunduru
them as well. And also we have calculated the
Anilkumar, "Airline Data Analysis", International
average delays to all the airports.
Journal of Scientific Research in Computer Science,
Engineering and Information Technology (IJSRCSEIT),
VII. FUTURE SCOPE ISSN : 2456-3307, Volume 5 Issue 1, pp. 22-29, January-
February 2019.
Further we can implement this project by Available at doi :
calculating the monthly wise, weekly wise, day https://doi.org/10.32628/CSEIT19514
wise and hourly wise flight statistics at popular Journal URL : http://ijsrcseit.com/CSEIT19514

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


29
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1838108
Li-Fi the Next Generation of Wireless Communication through Visible
Light Communication (VLC) Technology
Imran Siddique, Muhammad Zubair Awan, Muhammad Yousaf Khan, Azhar Mazhar
Department of Computer Science and Information Technology, Afro Asian Institute, Lahore Affiliated with
Government College University Faisalabad, Pakistan

ABSTRACT

Li-Fi stands for Light-Fidelity. This technology is very new and was proposed by the German physicist Harald
Haas in 2011. Light based communication system is the backbone of the future of the communication system.
Li-Fi is a wireless technology that uses light emitting diodes (LEDs) for transmission of data. The development
of the wireless communication leads to advance research in LiFi technology. The term Li-Fi states to visible
light communication (VLC) technology that uses as medium to deliver high-speed communication in a fashion
similar to Wi-Fi. Li-Fi comprises a wide range of frequencies and wavelengths, from the Infrared through
visible and down to the Ultraviolet spectrum. The immense use of Li-Fi may solve some bottleneck of data
transmission in Wi-Fi technology. With the innovation in technology and the number of users, the existing
radio-wave spectrum fails to accommodate this need. To resolve the issues of scalability, availability and
security, we have come up with the concept of transmitting data wirelessly through light using visible light
communication (VLC) technology. This paper objective is to study and describe the LiFi technology. The
improvement of the wireless communication leads to advance research in LiFi technology through Visible
Light Communications (VLC) Technology.
Keywords : Visible Light Communications (VLC), Light Emitting Diodes (LEDs), Radio Frequency (RF),
Infrared (IR), Remotely Operated Vehicles (ROVs), Line of Sight (LOS).

I. INTRODUCTION carrier signal instead of traditional Radio Frequency as


in Wi-Fi. Li-Fi technology communicates with the
With incessantly growing needs of telecommunication help of Visible Light Communication (VLC) spectrum
world, there is an increased thrust for higher and has no side effect as we know the light is very
bandwidth that facilitates faster and secure data much part of our life and so much faster. Moreover Li-
transmission. Present telecommunication industry Fi makes possible to have a wireless Internet in
relies on radio waves of electromagnetic spectrum for specific environments, where Wi-Fi is not allowed
data transmission. Inappropriately, the radio wave due to interferences or security considerations.
spectrum has certain key limitations: Capacity,
Efficiency, Availability and Security. So using the II. BACKGROUND
Light Emitting Diodes that offers in the transmission
of data much faster and reliable than the data that can The history of wireless communication based on
be transmitted through Wi-Fi. Li-Fi is a wireless electromagnetic waves dates back to the Photophone,
communication system in which light is used as a invented by Alexander Graham Bell in the late 19th

CSEIT1838108 | Received : 01 Jan 2019 | Accepted : 10 Jan 2019 | January-February -2019 [ 5 (1) : 30-37 ]
30
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

century. It was the first device used to communicate


without wires and Bell’s invention used the light
produced by the sun to carry the information [6].
There are around 1.4 million cellular staff radio
waves base stations installed, with over 5 billion
mobile phones.

transmission speed. Li-Fi is the new technology has


Mobile phones transmit over 600TB of data. With the
used to label the fast and cheap wireless
advancement in technology and the number of users,
communication system, which is the next optical
the existing radio wave spectrum fails to satisfy to
version of Wi-Fi [4].With the innovation in
this need [1].A new generation of wireless
technology and the number of users, the existing
communications is considered as the fourth stage of
radio-wave spectrum fails to accommodate this need.
the revolution mobile trade. This enables
To resolve the issues of scalability, availability and
connectivity between different objects and machines
security, we have come up with the concept of
all over the world [2].Currently internet have
transmitting data wirelessly through light using
become a major demand; people are in search of Wi-
visible light communication (VLC) technology.
Fi hot spots. This is the advanced scenario in today
communication system which enables the need of
B. VISIBLE LIGHT COMMUNICATION (VLC)
LEDs, Light Emitting
SYSTEMS

Diodes that provides in the transmission of data


Visible Light Communication (VLC) has enlarged
much faster and reliable than the data that can be
great interest in the last decade due to the quick
transmitted through Wi Fi [3].
developments in Light Emitting Diodes (LEDs)
fabrication. Efficiency, durability and long life span
III. METHODS AND MATERIAL
of LEDs make them auspicious residential lighting
equipment as well as alternative cheap and fast data
A. ISSUES IN WIRELESS COMMUNICATION
transfer equipment [5]. One of the ideas laid forward
for wireless optical communication is the visible light
Transfer of data from one place to another is one of
communication method. The signals in the 380-780
the most important daily activities. When the
nm wavelength interval of the electromagnetic
multiple devices are connected to the existing
spectrum are the light signals that can be detected by
wireless networks that connect us to the internet are
the human eye. It is possible to attain illumination
very slow. As the number of devices increasing the
and data transfer simultaneously by means of LEDs
internet access, the fixed bandwidth which is
that is the prominent lighting equipment lately. By
available makes it more and more difficult to utilize
this way, both internal lighting of a room and data
high data transfer rates and connect to a network.
transfer will be attained without the need of an
But, radio waves are just a small part of the spectrum
additional communication system. This technology is
available for data transfer. A solution to this problem
given the name of Visual Light Communication [6].
is by the use of the proposed system LiFi. Li-Fi is
transmission of data through light by sending data
Fundamental entries in a VLC system are the
through an LED bulb that varies in intensity faster
transmitter (LEDs), receivers (photo detectors),
than the human eye can follow and the faster data

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


31
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

modulation of data to optics and the optical technological communication at ultra-high speed
communication channel as shown in Fig -3: greater that of current wireless systems One of the
goals of researchers is to allow 100 megabits of data
transference per second (Mbps) in offices and homes
by modulation of light from promoted lighting
systems. If it is developed correctly, the possibility
exists that many of the problems associated with
present day infrared, radio wave and microwave
communications systems could be at least partially
Fig -1: Visible Light Communication (VLC) Systems resolved, and a more biologically friendly system
[6]
made existing to industries and the general public.

An auxiliary advantage is that VLC systems can


transmit data more securely over short distances than
radiofrequency/microwave communications devices
whose signals can be easily detected outside the
rooms and buildings they invent in [7].

Fig -2: Basic VLC configuration [5]


Fig -4 : Layered architecture of VLC

D. WHY USE VISIBLE LIGHT COMMUNICATION


(VLC)

Fig -3: Block diagram of a VLC architecture [5] The Gamma rays cannot be used as they could
be dangerous [10].
C. LED (LIGHT EMITTING DIODE) VLC X-rays has related health issues.
TECHNOLOGY Ultraviolet light is good for place without
people, but otherwise dangerous for the human
LED (Light Emitting Diode) Visible Light body.
Communications (VLC) systems are familiar as Infrared, due to eye safety regulation, can only
creating a possible valuable addition to future be with low power [10].
generations of technology, which have the potential Radio waves penetrate through the walls so they
to utilize light for the purposes of advanced arises security issues.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


32
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

Hence we left with the only the visible- “Light ✓ A good silicon photodiode, as a receiving
Spectrum”. element.

By using LiFi, it is possible to get the high speed of


100Mbps and more with the help of various
multiplexing techniques. The LiFi emitter system
consists of 4 main subassemblies:
Fig -5: The Electromagnetic Spectrum o Bulb
o RF power amplifier circuit (PA)
E. WHAT IS LI-FI? o Printed Circuit Board (PCB)

Li-Fi is a Visible Light Communications (VLC) system. H. WORKING STYLE OF LIFI


Like Wi-Fi it’s a bidirectional wireless communication
technology. But, where Wi-Fi uses radio waves to It is applied by using a light bulb at the downlink
transmit data, Li-Fi uses visible light from LED light transmitter. Normally the light bulb glows at a
bulbs fitted with a special chip. It’s a little like Wi-Fi, constant current supply however fast and elusive
but has reached speeds 100 times faster in testing, variations in current can be made to produce the
making it far more suited to the demands of the future optical outputs since it just uses the light, hence can
of data [8]. Li-Fi and Wi-Fi are quite similar as both be easily applied to any such area where radio
transmit data electromagnetically. frequency communication is often challenging [11].
The source of light continuously appears because the
However, Wi-Fi uses radio waves while Li-Fi runs on LED can be turned ON and OFF quickly and the
visible light. As we now know, Li-Fi is a Visible Light amount is rapidly modulated that human eye cannot
Communications (VLC) system. This means that it fastening even though it is flickering. The LEDs
accommodates a photo-detector to receive light activity of ON-OFF which invisible and enables the
signals and a signal processing element to convert the transmission of data using the binary codes i.e. 0s and
data into 'stream able' content [9]. 1s, for switch ON LED it is a logical ‘1’, and for
switch OFF LED it is a logical ‘0’. This method of
using rapid pulses of light to transmit information
F. LI-FI IS A WIRELESS TECHNOLOGY? wirelessly is technically referred to as visible light
communication (VLC) [12].
Li-Fi is a wireless technology as it uses visible light
communication or infrared or even near ultraviolet
spectrum. It does not use the radio frequency like the
Wi-Fi technology we use today [10].

G. LIFI SYSTEM CONSTRUCTION [11]

LiFi uses pulses of light to transmit data over wireless


networks. The main mechanisms of LiFi systems are:
Fig -6: Innovative Architecture for Future Li-Fi
✓ High brightness white LED, which acts as the Wireless Technology
transmission source.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


33
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

❖ TRANSMITTER Li-Fi makes use of a free band that does not


prerequisite any licensing.[15]
LED is used as the transmitter or sender of signal Li-Fi Can be used securely in aircrafts
here. The wavering in the LED is basically the signal without affecting airline signals unlike Wi-Fi.
that is to be sent. The quick ON (transmits 1) and The issues of the scarcity of radio frequency
OFF (transmits 0) of the LED is used to encode a bandwidth can be sorted out by Li-Fi.[15]
string of data signal. Now this technique can be Li-Fi also known as Green information
upgraded by using multiple LEDs or LEDs of different technology as it does not effect on the birds,
colours like red, green, blue for more complex data human body’s etc. corresponding to the radio
communications. Tentatively, the speed of data waves or any other communication waves.
communication through Li-Fi can be up to 10Gbps Li-Fi does not have any side effect on any
[13]. living thing. [17]

B. DISADVANTAGES OF LI-FI

High Power Required.


Low Range.
Fig -7: Transceiver LiFi based on VLC Reflection of light takes place. [14]
Cannot penetrate into walls.
Reliability and Network coverage are also
❖ RECEIVER major issues.
Any interference from the exterior light
Since Li-Fi uses visible light to transmit data, it is sources like normal bulbs, opaque materials
important to convert them into digital signals with in the route of transmission will cause
the help of modulation. It includes silicon photo disturbance in communication.[16]
diode which shows good response to visible
wavelength, optical concentrator and filter and an C. APPLICATIONS AREAS OF LI-FI
amplifier. The optical concentrator is used to
compensate for high spatial attenuation due to the A. Educational Systems
beam divergence from the LEDs to illuminate large
area [13]. Li-Fi is the latest technology that can provide fastest
speed for Internet access. So, it can be a good
IV. RESULTS AND DISCUSSION alternative of Wi-Fi at educational institutions and at
companies so that the people there can make use of
A. ADVANTAGES OF LI-FI Li-Fi with the high speed.
High speed.
High Data Transfer Rate. B. Traffic Management
High Connectivity.
Highly Secured and Reliable.[14] In traffic signals Li-Fi can be used to communicate
Low Cost. with passing vehicles (through the LED lights of the
Use Light as career. Vehicles) which can help in managing the traffic in a
Very high operating frequency. better manner resulting into smooth flow of traffic

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


34
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

and controlling the traffic struck and increasing Difference between LIFI and WIFI
accident ratios in developing countries. The LED car
lights can alert drivers when other vehicles are too Parameters Li-Fi Wi-Fi
close via LiFi technology. Transmitter LED Antenna
Frequency band 1000 times 2.4 GHz
C. Better Replacement for Other Technologies
of THz
Standard IEEE 802.15.xx IEEE 802.11xx
Li-Fi uses Visible Light Communication Technology No of users Under The Depend on
(VLC) Instead of Radio Waves. So, it can be easily All over Lamp.(LEDs) access Point.
Data Bits Radio waves
installed at those places where Bluetooth, Infrared,
Transmission
Wi-Fi can’t work properly.
Topology Point to Point Point to
Multipoint
D. Underwater Applications Communicat Based on Visible Based on Radio
ion Light Frequency
Li-Fi can even work underwater where Wi-Fi fails Communication Communication
completely, if the cables of the Remotely Operated (VLC).
Vehicles (ROVs) which operate at a long distance Availability Anywhere, Limited
from the surface are replaced with light (LEDs), available in
which is water-logged then it is easier to access the Airplanes and
internet [16]. Underwater
Power Less More
Consumptio
E. Exclusive Medical Applications
n
Environment Low Medium
Wi-Fi cannot be installed at Hospitals/Operation Impact
Theatres/Medical Laborites as these radio waves can Cost Low High
interfere with the Cellphones/Tablets PC which can Bandwidth Unlimited Limited
block the signals for monitoring equipments thus
proving hazardous to the patient’s health. Usage of G. Industrial Automation
Li-Fi provides more advanced monitoring techniques
and operation strategies. Anywhere in industrial areas data has to be
transmitted, Li-Fi is capable of interchanging slip
F. Cheaper Internet in Aircrafts rings, sliding contacts and short cables, such as
Industrial Ethernet. Due to the real time capability of
The passengers travelling in aircrafts have to Li-Fi, it is also an alternative to common industrial
prescribes to put their electronic devices in the Flight Wireless standards.
Mode or No Internet Mode, because in an aircraft as
these signals may interfere with the navigation H. Safety Environments
signals. So, Wi-Fi is practically impossible in this
scenario.Li-Fi can easily provide high speed Internet In explosion risk environments, the use of electrical
via every light source such as overhead reading LEDs equipment, including mobile phones, is generally
present inside the airplane. greatly restricted. The use of Li-Fi to pass data will
simplify the configuration of data networks in such

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


35
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

environments, and can permit new systems to [3] R.Mallika@Pandeeswari; “Survey of (LI-FI) Light
enhance security in these environments [19]. Fidelity” ,2nd Year-IT, Department of Information
Technology, Mepco Schlenk Engineering College
V. CONCLUSION [International Conference on Emerging Trends in
Applications of Computing (ICETAC 2K17)
Li-Fi is the trend of today and near future. It is one of http://www.ijetjournal.org ISSN: 2395-1303]
[4] Mr. Korde S. K; “Review Paper on Li-Fi (Light
the cheapest and efficient modes of data transfer. It
Fidelity)”
surely can replace the traditional Wi-Fi networks and
[Vol-2 Issue-3 2016 IJARIIE-ISSN (O)-2395-4396]
can further be extended to different platforms to
[5] Taner Cevik and Serdar Yilmaz; “An Overview Of
make it easily accessible and portable which will thus
Visible Light Communication Systems”
cater to the growing demands of the increasing
[International Journal of Computer Networks &
population. One of limitation however is that it only
Communications (IJCNC) Vol.7, No.6, November
works in direct line of sight (LOS) but this would
2015]
solve the problems like the shortage of radio- [6] Adam Böcker, Viktor Eklind Daniel Hansson, Philip
frequency bandwidth and allow internet (WI-FI) to Holgersson, Jakob Nolkrantz & Albin Severinson;
be used in places and situations where it cannot be “An implementation of a visible light
used currently like the hospitals or aircrafts. There communication system based on LEDs”
are billions of light bulbs worldwide, they just need [Department of Signals and Systems CHALMERS
to be replaced with LEDs ones that transmit data. UNIVERSITY OF TECHNOLOGY Gothenburg,
Visible Light Communication (VLC) is very cheaper Sweden 2015]
to Wi-Fi because it uses light rather than radio- [7] Roma Jain, Pallavi Kale, Vidya kandekar, Pratiksha
frequency signals. Kadam; “Wireless Data Communication Using Li-Fi
Technology” [DOI 10.4010/2016.846 ISSN 2321
VI. ACKNOWLEDGEMENTS 3361 © 2016 IJESC]
[8] Jon Mundy and Sacha Kavanagh; “What is Li-Fi?”
The authors would like to thank “Afro Asian Institute, [https://5g.co.uk]
Lahore” Affiliated with Government College [9] Mr. Prashant T Rayrikar; “Wi-Fi Technology Vs Li-
University Faisalabad (GCUF) for the support Fi Technology” [IRJET e-ISSN: 2395 -0056 Volume:
03 Issue: 05 | May-2016 p-ISSN: 2395-0072]
provided to Carry the research work presented in this
[10] Ekta, Ranjeet Kaur; “Light Fidelity (LI-FI)-A
paper.
Comprehensive Study” [International Journal of
VII.REFERENCES
Computer Science and Mobile Computing, Vol.3
Issue.4, April- 2014]
[11] Abhishek Gupta Swayanshu Shanti Pragnya; “The
[1] Roma Jain, Pallavi Kale, Vidya kandekar, Pratiksha
Future Technology LiFi” [IJSRD - International
Kadam; “Wireless Data Communication Using Li-Fi
Journal for Scientific Research & Development| Vol.
Technology” [ISSN 2321 3361 © 2016 IJESC]
4, Issue 08, 2016, ISSN (online): 2321-0613]
[2] Imran Siddique & Muhammad Zubair Awan; “An
[12] Dinesh Khandal, Sakshi Jain; “Li-Fi (Light Fidelity):
innovation in Mobile Communication Technology”
The Future Technology in Wireless
[International Journal of Research e-ISSN: 2348-
Communication” [International Journal of
6848, p-ISSN: 2348-795X Volume 05 Issue 23,
Information & Computation Technology. ISSN
December 2018]
0974-2239 Volume 4, Number 16 (2014), pp. ©

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


36
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

International Research Publications House (Electronics & Communication) Engineering from


http://www. irphouse.com] UCEST, Lahore, Pakistan.
[13] Anwesha Chakraborty, Trina Dutta, Sushmita
Mondal, Dr. Asoke Nath; “Latest advancement in Muhammad Zubair Awan
Light Fidelity (Li-Fi) Technology” [International Muhammad Zubair Awan received
Journal of Advance Research in Computer Science his Master Degree (MCS) in
and Management Studies ISSN: 2321-7782 (Online) Computer Science from University
e-ISJN: A4372-3114]
of Lahore, Pakistan and pursuing his
[14] Harsh Deep Ahlawat, Gautham Menon, Girish
MS in Computer Science from
Chhabra; “Advanced Light Fidelity (Li-Fi) System”
COMSATS University, Lahore, Pakistan
[International Research Journal of Engineering and
Technology (IRJET) e-ISSN: 2395 -0056 Volume: 03
Muhammad Yousaf Khan
Issue: 04 | Apr-2016 p-ISSN: 2395-0072]
Muhammad Yousaf Khan
[15] Sowbhagya M P, P Vikas Krishna, Darshan S, Nikhil
A R; “Evolution of Gi-Fi and Li-Fi in Wireless
received his B.Sc. degree in
Networks” [International Journal of Computer Computer System Engineering
Sciences and Engineering ,Volume-4, Special Issue- from Bahauddin Zakariya
3, May 2016 ISSN: 2347-2693] University Multan, Pakistan and
[16] Nikhita Adidam; “Li-Fi - The Future of Internet” completed his M.Sc. in Computer System
[International Journal of Computer Science & Engineering from Islamia University Bahawalpur,
Engineering Technology (IJCSET) ISSN : 2229-3345 Pakistan.
Vol. 6 No. 01 Jan 2015]
[17] Ashmita Shetty; “A Comparative Study and Azhar Mazhar
Analysis on Li-Fi and Wi-Fi” [International Journal Azhar Mazhar pursuing his B.Sc.
of Computer Applications (0975 – 8887) Volume Degree in Electrical Engineering
150 – No.6, September 2016] (2nd year) from Khwaja Fareed
[18] E Ramadhani , G P Mahardika; “ The Technology of University of Engineering &
LiFi: A Brief Introduction” Department of Information Technology (KFUEIT),
Informatics, Universitas Islam Indonesia, Rahim Yar Khan, Pakistan
Yogyakarta, Indonesia [IOP Conference Series:
Materials Science and Engineering]
Cite this article as :
[19] www.lifi-centre.com

About Authors:

Imran Siddique
Imran Siddique received his B.Sc.
degree in Computer Engineering
from Bahauddin Zakariya
University Multan, Pakistan and
pursuing his MS in Electrical

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


37
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19515

Demand Supply Oriented Taxi Suggestion System for


Vehicular Social Networks with Fuel Charging Mechanism
Selvi C1, Keerthana D2
1
Computer Science and Engineering, Velalar College of Engineering and Technology, Erode, Tamil Nadu, India
2
ME-CSE , Velalar College of Engineering and Technology, Erode, Tamil Nadu, India
ABSTRACT

Data mining depends on large-scale taxi traces is an important research concepts. A vital direction for analyzing
taxi GPS dataset is to suggest cruising areas for taxi drivers. The project first investigates the real-time demand-
supply level for taxis, and then makes an adaptive tradeoff between the utilities of drivers and passengers for
different hotspots. This project constructs a recommendation system by jointly considering the profits of both
drivers and passengers. At last, the qualified candidates are suggested to drivers based on analysis. The project
also provides a real-time charging station recommendation system for EV taxis via large-scale GPS data mining.
By combining each EV taxi‟s historical recharging actions and real-time GPS trajectories, the present
operational state of each taxi is predicted. Based on this information, for an EV taxi requesting a
recommendation, recommend a charging station that leads to the minimal total time before its recharging starts.
Keywords : Vehicular Social Networks, Hotspot location, Trajectory data mining, Supply-demand level,
Electronic Vehicle.

I. INTRODUCTION messaging. Social networking sites enable users to


share ideas, pictures, posts, activities, events and
A social networking service (SNS) may be a platform interests with folks in their network. The web
to create social networks or social relations among community services area unit generally thought of a
people that share similar interests, activities, social network service, although in an exceedingly
backgrounds or real-life connections. A social broader sense, social network service sometimes
network service consists of an illustration of every suggests that associate degree individual-centered
user usually a profile, his or her social links, and a service whereas on-line community services area unit
spread of extra services. Social network sites area unit group-centered.
web-based services that enable people to make a
public profile produce a listing of users with whom to The social media sites not solely remained a platform
share connections, and think about and cross the to initiate informal dialogues and a supporter of live
connections among the system. Social network sites messages, however became associate integral a part of
area unit varied and that they incorporate new info promoting ways of the many a business homes. Social
and communication tools like mobile property, photo, Networking shortly became how for whole
video sharing. The Most social network services area promoting and promotion on social sphere, whereby,
unit web-based and supply suggests that for users to the enterprises started victimization these on-line
act over the net, like e-mail and instant electronic communities or websites for developing contacts and

CSEIT19515 | Received : 01 Jan 2019 | Accepted : 10 Jan 2019 | January-February -2019 [ 5 (1) : 38-44 ]
38
Selvi C, Keerthana D Int J S Res CSE & IT. January-February-2019 ; 5(1) : 38-44

driving traffic to their several websites. The appliance However, various vary of novel applications, coming
of those sites has unfold to business homes that together transport networks, exploiting quality
started victimization the Social Networking sites as a pattern, socially aware recommendation systems on
platform to market their services and build whole the roads area unit a number of the factors towards
awareness. These social networking websites kind the whom the analysis community has shown concrete
most tool of social media promoting. The foremost interest. Finally, they bestowed some open analysis
ordinarily used websites Twitter and Face book. The issue for future direction.
main objectives of the Taxi Recommendation square
measure B. Vehicular Social Networks: Enabling Smart
Mobility
 To specialize in a way to maximize drivers‟ profits In this article stressed the importance of high-
whereas high the profit of passengers. potency and reliable transmissions in VSNs for
 To value 2 completely different levels of Demand sensible cities. Significantly, we tend to study a case
offer that square measure appropriate for busy on traffic anomaly detection for VSNs by mechanical
(peak) days and traditional operating days. phenomenon knowledge analysis. They believed that
 To provides a time period charging station VSNs can draw in depth attentions and analysis
recommendation system for taxis. efforts within the close to future because the
 To calculate waiting time beside the gap for the integrations of data technology and social network
recharging stations. services become additional compacted. Though VSNs
II. RELATED WORKS will be considered the mixing of social networks and
IoVs to enhance the standard of life for voters, the
A. Vehicular Social Networks: A Survey avenues of VSN studies aren't flat, and plenty of open
problems are still ahead.
Considered social networking in an exceedingly
transport environment; the authors investigated the C. Energy-Latency Trade-Off For Energy-Aware
possible applications of VSNs and communication Offloading In Mobile Edge Computing Networks
design. VSNs enjoy the social behaviors and quality
of nodes to develop novel recommendation systems In this paper, single and multi-cell MEC network
and route coming up with. Further, they gave an eventualities are thought-about at an equivalent time..
outline of various recommendation systems and path In terms of the mixed number nonlinear drawback
coming up with protocols supported crowd sourcing (MINLP) for computation offloading and resource
and cloud-computing with future analysis directions. allocation, we tend to propose AN reiterative search
They bestowed a progressive literature review on rule combining interior penalty operate with D.C.
socially-aware applications of VSNs, information (the distinction of 2 convex functions/sets)
dissemination, and quality modeling. Further, they programming (IPDC) to search out the optimum
mentioned the various communication protocols resolution. The residual energy of good devices‟
style and information dissemination techniques to battery is introduced into the definition of the
deal with the present gap between VSNs and ancient coefficient issue of energy consumption and latency.
ad-hoc networks that is that the terribly 1st issue to Numerical results show that the projected rule will
be thought-about by the analysis community to get lower total price (i.e., the weighted add of energy
understand the construct of VSNs in public accepted. consumption and execution latency) examination
with the baseline algorithms and also the energy-

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


39
Selvi C, Keerthana D Int J S Res CSE & IT. January-February-2019 ; 5(1) : 38-44

aware coefficient issue is of nice significance to keep a tradeoff between the driver‟s and passengers‟
up the life of good mobile devices. utilities.

III. METHODOLOGY 1) Demand Hotspots Scanning By Clustering


A. Existing Work
By clustering the pick-up points, information from
This work constructs an adaptive recommendation taxi trajectory can be extracted to identify candidate
system based on the supply-demand level, by which a demand hotspots. Traditional DBSCAN algorithm is a
tradeoff is made between the utilities of drivers and kind of density-based clustering methods, which can
passengers. Then the hotspot with the highest score is discover arbitrary clusters and deal with noise or
recommended to available taxis. It considers a outliers effectively First, the distance distribution
passenger‟s utility with the waiting time for vacant matrix is calculated, denoted by Dist nxn.
taxis, which is predicted by mining the pick-up
events.
Input: The pick-up points dataset to be clustered P
Output: The final set of clusters C
1: for pi,pj in P do
2: Dist[i][j] ← getManhattandis(pi,pj);
3: end for
4: Sort Dist in an ascending order line by line;
5: for the i-th column vector in Dist do
6: get average value as epsi;
7: end for
8: DBSCAN (epsi , fixed MinPts) ;
9: Select optimal Eps by the number of cluster and noise;
10: N ← 0;
11: for p in P do
12: N + getEpsNeighbourNum(p);
13: end for
14: Perform DBSCAN with optimal Eps and MinPts;
16: return clustering results C.

where n is the number of pick-up points we extract,


and dist(i,j) is the Manhattan distance between GPS
Fig 1 : Taxi Recommendation System point pi and pj. When the value of i increases, the
First pick-up points for each time segment from the number of clusters and noise both decrease.
taxi trajectory are extracted. Then an adaptive
2) Passenger‟s Waiting Time Prediction
Density-based Spatial Clustering of Applications with
Noise algorithm (I-DBSCAN) for clustering is utilized.
The arrival times of passenger for a particular vehicle
Passengers‟ expected waiting time is predicted based
and actual vehicle arrival time is taken. Then the
on the information of different hotspots. For the
average values of waiting times are calculated and
online part, we retrieve hotspots within certain limits
thus the passenger waiting time is predicted. The
for the correct time segment according to the time
following algorithm is used to predict the waiting
and location of available taxis. The essential
time. With the input of pick up events time stamp
knowledge of each hotspot is calculated for online
sequences, the waiting time is calculated.
recommendation. After evaluating the real-time
demand-supply level of the whole area, we can make

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


40
Selvi C, Keerthana D Int J S Res CSE & IT. January-February-2019 ; 5(1) : 38-44

Input: The pick-up events timestamp sequence Input: Available taxi’s current time curtime and location curloc,
Tp = {pe1,pe2,··· ,pen} candidate hotspots set H
Output: The estimated waiting time w for the hotspot Output: The recommended hotspot
1: λ ← n−1 pen−pe1; 1: MaxScore ← 0,MaxId ← 0;
2: Initiaize the passenger arrival events timetamp sequence 2: U ← ø,w ← ø;
Ta = {ae1,ae2,··· ,aen}; 3: Tracing trajectory and computing the driver’s recent spent
3: for i = 1 to n do time on each hotspots ST = {st1,st2,··· ,stn};
4: if i = 1 then 4: for hi in H do
5: aei = random(0,pei); 5: d ← getManhattandis(curloc,hi.core);
6: // Uniform distribution 6: V ← (hi.revenue−βhi.searchingtime−γd) ;
7: else 7: ε ← sti ∑k stk ;
8: aei = aei−1 + random(0,pei −aei−1); 8: U ∪(V + ε);
9: // Truncated exponential distribution with λ 9: w∪hi.waitingtime;
10: end if 10: end for
11: end for 11: Evaluate real-time demand-supply level α based on curtime
12: w ← 0; using previousAlgorithm
13: for i = 1 to n do 12: for U∗ i , w∗ i corresponding to each hotspot do
14: w + (pei −aei); 14: score ← (1−α)U∗ i + αw∗ i ;
15: end for expected waiting time w. 15: if score > MaxScore then
16: MaxScore ← score;
17: MaxId ← i;
3) Demand-Supply Level Evaluation 18: end if
19: end for
The following algorithm is used for demand supply 20: return recommended hotspot hMaxId.

level evaluation. Total time intervals among the


B. Proposed Work
trajectories and total free/busy counts are calculated
and α value is found out.
1) State Inference Algorithm
Input: Record of trajectory points for the taxi R =
{r1,r2,··· ,rn}
Output: The real-time demand-supply level α
1: S ← ø;
This module is used to update the state inference of
2: for each R do the particular vehicle. It includes vehicle id, entry
3: if r.location in area and r.state was FREE then
4: get r.timestamp as ta; time, time interval, x and y position. This module
5: while FREE IN THIS AREA do
6: get next record; also updates vehicle state inference details every
7: end while
8: get r.timestamp as tb; particular second. These details are stored in „State
9: get r.state as m;
10: ∆t ← (tb −ta + 1−m);
Inference‟ table and viewed by using data grid view
11: S ∪(∆t,m); control.
12: end if
13: end for
14: sum1 ← 0,sum2 ← 0;
15: for (∆ti,mi) in S do Find State Inference (Algorithm)
16: sum1 + ∆ti;
17: sum2 + mi;
18: end for
19: α ← sum2/(sum1 + sum2);
This module is used to find the state inference of the
20: return real-time demand-supply level α. particular vehicle by using state inference algorithm.
When the vehicle id is selected it will automatically
4) Adaptive Recommendation
display the time interval. This algorithm has
threshold value based on this value state inference of
The following Algorithm is carried out in which
a particular vehicle such as recharging, operating or
Input is Available taxi‟s current time curtime and
to station can be calculated. These details are updated
location curloc, candidate hotspots set H and Output
every particular second.
is the recommended hotspot. Real-time demand-
supply level α is taken from previous algorithm. Max
Score is found out based on revenue in various pick
up points. Hotspot with max score is recommended.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


41
Selvi C, Keerthana D Int J S Res CSE & IT. January-February-2019 ; 5(1) : 38-44

Input: an EV taxi v, time t, time interval Ik correspond to


rechargeable time and entry time. It also includes
t, v‟s state distribution {Pk(st),Pk(sr),Pk(so)} in Ik, travel vehicle state details such as recharging vehicle count,
distance d, v‟s real-time trajectory trj = p1,p2,...,pm waiting vehicle count and arriving vehicle count. By
Output: predict v‟s state at t. using these details waiting time will be calculated.
1: if Occupied field of pm is 1 then
2: v‟s state at t is operating;
3: else 3) Vechile Optimization
4: if ∃sp ⊂ trj and pm = ˆpm then Particle Swarm Optimization technique is used to
5: v‟s state at t is recharging; achieve the result. Instead of arithmetic operators, set
6: else {∃sp ⊂ trj and pm = ˆpm}
theory based approach is used here. The velocity and
7: v‟s state at t is operating;
8: else
position update rules in S-PSO also have the same
9: if Pk(st)=max{Pk(st),Pk(sr),Pk(so)} and d<τd then format as that in original PSO algorithms but, the
10: v‟s state at t is operating; particle‟s position in S-PSO is represented by a hard
11: else {Pk(st)=max{Pk(st),Pk(sr),Pk(so)} and d>τd} division, i.e., X ⊆ S, instead of the numerical digits in
12: v‟s state at t is to-station;
unique PSO algorithms, and the velocity is
13: else {Pk(so) = max{Pk(st),Pk(sr),Pk(so)}}
14: v‟s state at t is operating; represented by a set with possibilities. And this
15: else {Pk(sr) = max{Pk(st),Pk(sr),Pk(so)}} module are not in existing system , so we used it as
16: v‟s state at t is recharging; proposed technique. The algorithm is a very powerful
17: end if
optimization technique.
18: end if
19: end if 1.Initialize particles
2.Repeat for each particle do
3.Calculate the fitness value if fitness value is better than the best
fitness value (pbest ) in hist
4.Set current value as the new pbest
2) Waiting Time for Fuel Charging Calculation 5.end if
6.end for
Input: A charging station s, the number of charging piles Ns, 7.Choose the particle with the best fitness value of all the particles
current time t, the number of EVs being recharging and waiting at a for each particle do
s is l and n, respectively, the number of EVs arriving at s earlier 8.Calculate particle velocity according to velocity update equation
than v0 is k; 9.Update particle position according to position update equation
Output: Wv 0 (s), i.e., v0’s waiting time at s;. 10.end for
1: if n = 0 then 11.until maximum iterations or minimum error criteria is attained
2: if l + k < Ns then
3: W v 0 (s) = 0;
4: else {l + k > Ns}
5: M = l + k
6: when an EV taxi finishes recharging, M = M − 1 4) Node Failure Scenario
7: denote the time when Ns − M = 1 as t1;
8: W v 0 (s) = t1 − t;
9: end if Delivering the goods in road failure as well as vehicle
10: else {n > 0}
11: when an EV taxi finishes recharging, n = n − 1 failure scenario is also carried out. During the vehicle
12: denote the time when n = 0 as t2;
13: M = l + k routing if any of the node (vertex) in the graph is
14: after t2, when an EV taxi finishes recharging, M =M − 1
15: denote the time when Ns − M = 1 as t3; found to be failure, then alternate path is also found
16: W v 0 (s) = t3 − t;
17: end if
out from one hop before that failure node (vertex).
Node failure scenario is taken in to account. Any
vehicle if found to be failed in intermediate stage,
This module is used to calculate the waiting time of a then the solution works from the remaining available
particular station. It contains information such as vehicles.
station id, number of outlet when the station id is
selected it will automatically display the number of The solution that N-PSO finds for the optimization
outlets, minimum rechargeable time, maximum problem is thus the best position found by the swarm,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


42
Selvi C, Keerthana D Int J S Res CSE & IT. January-February-2019 ; 5(1) : 38-44

represented as a set of elements. In every iteration, The finding seat available between two location and
each element is updated by following two "greatest" average of seat counting Time details as followed:
values. The first one is the greatest solution (fitness)
AVERAGE TIME CALCULATION PSO-
it has achieved so far. (The fitness value is and stored.) NPSO
This value is call pBest. Another" greatest" value that
100
is tracked by the particle swarm optimizer is the
80
greatest value, obtained so far by any element in the

AVG Time
60 PSO
population. This greatest value is a overall best and 40 NPSO
called gBest. 20
0
v[ ] = v[ ] + c1 * rand() * (pbest[ ] - present[ ]) + c2 * rand() * 1 2 3 4 5
(gbest[ ] - present[ ]) + (( c3 * rand()) + a[t])) + (( c4 * rand( )) - Vehicle
s[t])) - (a)present[ ] = present[ ] + v [ ] - (b)
where, a[t]=v[ ]+ pbest+ gbest
a[t]=v[ ]- pbest- gbest
Fig 2. Average Time Calculation PSO- N-PSO
v[ ] is the element velocity, present[ ] is the current element
(solution),
rand ( ) is a random number between (0,1); c1, c2 are learning Example:
factors usually c1 = c2 = 2. Number of Seat Count, C
Number of Distances, D
IV. ARCHITECURE Total Number of Vehicle TV
Total Number of distances TD
N number of Iteration N

AVG= [(C+D) * (TV+TD)/ N] * 1000

VI. CONCLUSION

The project constructs an adaptive recommendation


system by jointly considering the benefits of drivers
and passengers. First, a spatio-temporal clustering
method named I-DBSCAN is leveraged to group pick-
up locations into different clusters. Second, to
improve the drivers‟ utility, kinds of metrics
including expected revenue, driving distance,
searching time and preference are taken into
consideration. Recent advances in designing to find
the shortest path for minimum traveling costs and
V. EXPERIMENTAL RESULTS
number of vehicles without violating the constraints
and loading capacity of vehicle with Time Windows.
Experimental results illustrate the effectiveness and
To optimize the objective of the work, SBPSO
efficiency of the proposed algorithm, for it provides
algorithms have been found to have a global search
better results than the existing results. Experimental
ability, fast convergence speed, strong robustness, etc.,
result for comparison of PSO and N-PSO algorithm.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


43
Selvi C, Keerthana D Int J S Res CSE & IT. January-February-2019 ; 5(1) : 38-44

VI. REFERENCES Transportation Systems, vol. 17, no. 11, pp.


3318–3324, 2016.
[1] Z. Ning, F. Xia, N. Ullah, X. Kong, and X. Hu,
[9] H. Zhou, P. Wang, and H. Li, “Research on
“Vehicular social networks: Enabling smart
adaptive parameters determination in DBSCAN
mobility,” IEEE Communications Magazine, vol.
algorithm,” Journal of Xian University of
55, no. 5, pp. 16–55, 2017.
Technology, vol. 9, no. 7, pp. 1967–1973, 2012.
[2] A. Rahim, X. Kong, F. Xia, Z. Ning, N. Ullah, J.
[10] G. Qi, G. Pan, S. Li, Z. Wu, D. Zhang, L. Sun,
Wang, and S. K. Das, “Vehicular social
and L. T. Yang, “How long a passenger waits for
networks: A survey,” Pervasive & Mobile
a vacant taxi – large-scale taxi trace mining for
Computing, DOI: 10.1016/j.pmcj.2017.12.004,
smart cities,” in Proc. IEEE Green Computing
vol. 43, 2017
and Communications, pp. 1029–1036, 2013.
[3] W. Hou, Z. Ning, L. Guo, and X. Zhang,
[11] D. Shao, W. Wu, S. Xiang, and Y. Lu,
“Temporal, functional and spatial big data
“Estimating taxi demand-supply level using taxi
computing framework for large-scale smart
trajectory data stream,” in Proc. IEEE ICDMW,
grid,” IEEE Transactions on Emerging Topics in
pp. 407– 413, 2015.
Computing, DOI: 10.1109/TETC.2017.2681113,
[12] Z. Ning, F. Xia, X. Hu, Z. Chen, and M. S.
2017.
Obaidat, “Social-oriented adaptive transmission
[4] J. Zhang, X. Hu, Z. Ning, C. H. Ngai, L. Zhou, J.
in opportunistic Internet of smartphones,” IEEE
Wei, J. Cheng, and B. Hu, “Energy-latency
Transactions on Industrial Informatics, vol. 13,
trade-off for energy-aware offloading in mobile
no. 2, pp. 810–820, 2017.
edge computing networks,” IEEE Internet of
[13] X. Wang, Z. Ning, and L. Wang, “Offloading in
Things Journal, DOI:
Internet of vehicles: A fog-enabled real-time
10.1109/JIOT.2017.2786343, 2017
traffic management system,” IEEE Transactions
[5] X. Kong, F. Xia, Z. Ning, A. Rahim, Y. Cai, Z.
on Industrial Informatics, DOI:
Gao, and J. Ma, “Mobility dataset generation for
10.1109/TII.2018.2816590, 2018.
vehicular social networks based on floating car
[14] K. Zhang, Z. Feng, S. Chen, K. Huang, and G.
data,” IEEE Transactions on Vehicular
Wang, “A framework for passengers demand
Technology, vol. 67, no. 5, pp. 3874–3886, 2018
prediction and recommendation,” in Proc. IEEE
[6] Z. Ning, X. Wang, and J. Huang, “Vehicular fog
SCC, pp. 340–347, 2016.
computing: Enabling real-time traffic
management for smart cities,” IEEE Wireless
Cite this article as :
Communications, 2018
Selvi C, Keerthana D, "System for Vehicular Social
[7] X. Kong, X. Song, F. Xia, H. Guo, J. Wang, and
Networks with Fuel Charging Mechanism",
A. Tolba, “LoTAD: long-term traffic anomaly
International Journal of Scientific Research in
detection based on crowdsourced bus trajectory
Computer Science, Engineering and Information
data,” World Wide Web, vol. 21, no. 3, pp.
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
825–847, 2017.
5 Issue 1, pp. 38-44, January-February 2019.
[8] J. Tang, H. Jiang, Z. Li, M. Li, F. Liu, and Y.
Available at doi :
Wang, “A two-layer model for taxi customer
https://doi.org/10.32628/CSEIT19515
searching behaviors using GPS trajectory data,”
Journal URL : http://ijsrcseit.com/CSEIT19515
IEEE Transactions on Intelligent

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


44
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT183812

Aspect Ranking Technique for Efficient Opinion Mining using


Sentiment Analysis : Review
Prof. Sonali D. Borase#1, Prof. Prasad P. Mahale*2
Assistant Professor, Department of Computer Engineering, NMIMS Mukesh patel school of technology,
#

Shirpur, Maharashtra, India


* Assistant Professor, Department of Computer Engineering, SES’s R.C Patel institute of technology, Shirpur,
Maharashtra, India

ABSTRACT

Opinion mining, also called sentiment analysis, is the field of study that analyses people’s opinions, sentiments,
evaluations, appraisals, attitudes, and emotions towards entities such as products, services, organizations,
individuals, issues, events, topics, and their attributes. Even though facts still play a very important role when
information is sought on a topic, opinions have become increasingly important as well. Opinions expressed in
blogs and social networks are playing an important role influencing everything from the products people buy to
the presidential candidate they support. Thus, there is a need for a new type of search engine which will not
only retrieve facts, but will also enable the retrieval of opinions. Such a search engine can be used in a number
of diverse applications like product reviews to aggregating opinions on a political candidate or issue. This paper
consist review works have been designed for opinion mining by using classification and ranking techniques.
Keywords : Sentiment Analysis, Opinion Mining, POS, Ranking Algorithm, Feature Selection Method,
Semantic Orientation.

I. INTRODUCTION

Sentiment analysis, also called opinion mining, is the reach opinions/sentiments regarding a particular
process of extracting subjective information, such as topic, product, event or individual. Sentiment
opinions, sentiments and attitudes in the source analysis process can be modelled as a classification
materials towards an entity. It is an interdisciplinary problem. Sentiment analysis can be conducted at
research field that combines tools and techniques different levels of detail. Based on the levels,
from natural language processing, text mining and sentiment analysis can be broadly divided into three
computational linguistics [1]. Opinions are important main levels: document level, sentence-level and
factors and influencers of the decision-making aspect-level sentiment analysis [3]. Document-level
process. The determination of people’s opinions sentiment classification aims to determine the overall
toward a particular event can be extremely important sentiment orientation of an entire document, such as
in several fields, such as management sciences, a review text, assuming that each document contains
political science, economics and other disciplines of information regarding a single entity. Sentence-level
social sciences [2]. The Web provides a rich and sentiment analysis aims to identify subjective and
progressively expanding source of information to objective sentences. In sentence-level sentiment

CSEIT183812 | Received : 01 Jan 2019 | Accepted : 10 Jan 2019 | January-February -2019 [ 5 (1) : 45-49 ]
45
Prof. Sonali D. Borase, Prof. Prasad P. Mahale,. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 45-49

analysis, the sentiment orientation of subjective accordingly. Each file form target domain is assigned
sentences is also identified.[6] The classification of score on which basis it is classified as positive or
review documents at document or sentence level of negative. Each word has weight assigned to it.
granularity does not fully reveal the opinions Summation of weights of words in each sentence
regarding different features of a particular entity. gives score to sentence. Then addition of all sentence
Hence, aspect-level sentiment analysis concerns the score is nothing but score of file. On the basis of this,
classification of sentiments by focusing on the file is classified. Clustering helps in reducing
particular features/aspects of entities. mismatch between domain specific words of source
and target domains. Two sets of lexicons are extracted
II. LITERATURE SURVEY as an output with polarity which is compared with
SentiWordNet..[1]
Blenty proposed algorithm works on word level. POS
tagged words are extracted from pre-processing steps
are used. Each word acts as feature. Algorithm is Pre-processing of dataset
executed for total number of features provided in
input dataset. According to this weight words are
classified into two categories. Classified words are
having POS tag, polarity tag, and weight value. From Apply classifier on dataset.
this list, common and uncommon words are picked
and used for bipartite graph clustering.

Probability distribution of class c is calculated based Applying clustering on dataset.


on term frequency. Classification process predicts the
polarity of the target domains lexicon from source Fig 1 : major phases in entropy based classifier
domain. The clustering algorithm is applied on
classified word lists and documents until it reaches Nyaung proposed system in which review crawler
convergence. All extracted words from source crawl the related web pages and retrieves review
domain are tagged, and weight is calculated for each comments only. The filtered review comments will
word using mutual information available for words. be proceeding for more processing steps. Subjective
Target words are extracted and compared with the sentences express the reviewer’s sentiment about the
source. If they match then they will be categorized as product and objective sentences do not have any
domain independent, otherwise domain-specific. direct or support of that sentiment.
Domain-independent words are from both source Subjectivity/objectivity determination works in two
and target domains; whereas, domain-specific are phases – training and classification. For training
from target domain only. A graph is constructed phase, manually labelled sentences are used as trained
between domain dependent and domain-independent data, which is later used to identify subjective
words. Co-occurrence relationship between these unigrams for new dataset. In the second phase, the
words represents edge. Occurrence of domain- classification is centred on the probability of
specific word along with domain-independent word unigrams from test dataset using the training data. In
means that both a related to each other and assign the pre-processing, only subjective sentences are
edge. Using domain-independent words weight is submitted to a pipeline for Parts-Of-Speech (POS)
assigned to domain-specific words and classified tags. POS tagging is used for sentence splitting and to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


46
Prof. Sonali D. Borase, Prof. Prasad P. Mahale,. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 45-49

assign lexical categories to the words in text. The feature candidates are identified by POS tags and
feature and opinion pairs, the parser from previous only the features which are stored in the domain
step are analysed and generate all possible ontology are valid. Maximum Entropy modelling for
information components from them. The extracted prediction of feature and opinion pairs in a different
pairs from dependency relation can contain which way it has been used so far, using weights for both to
are relevant or not. Therefore, maximum entropy emphasize the importance of each one of them in the
model is used to predict opinion-relevant product relation of feature and opinion pairing task.
feature or not by choosing the class with the highest Maximum entropy model is used to predict which
conditional probability p according to this model. feature word should be related with the opinion
Domain ontology to get the domain related features word with maximum probability.[3]
and to define the synonym set for features. Product

Table 1: Summery

Year Author Method Limitation

2018 Blenty et Modified Naïve Byes Algorithm It is not support


al unstructured keyword.
2018 Deshmukh Extracts and classifies opinion words from one domain called It is not consider non-
et al source domain and predicts opinion words of another domain word feature
called target domain using a semi-supervised approach,
Which combines modified maximum entropy and bipartite graph
clustering.
2017 Aytug et The presented framework obtains individual feature rankings with
al information gain, chi-square, gain ratio, symmetrical uncertainty,
Pearson correlation coefficient, ReliefF algorithm and
probabilistic significance measure-based feature selection
methods. Then, these individual lists are amalgamated into a
single ranking list via a genetic algorithm.
2015 Dim En Polarity of extracted opinion word calculated with Comparative sentence
Nyaung et sentiword net. not included
al
2014 Zhang et al Probabilistic aspect ranking algorithm. It used only document
level sentiment
analysis.

Deshmukh et al proposed method collected data from identifying polarity of each aspect n finally the
site that provides with reviews on electronic gadgets product itself. Modify Naivè Bayes model to create
and gives the respective user reviews. Pre-processing an aspect ranking algorithm which eventually will
phase on data through Natural Language processing is rank the products where this step calculates the
done and then feature/aspect list is displayed. weight of features from reviews. [2]
Extracting nouns as features and adjectives as
opinions. Next step is Enhanced Aspect Identification Zhang et al proposed system in which given the
and Opinion Word Extraction In this technique High consumer reviews of a particular product, we first
frequency aspects are ranked and the corresponding identify aspects in the reviews by a shallow
opinion words for the aspects are extracted for dependency parser and then analyse consumer

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


47
Prof. Sonali D. Borase, Prof. Prasad P. Mahale,. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 45-49

opinions on these aspects via a sentiment classifier. measures like Information gain-based feature ranking
System develop a probabilistic aspect ranking is a filter-based feature selection method, which is
algorithm, which effectively exploits the aspect widely utilized in text mining domain. The second is
frequency as well as the influence of consumers’ Chi-squared feature ranking evaluates the merit of
opinions given to each aspect over their overall each feature individually with the chi-squared
opinions on the product in a unified probabilistic statistical measure. Next is Information gain is a
model. The overall opinion in a review is generated biased measure towards features with high values.
based on a weighted aggregation of the opinions on Similar to gain ratio, the symmetrical uncertainty
specific aspects, where the weights essentially coefficient aims to eliminate the bias of information
measure the degree of importance of these aspects. A gain towards the attributes with higher values.
probabilistic regression algorithm is developed to Pearson correlation coefficient is used to measure the
infer the importance weights by incorporating aspect correlation between two attributes. Relief algorithm
frequency and the associations between the overall is a filter-based method that uses a feature relevance
opinion and the opinions on specific aspects. In order criterion to rank the features. Proposed genetic rank
to evaluate the proposed product aspect ranking aggregation-based feature selection model uses Filter-
framework. A product aspect ranking framework to based feature selection methods can obtain different
automatically identify the important aspects of rankings for the same data set. The aggregation of
products from numerous consumer reviews. A several different feature rankings may be beneficial
probabilistic aspect ranking algorithm to infer the to obtain an enhanced ranking of the features. The
importance of various aspects by simultaneously proposed method models the feature selection as a
exploiting aspect frequency and the influence of rank aggregation problem and the feature rankings
consumers’ opinions given to each aspect over their obtained by different filter-based methods are
overall opinions on the product.[4]. combined. Given a set of rankings of the same
candidate sets, the rank aggregation seeks to find a
Aytug at al evaluated Feature selection is the process single better ranking from these multiple rankings.
of obtaining an appropriate feature subset from the The rank aggregation problem can be modelled as an
data set so that the classification optimization problem where the objective is to
obtain a final ranking that is the closest to the all
Algorithms can deal efficiently with high- individual ranking lists. [5]
dimensional feature spaces. Feature selection
methods aim to eliminate irrelevant or redundant III. CONCLUSION
features and to reduce the training time required to
build a classification model Feature selection method This paper give brief information on Opinion Mining
can be broadly divided into two groups: filter-based and the recent literature review putting light on the
and Wrapper-based feature selection methods. steps involved in Opinion Mining. Opinion Mining is
the field of study that tracks the mood or attitude of
Filter-based methods evaluate the usefulness of people towards a particular product or topic. It helps
features based on heuristics/evaluation metrics, to find the product with best quality, which is
whereas wrapper-based methods select the features predicted based on the user reviews. It is also used for
based on the performance of a machine learning recommendation system and business intelligence.
algorithm to optimize the predictive performance. There are many challenges and future developments
System evaluated the different individual filter-based possible in Opinion Mining approach like short

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


48
Prof. Sonali D. Borase, Prof. Prasad P. Mahale,. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 45-49

length and irregular structure of the content such as Cite this article as :
named entity recognition, parsing. Opinion mining is
to find the opinion of a person from sentences and Prof. Sonali D. Borase, Prof. Prasad P. Mahale,
then classify them on the basis of polarity. Opinion "Aspect Ranking Technique for Efficient Opinion
mining helps people to know the semantic Mining using Sentiment Analysis : Review ",
orientation for a product by classifying it as positive, International Journal of Scientific Research in
negative or neutral. Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
IV. REFERENCES 5 Issue 1, pp. 45-49, January-February 2019.
Available at doi :
[1]. J. Deshmukh and A. Tripathy, "Entropy based https://doi.org/10.32628/CSEIT183812
classifier for cross-domain opinion mining", Journal URL : http://ijsrcseit.com/CSEIT183812
Applied Computing and Informatics, vol. 14,
no. 1, pp. 55-64, 2018.
[2]. B. Alengadan and S. Khan, "Modified
Aspect/Feature Based Opinion Mining for a
Product Ranking System", 2018 IEEE
International Conference on Current Trends in
Advanced Computing (ICCTAC), pp. 1-5, 2018.
[3]. D. Nyaung and T. Lai Thein, "Feature-Based
Summarizing and Ranking from Customer
Reviews", International Journal of Computer
and Information Engineering, vol. 9, no. 3, pp.
734-739, 2015.
[4]. Zheng-Jun Zha, Jianxing Yu, Jinhui Tang,
Meng Wang and Tat-Seng Chua, "Product
Aspect Ranking and Its Applications", IEEE
Transactions on Knowledge and Data
Engineering, vol. 26, no. 5, pp. 1211-1224,
2014.
[5]. A. Onan and S. Korukoglu, "A feature selection
model based on genetic rank aggregation for
text sentiment classification", Journal of
Information Science, vol. 43, no. 1, pp. 25-38,
2016.
[6]. B. Liu, "Sentiment Analysis and Opinion
Mining," Synthesis Lectures on Human
Language Technologies, vol. 5, no. 1, pp. 1-167,
May 2012

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


49
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19516

Big Data Anonymization in Cloud using k-Anonymity Algorithm


using Map Reduce Framework
Anushree Raj1, Rio G L D’Souza2
1Department of M.Sc. Big Data Analytics, St Agnes Autonomous College, Mangalore, Karnataka, India
2Department of Computer Science and Engineering, St Joseph Engineering College, Mangalore, Karnataka,
India

ABSTRACT

Anonymization techniques are enforced to provide privacy protection for the data published on cloud. These
techniques include various algorithms to generalize or suppress the data. Top Down Specification in k
anonymity is the best generalization algorithm for data anonymization. As the data increases on cloud, data
analysis becomes very tedious. Map reduce framework can be adapted to process on these huge amount of Big
Data. We implement generalized method using Map phase and Reduce Phase for data anonymization on cloud
in two different phases of Top Down Specification.
Keywords: Anonymization, Big Data in cloud, k-Anonymity, Map Reduce, Privacy Preserving

I. INTRODUCTION algorithm [8]. Using data anonymization key pieces


of confidential data are obscured in a way that
Security and privacy are a significant obstacle that is maintains data privacy. Data can still be processed to
preventing the extensive adaptation of public cloud gain useful information. Anonymization data can be
[1]. Cloud Computing is a revolutionary computing stored in cloud and even processed without concern
approach, which provides massive storage and that other individuals may capture the data [9]. Later,
computational capability [2]. It provides user to the results can be collected and mapped to the
implement applications cost effectively without original data in a secure area.
investment and infrastructure. It enables users to
have flexible, on demand access to computing In this paper we propose an anonymization model
resources via Internet [3]. These advantages are the using the Top Down Specification techniques on
causes of security and privacy problems, which huge amount of data along with Map reduce
emerge because data owned by different users are framework. The Top Down Specification has two
stored in cloud servers and users are losing their own phases which includes the map phase which reduces
control on data. the huge amount of data into small chunks of data
and the reduce phase combines the intermediate
Data anonymization is the process of hiding identity results to give back the k anonymous data [10].
or sensitive data from original data. Various Anonymization takes place at both the phases.
algorithms are in use for implementation [4], [5], [6], Generalization method is used on Data that usually
[7]. The most efficient algorithm for data contain sensitive information and this proves the
anonymization is the Top Down Specialization importance of employing anonymity approaches.

CSEIT19516 | Received : 01 Jan 2019 | Accepted : 10 Jan 2019 | January-February -2019 [ 5 (1) : 50-56 ]
50
Anushree Raj, Rio G L D'Souza Int J S Res CSE & IT. January-February-2019 ; 5(1) : 50-56

Generalization is the most common method to make Sensitive Attribute: Attribute or data that an
the data anonymous to provide the privacy. In this individual is sensitive about revealing.
paper we discuss on data anonymization technique in
section II and then show the implementation using There are two main ways through which k-
Map Reduce in section III. Anonymity can be achieved:
Bottom Up Generalization: In Bottom-Up
II. DATA ANONYMIZATION Generalization strategy, the data is initialized to its
current state and generalizations are carried out over
Data anonymizaton is a deep study, broadly adopted attribute values until k-Anonymity is not violated
for privacy preserving in data publishing. [16].
Anonymization techniques highly focus on data
security in public cloud [11]. The main feature of Top Down Specialization: In Top-Down
anonymization techniques is hiding few key data to Specialization strategy, all the attribute values are
maintain data privacy. Data can further be processed initialized to the root value of the hierarchy tree [17].
to achieve necessary information by analysing it. The The specialization is carried out iteratively over the
simple way of preserving privacy is by removing all attribute values, until the k-Anonymity is violated.
those information, which will directly link to data of Top Down Specification
an individual. This process s referred to
anonymization [12]. There are many approaches to Let D be a data set to which the anonymization
gain data anoymization, such as k-Anonymity, l- technique is applied. R denotes the records in the
Diversity and t-Closeness which helps to make the data set. Let A be the attributes, which can identify N
data more private and anonymous, to protect data individuals through Ai where 1 ≤ i ≤ N, k is the
access from any third party users. anonymity parameter used in the anonymization
levels. Anonymization levels are the single
k- Anonymity: mapreduce job execution level from multiple
k-Anonymity is used to hide some key information of mapreduce jobs. The domain values are represented
the users in a database of k users. Then the attackers by the taxonomy tree [18],[21]. Q determines the
can make sure there are only k users but cannot quasi identifier. A record can contain any N number
interpret the information of an individual user in of Quasi Identifiers, Qi ϵ R, 1 ≤ i ≤ N. the sensitive
particular [13], [14]. This method is used to values are denoted by S. In Top Down Specification, a
anonymize the quasi-identifiers to provide privacy to data set is anonymized through specialization
the data. The main aim of k-Anonymity is to make operations. The domain values are replaced with all
each record indistinguishable from at least k-1 other its sub domain values.
records. k-Anonymity assigns properties in specific
ways [15]. The k anonymity data attributes are of In Top Down Specialization, a hierarchy tree is used
three types to represent the attribute values which is initiated by
a root value. The specialization performed iteratively
Key Attribute: Attribute which can identify n over the attribute values, till the k - anonymization is
individual directly. satisfied. The specialization is performed by
Quasi - identifier: Attribute which can be linked with substitution of the parent attribute value by its child
external information to identify an individual. value in taxonomy tree. For example, the patient
record of a hospital is anonymized and published for

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


51
Anushree Raj, Rio G L D'Souza Int J S Res CSE & IT. January-February-2019 ; 5(1) : 50-56

data analysis using Top Down Specialization which is


a generalization algorithm then Table I and II
represent the anonymization for the quasi identifiers.
Here Q = {AGE, SEX}

Table 1. Original Data

NAME AGE SEX EDUCATION PROFESSION Figure 3. Taxonomy tree for Sex
SAM 34 M MASTER BUSINESS
FUNG 45 M DOCTRATE DOCTOR Table 2. Anonymized data
SONA 53 F MASTER PRINCIPAL
RAFI 25 M BACHELORS HR NAM SE PROFESSI
ZAMA 37 F MASTERS MANAGER AGE EDUCATION
E X ON
ALICE 28 F BATCHELORS BUSINESS
[1- AN GRADUATE
BOB 52 M DOCTRATE PROFESSOR * BUSINESS
34] Y SCHOOL
[35- AN GRADUATE
The specialization is performed by replacing the * DOCTOR
49] Y SCHOOL
parent attribute value by its child value in Taxonomy
[50- AN GRADUATE PRINCIPA
Tree *
63] Y SCHOOL L
[1- AN
TAXONAMY TREES AND QUASI IDENTIFIERS * BACHELORS HR
34] Y
[35- AN GRADUATE MANAGE
*
49] Y SCHOOL R
[1- AN
* BACHELORS BUSINESS
34] Y
[50- AN GRADUATE PROFESS
*
63] Y SCHOOL OR

The above table describes the anonymization using


generalization algorithms like TDS for the selected
Figure 1. Taxonomy tree for Education Quasi Identifiers with the taxonomy tree hierarchy.
Algorithm 1 TDS

1) Identify every value in HT to the top most value of


the hierarchy tree.
2) Let ListAttributei initially include the top most
value.
3) While some ca(categorical attribute) in
ListAttributei is valid and useful, do step 4, 5, 6
Figure 2. Taxonomy tree for Age 4) Find the Best specialization from ListAttributei
5) Perform Best on HT and update ListAttributei

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


52
Anushree Raj, Rio G L D'Souza Int J S Res CSE & IT. January-February-2019 ; 5(1) : 50-56

6) Update InfoGain(ca) and AnonymityLossValue(ca), denoted by AfterSpecification(Q).


and validity for ca ∈ ListAttributei BeforSpecfication(Q) - AfterSpecification(Q) is the
7) End while loss of anonymity.
8) return Generalized HT and ListAttributei AnonymityLossValue(v) =
AVERAGEloss(BeforSpecfication(Q) -
Initially ListAttributei contains only the topmost AfterSpecification(Q))
value ListAttributei, in which case it returns the
masked table together with ListAttributei. In order to III. MAP REDUCE FRAMEWORK
maintain the data utility we have to find the best
specialization, otherwise the data utility will decrease Map Reduce is a programming model, which
considerably and can leave data useless for mining computes large volume of data sets with parallel and
purpose. distributed algorithm on cluster. A Map Reduce
program is composed of two user defined functions
The metrics used to find the best specialization are, [22]. A Map function which performs sorting of data
the Information Gain (InfoGain) and Anonymity loss and Reduce function that performs summary
value (AnonymityLossValue). For each specialization operation. Map Reduce is a framework for processing
SP:P child (P) is calculated. parallelizable problems across large datasets using
The Information Gain (InfoGain) and Anonymity loss large number of computers.
value (AnonymityLossValue) values are calculated to
generalize the record R ϵ D using the SC(R). A typical MR-MPI program makes at least three calls
to the MapReduce -MPI library, to perform map(),
SC(𝑅) collate(), and reduce() operations on a MapReduce
𝐼𝑛𝑓𝑜𝐺𝑎𝑖𝑛(𝑅)
if 𝐴𝑛𝑜𝑛𝑦𝑚𝑖𝑡𝑦𝐿𝑜𝑠𝑠𝑉𝑎𝑙𝑢𝑒(𝑅) ≠ 0 object it creates. In a map() key-value pairs are
𝐴𝑛𝑜𝑛𝑦𝑚𝑖𝑡𝑦𝐿𝑜𝑠𝑠𝑉𝑎𝑙𝑢𝑒(𝑅)
=
else generated by each processor. The key-value pairs
{ 𝐼𝑛𝑓𝑜𝐺𝑎𝑖𝑛(𝑅)
produced are stored locally by each processor; a map()
………………. (1)
thus requires no inter-processor communication.
1) Information gain,
Users call the library with a count of tasks to perform
IG(R) is only positive values
and a pointer to a user function; the MapReduce -
|Rc| MPI map() operation invokes the user function
IG(R) = 𝐼(𝑅𝑐) − ∑ ( ) I(Rc)
|Rr| multiple times as a callback. Depending on which
c variant of map() is called, the user function may be
………………… (2)
passed a file name, a chunk of bytes from a large file,
I(Rx) is the entropy measure of Rx where Rx is the
a task ID, or a v pair. Options for assigning map tasks
number of records. (IRx, Sl )is the records with
to processors are specified by the user and include
sensitive value and entropy is computed by,
assigning consecutive chunks of tasks to processors,
|Rca, S| |Rca, S| striding the tasks across processors, or using a master-
𝐼(𝑅𝑥) = − ∑ ( ) X log2 ( ) slave model that is useful when tasks have widely
|Rr| |Rr|
c varying workloads [19].
…………………… (3)
2) Anonymity Loss Value (AnonymityLossValue)
The basic data stored and operated on by any
Here the anonymity before specialization is denoted
MapReduce framework are key-value pairs [20]. In
by BeforSpecfication(Q) and after specialization is
the MapReduce -MPI library, individual keys or

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


53
Anushree Raj, Rio G L D'Souza Int J S Res CSE & IT. January-February-2019 ; 5(1) : 50-56

values can be of any data type or length, or gathered by enterprises and government agencies.
combinations of multiple types (one integer, a string Such data is subjected to, be abused by internal and
of characters, two integers and a double, etc); they external adversaries. Many anonymization
are simply treated as byte strings by the library. A techniques came into existence. They are k-
key-value pair always has a key; its value may be Anonymity, l-Diversity, t-Closeness and m-Privacy.
NULL. Key-value pairs are stored within a These privacy preserving data mining algorithms
MapReduce object. A user program may create one or cannot work for big data analytics, as such data is
more MapReduce objects to implement an algorithm computed using MapReduce programming paradigm
[21]. Various MapReduce operations (map, reduce, in cloud computing environment. MapReduce is used
etc) are invoked on an object and its key-value pairs; to control the parallel processing power of cloud
key-value pairs can also be passed between and computing infrastructure. In this paper, we proposed
combined across objects. algorithms which are used to parallelize k-
Anonymity using MapReduce programming
Conceptually the map and reduce functions supplied paradigm with Hadoop. First, we studied k-
by the user have associated types. Anonymity model and its usefulness in anonymizing
map (k1,v1) → list(k2,v2) data. We implemented the Top Down Specialization,
reduce (k2,list(v2)) → list(v2) a generalization method on k-anonymity so as to
anonymize the original data. We then implemented
Map Reduce approach involves the following four algorithms that could parallelize k-Anonymity using
steps: Map Reduce framework. The proposed algorithm is
capable of anonymizing big data in order to support
1. Map processors, assigns the K1 input Key value, privacy preserving. In future, we can propose an
and provides the processor with all the input data algorithm using MapReduce programming towards
associated with that key value. efficiently anonymizing big data for preserving
2. Run the user provided Map() exactly once for privacy.
each key value, generating output organized by
key values K2. V. REFERENCES
3. Reduce processors, assigns the K2 key value with
all the Map generated data associated with that [1] L. Hsiao-Ying and W.G. Tzeng, “A Secure
key value. Erasure Code-Based Cloud Storage System with
4. Run the user provided Reduce () exactly once for Secure Data Forwarding,” IEEE Trans. Parallel
each K2 key value produced by the Map and Distributed Systems, vol. 23, no. 6, pp.
5. Produce the final output –The Map Reduce 9951003, 2012.
system collects all the reduce output and sorts it [2] D. Zissis and D. Lekkas, "Addressing Cloud
by K2 to produce final outcome Computing Security Issues," Future Generation
Computer Systems, vol. 28, no. 3,pp. 583-
IV. CONCLUSION 592,2011.
[3] RuilinLiu, Hui Wang, ”Privacy –Preserving
Privacy is very important to protect the sensitive data Data Publishing “ IEEE ,2010.
from the attacker. To provide data privacy the [4] B.C.M. Fung, K. Wang, and P.S. Yu,
anonymization methods may be used. The data "Anonymizing Classification Data for Privacy
mining activities utilize huge amount of data Preservation," IEEE Trans. Knowledge and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


54
Anushree Raj, Rio G L D'Souza Int J S Res CSE & IT. January-February-2019 ; 5(1) : 50-56

Data Eng., vol. 19, no. 5, pp. 711-725, May Systems, Volume 10 Issue 5, October 2002,
2007. Pages 571 - 588
[5] X. Xiao and Y. Tao, "Anatomy: Simple and [15] Meyerson A and Williams R, On the
Effective Privacy Preservation." Proc. 32nd complexity of optimal k-anonymity. In
Int'l Conf. Very Large Data Bases (VLDB'06), PODS04: Proceedings of the twenty fourth
pp. 139-150,2006. ACM SIGMOD-SIGACT-SIGART symposium
[6] K. LeFevre, D.J. DeWitt, and R. Ramakrishnan, on Principles of database systems, pages 223–
"Incognito: Efficient Full-Domain K- 228, 2004
Anonymity," Proc. ACM SIGMOD Infl Conf. [16] Ke Wang, Philip S. Yu, Sourav Chakraborty,
Management of Data (SIGMOD '05), pp. 49-60, "Bottom-Up Generalization: A Data Mining
2005. Solution to Privacy Protection", Fourth IEEE
[7] K. LeFevre, D. l DeWitt, and R. Ramakrishnan, International Conference on Data Mining,
"Mondrian Multidimensional K-Anonymity," 2004. ICDM '04 .. Pages 249 – 256
Proc. 22nd Int'! Conf. Data Eng. (ICDE '06), [17] Xuyun Z, Laurence T Yang ,”A scalable Two
2006. phase Top Down Specialization Approach for
[8] 1 Xu, W. Wang, 1 Pei, X. Wang, B. Shi, and A. Data Anonymization using Map Reduce on
W. Fu. , "Utility-based anonymization using cloud“, IEEE Transaction on Parallel and
local recoding ", In ACM SIGKDD, 2006. Distributed Systems ,TPDSSI-2012.
[9] S.Yu, ”Anonymizing Classification Data for [18] Zhang X, Yang LT, Liu C, Chen J. A scalable
Privacy Preservation“. IEEE Transactions on two‑phase top‑down specialization approach
Knowledge and Data Engineering ,vol 19 no 5 for data anonymization using MapReduce on
,2007 cloud. IEEE Trans Parallel Distrib Syst.
[10] Benjamin C.M Fung, Ke Wang, Philip S.Yu 2014;25(2):363–73
“Top Down Specialization for Information and [19] Jefrey Dean and Sanjay Ghernawat “Map-
Privacy Preservation”. Reduce: Simplified Data Processing on Large
[11] Bayardo R and Agrawal R, Data privacy Clusters” Google,Inc.2004
through optimal k-anonymization. In ICDE05: [20] J. Dean and S. Ghemawat, "Mapreduce:
The 21st International Conference on Data Simplified Data Processing on Large Clusters,"
Engineering, pages 217– 228, 2005. Comm. ACM, vol. 51, no. 1, pp. 107-113,2008
[12] HIPAA 2012 “k-anonymity : A model for [21] Al‑Zobbi M, Shahrestani S, Ruan C.
Protecting Privacy “ International Journal Sensitivity‑based anonymization of big data. In:
Uncertain Fuzz .vol 10 ,no,5 pp 557-570 ,2002. Local computer networks work‑ shops (LCN
[13] Ciriani, S. De Capitani di Vimercati, S. Foresti, workshops), 2016 IEEE 41st Conference on.
and P. Samarati, "k-Anonymity Data Mining: A IEEE; 2016. p. 58–64
Survey", Springer US, Advances in Information [22] Al‑Zobbi M, Shahrestani S, Ruan C.
Security (2007) Implementing a framework for big data
[14] Latanya Sweeney, School of Computer Science, anonymity and analytics access control. In:
Carnegie Mellon University, Pittsburgh, Trustcom/BigDataSE/ICESS, 2017 IEEE. IEEE;
Pennsylvania, "Achieving k-anonymity privacy 2017. p. 873–80.
protection using generalization and [23] Al‑Zobbi M, Shahrestani S, Ruan C.
suppression", International Journal of Multi‑dimensional sensitivity‑based
Uncertainty, Fuzziness and Knowledge-Based anonymization method for big data. In: Elk‑

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


55
Anushree Raj, Rio G L D'Souza Int J S Res CSE & IT. January-February-2019 ; 5(1) : 50-56

hodr M, Shahrestani S, Hassan Q, editors.


Networks of the future: architectures,
technologies, and implementations. Boca
Raton: Chapman and Hall/CRC Computer and
Information Science Series, Taylor & Francis;
2017. p. 448.

Cite this article as :

Anushree Raj, Rio G L D'Souza, "Big Data


Anonymization in Cloud using k-Anonymity
Algorithm using Map Reduce Framework",
International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 50-56, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT19516
Journal URL : http://ijsrcseit.com/CSEIT19516

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


56
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1| ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19518

Improvement of Privacy and Security in Hybrid Cloud with


Attribute Group Based Access Control
Kayalvili S1, Sowmitha V2
1Computer Science and Engineering, Velalar College of Engineering and Technology,Erode, Tamilnadu, India
2 ME-CSE, Velalar College of Engineering and Technology, Erode, Tamilnadu, India
ABSTRACT

Cloud computing enables users to accumulate their sensitive data into cloud service providers to achieve
scalable services on-demand. Outstanding security requirements arising from this means of data storage and
management include data security and privacy. Attribute-based Encryption (ABE) is an efficient encryption
system with fine-grained access control for encrypting out-sourced data in cloud computing. Since data
outsourcing systems require flexible access control approach Problems arises when sharing confidential
corporate data in cloud computing. User-Identity needs to be managed globally and access policies can be
defined by several authorities. Data is dual encrypted for more security and to maintain De-Centralization in
Multi-Authority environment.
Keywords:ABE-Attribute Based Encryption, Authority, Fine Grained, Outsourcing, Dual Encryption.

I. INTRODUCTION Cloud computing provides the tools and technologies


to create data/compute intensive parallel applications
Cloud computing is internet-based computing during with far morereasonablecosts compared to ancient
whichmassiveteams of remote serverssquare measure parallel computing techniques. In Single-authority
networked to permit sharing of data-processing tasks, ABE, every user’s keys area unit generated
centralized information storage, and on-line access to victimizationtotally different random and on the QT
pc services or resources. Clouds are often classified as shared values specified keys generated for various
public, personal or hybrid. Cloud computing that users can't be combined, that prevents collusion
depends on sharing computing resources instead of attacks.
having native servers or personal devices to handle
applications. The mostsanctioning technology for The main objective of this paper isn't to use a central
cloud computing is virtualization. In basic ABE authority to manage users and keys, and
systems, the data shared is oftenat intervals one solelystraightforward trust relations have to be
domain or organization. However, in reality, compelled to be fashioned by sharing the general
infolikes drivers’ licenses andregistration info in public key between every attribute authority (AA).
universities area unit organized by totally different User identities square measuredistinctive by
government departments. The management of combining a user’s identity with the identity of the
attributes and key distributions can't be undertaken AA wherever the user is found. Once a key request
by an equivalent attribute authority. Moreover, has to be created to associate authority outside the
access methodscould also be distributed supported domain, the request has to be performed by the
attributes of various authorities. authority within the current domain instead of by

CSEIT19518 | Received : 01 Jan 2019 | Accepted : 10Jan2019 | January-February-2019 [ 5 (1) : 57-62 ]


57
Kayalvili S, Sowmitha V.Int J S Res CSE & IT. January-February-2019 ;5(1):57-62

the users, so, user identities staypersonal to the AA B. Persona: An Online Social Networkwith User-
outside the domain, which is able to enhance privacy Defined Privacy
and security. Here, once the user ought to requests
associate attribute secret key, if the attributes square These network help users share in order with their
measuresituated outside the domain, the request by friends. However, users entrust thesocial network
the supply AA within the domain to the target AA is provider with such personal information as sexual
employedinstead of by requests by users themselves. preferences, supporting and religious views, phone
numbers, occupations, identities of friends, and
II. RELATED WORKS photographs. Although sites over privacy controls
that let users restrict how their data is viewed by
A. AData Outsourcing Architecture Combining other users, sites provide inadequate controls to
Cryptography and Access Control restrict data sharing with shared affiants or
application developers.
The recent adoption and diffusion of the information
out-sourcing paradigm, whereverknowledge house C. Fuzzy Identity-Based Encryption
owners store their knowledge on external servers,
there are increasing general demands and issues for In this paper, discussed the use fullness of using
knowledge confidentiality. Besides well-known risks biometric s in Identity-Based and then discuss their
of confidentiality and privacy breaks, threats to out- contributions. Using biometrics in Identity- Based
sourced knowledgeembody improper use of Encryption in many situations, using biometric–based
information: the server may use substantial individuality in an IBE system has a number of
components of a setof knowledge gathered and important advantages over “standard” IBE. In
arranged by the information owner, doubtless standard Identity-Based Encryption scheme a user
harming the information owner's marketplace for with a positive identity, for example, “Bob Smith”,
any professional duct or service that comes with that will need to go to an authority to obtain the private
assortmentof data. The projected novel access key corresponding to the identity. In that process the
management model and design that eliminates the user will need to “prove” to the authority that he is
necessity for a reference monitor and depends on indeed entitled to this identity. That will typically
cryptography to make sure confidentiality of involve presenting supplementary documents or
knowledgehold on a server. Knowledgearea unit credentials. Typically, there exists a trade-off
encrypted because theknowledge owner stores them between a system that is expensive in this step and
on associate in nursing external server. one that is less reliable.
Authorizations and codingarea
unitunitedthereforepermitting access D. Mediated Cipher text-Policy Attribute-Based
managementsocial control to be outsourced besidethe Encryption and Its Application
information. The nice advantage is that the
information owner, whereas specifying the policy, Distributed data systems needversatile access
wants not b econcerned in its social control. management models thattranscend discretionary,
necessary and role-based access management. On the
opposite hand, the present trend of service-based data
systems and storage outsourcing needenhanced
protection of knowledgetogether with access

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


58
Kayalvili S, Sowmitha V.Int J S Res CSE & IT. January-February-2019 ;5(1):57-62

managementways that area unit cryptographically control can be achieved by dual encryption method
implemented. The conception of Attribute-Based which takes advantage of the attribute-based
coding (ABE) fulfills the samenecessities. It provides encryption and discerning group key distribution in
a chicmanner of encrypting knowledgesuch the each attribute group. This work demonstrates how to
encryptor defines the attribute set that the apply the proposed mechanism to securely manage
decryptorhas topossesso as to decrypt the cipher- the outsourced data. The analysis results point to that
text. . A secret key holder willdecipher the cipher the proposed system is efficient and protected in the
text if the attributes related to his secret key satisfy data outsourcing systems.
the access policy related to the cipher text. In these
schemes, a ciphertext is related toassociate in nursing The concept of attribute-based encryption (ABE) is a
access policy and therefore the user secret promising approach that fulfills these requirements.
keysrelated toa collection of attributes. ABE features a device that enables an entrance
control over encrypted data using access policies and
III. METHODOLOGY qualified attributes among private keys and cipher
texts. Especially, cipher text-policy ABE (CP-ABE)
1. EXISTING WORK provides a scalable way of encrypting data such that
the encryption defines the attribute set that the
When the system is set up, the public key of each AA decryptor needs to have classify to decrypt the cipher
and the basic parameters are distributed, which text.
simplifies the process of trust establishment. The key
issuing protocol for privacy only needs to use the 2. PROPOSED WORK
public key of AA to realize the trust between AA.
Present data outsourcing systems require supple  Attribute Group - Based Encryption is managed.
access control approaches. In many cases, it is It adapts a dual encryption approach to overcome
attractive to provide differentiate access services such the user access control problem.
that data access policy are defined over user  Multiple attribute groups are included and data is
attributes or roles. Some of the most difficult issues in distributed among them. User privileges may be
data outsourcing scenario are the enforcement of varying for data maintained by different attribute
union policies and the support of policy updates. groups.
Cipher text-policy attribute-based encryption is a  The data owner maintains all the membership
hopeful cryptographic solution to these issues for lists to enable the direct user management in
enforces access control policies defined by a data more than one place.
owner on outsourced information.  Keys are assigned randomly and independently
from each other. But group key mechanism is
But, the problem of applying the attribute-based common for users in single group
encryption in an outsourced architecture introduces  All the data is maintained by more authority
several challenges with regard to the attribute and groups.
user revocation. So the project also proposes an access
control mechanism using cipher text-policy attribute-
based encryption to enforce right to use control
policies with well-organized element and user
revocation capability. The fine-grained way in

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


59
Kayalvili S, Sowmitha V.Int J S Res CSE & IT. January-February-2019 ;5(1):57-62

command button is user to terminate the current


modules.
d) Attribute Key Generation

This module is used to process the key generation


process and the access structure modules is used to
create the access specification for each and every user
for specifying the details with the rights of select,
insert, update, delete operation those process is
selected by the check box control, attribute identity
number and user identity numbers are selected by
user from the Combo Box control.
Fig 1.1. Privacy and Security Improvement

e) Encryption
a) Trusted Key Pair
Encrypt the text using public key for the purpose of
The trusted key pair created in the application for other users not knows the given message. So, the
further process, following key generate the public unrestricted key is extract using get key grasp button
key and master key, for the purpose of encrypt and displayed in the label control, the message is
the message and group key is created. These entered in the textbox manage then the given
information are generate by form command encrypted communication is display in the label
button event and showed in the multiline text control. The encrypted message saved in the
mode. These key saved in the relevance using application using creates cipher text and save
save authority key event. command button event.

b) Attribute Creation f) Decrypt Cipher Text

It contains details such as attribute id, attribute It is used to retrieve the plain data in the application.
names are entered by user in the textbox controls, The specified cipher text is entered the data is
and saved by the save command button. The showed to the user. In these module user
delete button is used is used to delete the specific characteristics number and nobody texts are selected
record and close command button is user to from the combo box control, group identity is display
terminate the current modules. in the label controls. The significance is decrypted in
the cipher text grid view control using the decrypt
c) User Creation authority button occurrence.

Used to create the client facts for accessing the The following Table describes experimental result for
attribute with privilege level. The client id, user proposed system secure attribute selection analysis.
first name and passwords are enter by user in the The table contains number of time slot interval and
textbox control these details are saved by the save given attribute to calculate average numbers of
command button event. The delete button is used selection attribute details are shown.
is used to delete the specific record and close

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


60
Kayalvili S, Sowmitha V.Int J S Res CSE & IT. January-February-2019 ;5(1):57-62

Member Node Existing System Proposed RATIO OF SELECTION ATTRIBUTE SELECTION


(%) System (%) (AS)- [Proposed System]

8 72.54 78.62 1.2


1
12 76.13 78.11 RATIO OF

Ratio [%]
0.8
SELECTION
0.6
16 82.42 83.13 ATTRIBUTE
0.4 SELECTION (AS)
24 86.66 84.67 0.2
0
30 88.13 89.78 1 3 5 7 9 11
Number of Attribute
32 80.44 82.66
38 78.33 80.21 Fig 1.2 Attribute selection- Ratio Analysis
42 87.22 89.76
46 79.22 80.65 VII. CONCLUSION

 Authorization policies and the support of policy


IV. CIPHER TEXT-POLICY ATTRIBUTE-BASED updates are studied.
ENCRYPTION  Proposes a cryptographic approach for
STEP 1: communication between users and attribute
The setup algorithm is executed which is a authorities.
randomized algorithm that takes no input other than  Allows a data owner to define the access
the implicit security parameter of the public key PK control policy and enforce it on his outsourced
and a master key MK. data.
 Efficient and scalable to securely manage the
STEP 2: outsourced data.
The attribute key generation algorithm is
executed which takes input the master key MK, a set of
VIII. FUTURE WORK
attributes  L, and a set of user indices U u as
parameters. It outputs a set of private attribute keys SK
for each user in U that identifies with the attributes set.  It allows a data possessor to define the access
control guidelines and enforce it on his
STEP 3: outsourced data.
The key encrypting key (KEK) generation  It also features a method that enables more
algorithm is executed in this module, which takes a set fine-grained access control with efficient
of user indices U u as input, and outputs KEKs for attribute and user revocation capability. The
each user in U, which will be used to encrypt attribute
proposed scheme will be efficient and scalable
group keys Ki for each Gi G.
to firmly manage the outsourced data.
.
 The innovative system is designed such that
that enhancement can be integrated with
current modules easily with less integration
work.

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


61
Kayalvili S, Sowmitha V.Int J S Res CSE & IT. January-February-2019 ;5(1):57-62

IX. REFERENCES ACM TOCS , 1(3):239{248, August 1983.


Germany, Septemb er 2007.
[11]. J. Crampton, K. Martin, and P. Wild. On key
[1]. S. Vimercati, S. Foresti, S. Jajodia, S. Paraboschi,
assignment for hierarchical access control. In
and P. Samarati, "A Data Outsourcing
Proc. of the 19th IEEE CSFW'06 , Venice, Italy,
Architecture Combining Cryptography and
July 2006.
Access Control," Proc. ACM Workshop
[12]. G. Miklau and D. Suciu. Controlling access to
Computer Security Architec-ture (CSAW ’07),
published data using cryptography. In Proc. of
Nov. 2007.
the 29th VLDB Conference , Berlin, Germany,
[2]. L. Ibraimi, M. Petkovic, S. Nikova, P. Hartel,
Septemb er 2003.
and W. Jonker, "Mediated Ciphertext-Policy
[13]. H. Hacig Aum Aus, B. Iyer, and S. Mehrotra.
Attribute-Based Encryption and Its
Providing database as a service. In Proc. of 18th
Application," Proc. Int’l Workshop Information
ICDE , San Jose, CA, USA, February 2002.
Security Applications (WISA ’09), pp. 309-323,
2009.
[3]. R. Baden, A. Bender, N. Spring, B.
Cite this article as :
Bhattacharjee, and D. Starin,"Persona: An
Online Social Network with User-Defined
Kayalvili S, Sowmitha V, "Improvement of Privacy
Privacy,"Proc. ACM SIGCOMM ’09, Aug. 2009.
and Security in Hybrid Cloud with Attribute Group
[4]. A. Sahai and B. Waters, "Fuzzy Identity-Based
Based Access Control", International Journal of
Encryption," Proc.Eurocrypt ’05, pp. 457-473,
Scientific Research in Computer Science, Engineering
2005.
and Information Technology (IJSRCSEIT), ISSN :
[5]. V. Goyal, O. Pandey, A. Sahai, and B. Waters,
2456-3307, Volume 5 Issue 1, pp. 57-61, January-
"Attribute-BasedEncryption for Fine-Grained
February 2019.
Access Control of Encrypted Data," Proc. ACM
Available at doi :
Conf. Computer and Comm. Security, pp. 89-
https://doi.org/10.32628/CSEIT19518
98, 2006.
Journal URL :http://ijsrcseit.com/CSEIT19518
[6]. J. Anderson. Computer security planning
study.Technical Rep ort 73-51, Air Force
Electronic System Division, 1972.
[7]. J. Saltzer and M. Schro eder. The protection of
information in computer systems.
Communications of the ACM , 17(7), July 1974.
[8]. N. Provos . Encrypting virtual memory. In
Proc. of the 9th USENIX Security Symposium ,
Denver, Colorado, USA, August 2000.
[9]. A. Harrington and C. Jensen. Cryptographic
access control in a distributed ¯le system. In
Proc. of the 8th SACMAT , Como, Italy, June
2003.
[10]. S. Akl and P. Taylor. Cryptographic solution to
a problem of access control in a hierarchy.

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


62
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195113

Survey on Factors Determining the Share Price Movement in


Indian Stock Market - A Big Data Approach
Nitha K. P.*1, Dr. S. Sivakumari2
1Research Scholar, Department of Computer Science and Engineering, Faculty of Engineering, Avinashilingam
Institute for Home Science and Higher Education for Women Coimbatore, Tamil Nadu, India
2 Professor and Head, Department of Computer Science and Engineering, Faculty of Engineering,
Avinashilingam Institute for Home Science and Higher Education for Women Coimbatore, Tamil Nadu, India

ABSTRACT

The aim of investors of any stock is to earn reasonable returns. This paper presents a framework on share price
movement and investigates the factors determining and predicting the share prices in Indian stock market through
big data approach. This study is also trying to focus on the fundamental factors, technical factors and sentimental
factors that stimulate the share price movement. The methodology used mainly focuses on the implications of the
theory of financial markets and features affecting the share price movement. Stock market is one of the burning
areas where data is growing day by day. Big data Analytics is a very promising area and buzzword for the next
generation information technologies. Because of the heterogeneity and other complexities of data, big data
architecture and design, is needed which specifically deals with the stock market data and analyze the heterogeneous
data for the future prediction of the market. Variety and large Volume of factors exists. Moreover, the velocity with
which data generates and the effect of them in the share price movement emphasizes, role of big data in predicting
the share prices in Indian Stock Market.
Keywords : Stock Market, Big Data, Stock valuation model, Technical, Fundamental and Sentimental Factors

I. INTRODUCTION ratio, dividend yield, and dividend payout. This study


also examines the importance of technical factors and
For an investor, predicting the share price is sentimental factors, which highly influences the
significant and critical. Data required for the stock share price movement of Indian companies.
market prediction is placed and generated at distinct
places of the world. Industry specific factors such as A. FUNDAMENTAL FACTORS
EPS, book value, P/E ratio and dividend yield has 1) Book Value (BV) : It is also known as net asset
influenced the share price movement of Indian value per share because it measures the amount
companies a lot. The present study has been of assets, which the corporation has on behalf of
undertaken with an objective to review the existing each equity share. BV shows the net investment
literature by examining the relationship between per share made in the business by the
stock prices and company specific intrinsic factors or shareholder. It is the value at which an asset is
fundamental factors such as book value per share, carried on a balance sheet.
dividend per share, earning per share, price earnings 2) Earning Per Share (EPS): The ratio of the profit

CSEIT1838116 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 63-69 ]
63
Nitha K. P., Dr. S. Sivakumari Int J S Res CSE & IT. January-February-2019 ; 5(1) : 63-69

after tax of the company for any financial year associated with the matching price chart of the
after payment of preference dividend. The equity security. Technical indicators offer many uses such
shareholders are the sole appellants to the net as confirm the trends, generate buy/ sell signals,
earnings of the corporation after making forecast the trend of future prices. Technical
payment of dividend to the preference analysis is a technique, which gives an idea about
shareholders. The significance of this ratio flows future share prices of selected companies in which
from the fact that higher the earnings per share we invest. Based on the knowledge of technical
the more is the scope for a higher rate of analysis one can predict the perfect investment
dividend and of retained earnings, to build up decision of the stock market. By using the nominal
the internal power of the company. pointers, the future market of securities will know
3) Dividend per share (DPS): Dividend is the in which stock to invest. The more accurate
portion of the profit after tax, which is prediction of stock prices of selected companies the
distributed to the shareholders for their investor to carry out fundamental analysis of stock
investment bearing risk in the company. It has a prices, they can predict of future trend of stock
significant influence on the market price of the prices.
share. DPS shows how much the company has
paid out as dividend. It refers to the actual 1) Moving Averages: Most visual aids shows a lot of
amount of dividend (gross) declared per share. discrepancy in price movement. This can make it
The net profit after taxes belong to shareholders difficult for traders to get an idea of a security’s
but the income that they really receive is the overall trend. One simple method traders use to
amount of earnings distributed and paid as cash combat this is to apply moving averages. A
dividend. moving average is the average price of a security
4) Price Earnings Ratio (P/E ratio): This ratio over a set amount of time. They smooth a data
enables an investor to make appropriate series and made it easier to spot tendencies,
calculation of the time required to cover the something that is especially helpful in volatile
investment in a company’s stock. P/E ratio markets.
expresses the relationship between the market 2) Simple Moving Average (SMA): A simple moving
price of a company’s share and its earnings per average is formed by computing the average
share. It indicates the extent to which the (mean) price of a security over a specified
earnings of each share are covered by its price number of periods. For example: a10 day moving
average(SMA) is taken as the closing prices for
B. TECHNICAL FACTORS the last 10 days added together and dividing the
total by 10.
Technical indicators are mathematical formulas 3) Exponential Moving Average (EMA): Exponential
that, when applied to security prices clearly flash moving average is the moving average that is
either buy or sell indications. Price data includes formed by applying weight to the recent price
any combination of the open, high, low, or close changes.
over a period and most of the indicators use only the
closing prices, technical indicators are usually EM A(c) = ((price(c) EM A (p) ∗ K) + EM A (p))(1)
shown in a graphical form above or below a
security’s price chart for the analysis. Once shown Where c=current, p=previous
in graphical form, an indicator can then be

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


64
Nitha K. P., Dr. S. Sivakumari Int J S Res CSE & IT. January-February-2019 ; 5(1) : 63-69

The formula for the smoothing constant K is: polarity” sentiment classification looks, for
instance, at emotional states such as “angry,”
K = 2/ (1 + N) (2) ”sad,” and “happy.”

Where, N = number of periods for EMA Existing methods applied to extract the sentiment
4) Relative Strength Index (RSI) : The relative analysis, which grouped mainly into four
strength index (RSI) is an extremely useful categories: keyword spotting, lexical affinity,
oscillating momentum indicator that was statistical methods, and concept-level techniques.

developed by J. Welles Wilder and is one of the Keyword spotting categorizes text based on the
presence of unambiguous affect words such as
most widely used indicators in technical analysis.
happy, sad, afraid, and bored. Lexical affinity
RSI oscillates between zero and 100. The most
detects an affect words and assigns a probable
popular is the 14 days RSI where the RSI is
affinity to particular emotions. Statistical
calculated based on 14 days values. Traditionally methods has control on elements from machine
the stock is considered to be over bought when learning such as latent semantic analysis, support
RSI is above 70 and over sold when RSI is below vector machines, “bag of words” and Semantic
30. Signals can also be generated by looking for Orientation Point-wise Mutual Information.
divergences and central line crossovers. The
basic formula is: Concept-level approaches leverage on elements
from knowl- edge representation such as
RSI = 100[100/(1 + RS)] (3) ontologies and semantic networks and, hence, are
also able to detect semantics that are expressed in
Where, RS = average of upward price change over a a subtle manner, e.g., through the analysis of
select number of days / average of downward price concepts that do not explicitly convey relevant
change over the same number of days information, but which are implicitly linked to
other concepts that do so.The main data sources
C. SENTIMENTAL FACTORS for the sentimental analysis are taken from
Twitter as they provide an API to extract the
Sentiment analysis extracts subjective tweet data from the site. The following features
information in source materials with the help of capture useful characteristic of Twitter and
natural language processing, text analysis and writers for opinion retrieval.
computational linguistics. Sentiment analysis
aims to determine the attitude of a speaker or a II. METHODS FOR STOCK MOVEMENT
writer with respect to some topic or the overall PREDICTION
contextual polarity of a document. The
elementary job in sentiment analysis is to The Support Vector Machine (SVM) been
categorize the feelings of a given text in the recognized as being able to efficiently handle
document, sentence, and the feature/aspect level high dimensional data and has shown to perform
of the expressed opinion in a document, or a well on many classification and clustering
sentence or an entity feature/aspect is positive, techniques
negative, or neutral. Advanced, “beyond 1) Price only: In this method, only past data or
prices predicts the stock movement. The

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


65
Nitha K. P., Dr. S. Sivakumari Int J S Res CSE & IT. January-February-2019 ; 5(1) : 63-69

purpose of this method is to investigate topics. Latent Dirichlet Allocation (LDA)


whether there are patterns in the history of (Blei, Ng, & Jordan, 2003) is a generative
the stock or not. In addition, this model was probabilistic model of a corpus. The basic idea
used as a standard to estimate whether is that documents are represented as random
combination of the sentiments is effective for mixtures over latent topics, where each topic
the approach by comparing with other is characterized by a distribution over words.
sentiment models. Features used for the Therefore, the LDA chosen as a simple topic
training of SVM are price t1 and price t2 model to discover these hidden topics.
which are the price movements (up, down) at 5) JST-based method: The method refers to
the transaction dates t 1 and t 2, opinion often expressed on a topic or aspect.
respectively. When people post the message on the social
2) Human sentiment: In addition to historical media to express their opinion for a given
prices, this model integrates the sentiments stock, they tend to talk their opinions for a
annotated by human into the prediction certain topic such as profit and dividend.
model. According to the Message Board Based on pairs of topic-sentiment, they would
dataset, the users explicitly attach a sentiment think that the future price of that stock
label with their posts. The sentiment tags like would go up or down. From that intuition, a
strong buy, buy, hold, sell and strong sell are new feature topic-sentiment for the stock
usually seen labels attached with the posts. prediction model results.
Instead of using all the messages, only the 6) Aspect-based sentiment: As an alternative of
messages with annotated sentiments by the considering, the blends of hidden topics and
users need to be considered, and all the other sentiments as in the previous model, in this
irrelevant messages can be discarded. From model the combinations not hidden. Each
these messages, only the explicit sentiment message represented as a list of topics and
will be taken and remove other information their corresponding sentiment values are
such as message content. The perseverance of considered. The topic selected as the
this method is that how attitude interpreted consecutive nouns in the sentence will
by human can be used to predict the stock. provide tremendous results in prediction.
Because human marks the sentiments, this
feature is one of the strongest features for III. IMPORTANCE OF BIG DATA IN SHARE PRICE
stock prediction. PREDICTION
3) Sentiment classification: To utilize the
remaining 84.4% of the messages without the Data required for the stock market prediction placed
explicit sentiments, the build model extracted and generated at different places of the world.
the sentiments for those messages. A Distinguished agencies are maintaining this data in its
classification model will be trained from the own format. So, to cater need of current scenario, big
messages with annotated sentiments on the data architecture for the stock market has framed. Big
training dataset. Then it can used to classify data architecture for stock market divided into three
the remaining messages into five classes parts. i) First part is to identify different data and its
(Strong Buy, Buy, Hold, Sell, and Strong Sell). sources required for the future prediction of the
4) LDA-based method: In this model, we market.ii) Second Part is to store, acquire and process
consider each message as a mixture of hidden of the heterogeneous, unstructured and temporal

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


66
Nitha K. P., Dr. S. Sivakumari Int J S Res CSE & IT. January-February-2019 ; 5(1) : 63-69

data.iii) Last part of the architecture is to attain its Stock market is affected by all information. As per
goal, the prediction of the stock price. the efficient market hypothesis, stock market must
react on all of the information. Small and Middle
News and Information flows from any corner have class investors have not such type of resources to
impact on investment. Every second data generated know all happenings instantaneously and take their
for the market and its format is different from positions respectively. Loss fear factor had been
previous dataset. To store and integrate these playing a crucial role for the small and middle class
temporal data for the stock market is the biggest population and that is why they are keeping
challenge. In current scenario, cost required for these distance from the market. Easy availability of this
data is affordable by few big investors. Investors from integrated data from all corners of the globe is the
third and fourth world countries have acute scarcity key issue.
of resources. Big data architecture with latest
technologies will provide easy access of information Due to temporal nature of the data required in the
flow; it is relatively affordable to all investors. Little, stock market, big data architecture ,being
and slow access of information became big reason for implemented in such a way that all data must be
the big loss in the market for small investors. All updated and integrated on timely basis and reports
information need would be integrated very fast and should be generated based on the current
provide information to the small and medium information. To achieve efficient market hypothesis,
investors at very reasonable & affordable cost. these information must reach all participates of the
market. Due to shared resources and other above
Traditional database will not be able to handle and dis- cussed benefits of big data technologies, small
integrate this type of heterogeneous, temporal, investors too have equal opportunity to access
unstructured and dynamic data stored at different excess relevant data. This will increase participation,
places, so a new architecture is required for the participation will increase more investments,
integration of the heterogeneous database. Data will investments will increase more stable market with
not be available at one place or one database and it market hypothesis, more realistic market hypothesis
is in the different format too. It requires an will lead to more accurate prediction of the market
integrated mechanism which provides all temporal and finally it leads to stable and mature market.
data at one place and easy to process for the
decision-making process in desired format. Cloud based big data engine is used to integrate
different heterogeneous and unstructured data. Cloud
based technology will provide all analytical services
to the clients at reasonable costs. Distributed file
systems like HDFS (Hadoop Distributed File System)
and parallel processing in the cloud will increase
efficiency in the processing of data with quick
response. Big data integration will affect greatly on
the growing markets like India. There is no need to
store same data repeatedly. Intermediaries and other
interested parties gain benefits by two ways. First,
Fig. 1. Big Data Analysis Pipeline Architecture they can get data without wasting time and cost as
per their own desired format. Secondly, they exploit

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


67
Nitha K. P., Dr. S. Sivakumari Int J S Res CSE & IT. January-February-2019 ; 5(1) : 63-69

the benefit of other existing assistances of Big or from Tweets for Event Prediction, ,
cloud computing like SaaS, PaaS, and IaaS for International Conference on Intelligent
forecasting the share prices. A model developed based Computing, Communication & Convergence.
on the framework of big data analytics along with the Procedia Computer Science pp: 84 89,July 2015
blend of the fundamental, Technical and Sentimental [7]. Raghavendra Kune1,Pramod Kumar
factors will surely be a promising model for the Konugurthi1, Arun Agarwal,Raghavendra Rao
accurate prediction of stock prices in Indian stock Chillarige and Rajkumar Buyya "The anatomy
market. of Big Data Computing", SOFTWARE:
PRACTICE AND EXPERIENCE,Wiley Online
IV. REFERENCES Library (wileyonlinelibrary.com). DOI:
10.1002/spe.2374,9 October 2015
[1]. V. Sandhiya, T.Revathi, A.Jayashree, A.Ramya, [8]. Dr B Charumathi and Suraj E S "An Artificial
S.Sivasankari "Stock Mar- ket Prediction on Neural Network Approach on Improving
Bigdata Using Machine Learning Algorithm", Relative Valuation Accuracy for Bank Stocks in
Inter- national Journal of Engineering Science India Asian Research Journal of Business
and Computing,Volume 7 Issue No.4,April Management Issue 2 (Vol. 3) Issn: 2321-
2017 9246,2015
[2]. Rupawari Jadhav, M. S. Wakode "Survey : [9]. Kavitha S, Raja Vadhana P, Nivi A N "BIG
Sentiment Analysis of Twitter Data for Stock DATA ANALYTICS IN FINANCIAL
Market Prediction", International Journal of MARKET", IJRET: International Journal of
Advanced Research in Computer and Research in Engineering and
Communication Engineering,Vol. 6, Issue 3, Technology,Volume: 04 Issue: 02 - Feb-2015
March 2017. [10]. Aditya Bhardwaja, Yogendra Narayanb,
[3]. Venkata Sasank Pagolu,Kamal Nayan Reddy Vanrajc, Pawana, Maitreyee Duttaa Sentiment
Challa,Ganapati Panda,Babita Majhi "Sentiment Analysis for Indian Stock Market Prediction
Analysis of Twitter Data for Predicting Stock Using Sensex and Nifty, , 4th International
Market Movements", International conference Conference on Eco-friendly Computing and
on Signal Processing, Communication, Power Communication Systems. Procedia Computer
and Embedded System (SCOPES)- 2016 Science 70 (2015) pp: 85-91
[4]. Varunesh Nichante, Prof. Sulabha Patil "A [11]. Krishna Kumar Singh, Priti Dimri, Krishna
Review: Analysis of stock market by using big Nand Rastogi "The Impli- cations of Big Data in
data analytic technology", International Journal Indian Stock Market", International Journal of
on Recent and Innovation Trends in Computer Applications, Volume 99,Issue:5,
Computing and Communication,Volume: 4 August 2014
Issue: 1,ISSN: 2321-8169,January 2016 [12]. Dr B Charumathi and Suraj E S "Comparing
[5]. Aparna Nayak, M. M. Manohara Pai and Stock Valuation Models for Indian Bank Stocks
Radhika M. PaiPrediction Mod- els for Indian International Journal of Accounting and
Stock Market, , Twelfth International Multi- Taxation, Vol. 2, No. 2, pp. 111-127,June 2014
Conference on Information Processing-2016 [13]. Carson Kai-Sang Leung,Richard Kyle
(IMCIP-2016),pp : 441-449 MacKinnon,Yang Wang "A Ma- chine
[6]. P.G Preethia,V. Umab ,Ajit kumar ,Temporal Learning Approach for Stock Price
Sentiment Analysis and Causal Rules Extraction

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


68
Nitha K. P., Dr. S. Sivakumari Int J S Res CSE & IT. January-February-2019 ; 5(1) : 63-69

Prediction", IDEAS14, Porto, Portugal,ACM


978-1-4503-2627-8/14/07,July 2014
[14]. Yoosin Kim1, Seung Ryul Jeong1, Imran Ghani
"Text Opinion Mining to Analyze News for
Stock Market Prediction", International Journal
in Advanced Soft Computing and Applications,
Volume 6,Issue:1, March 2014
[15]. Qasem A. Al-Radaideh,Adel Abu Assaf,Eman
Alnagi "Predicting Stock Prices Using Data
Mining Techniques", The International Arab
Con- ference on Information Technology
(ACIT2013)
[16]. Dr B Charumathi and Suraj E S "Value
relevance of earnings and book value for bank
stocks An empirical analysis using ohlson
valuation model Global Journal of Management
and Research, July 2013.
[17]. Phaisarn sutheebunjard "Stock exchange of
Thailand Index prediction using back
propogation networks Second international
conference on computer and network
technology, IEEE 376-381, Thailand,2010
(2010).

Cite this article as :

Nitha K. P., Dr. S. Sivakumari, "Survey on Factors


Determining the Share Price Movement in Indian
Stock Market - A Big Data Approach ", International
Journal of Scientific Research in Computer Science,
Engineering and Information Technology
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
63-69, January-February 2019. Available at doi :
https://doi.org/10.32628/CSEIT195113
Journal URL : http://ijsrcseit.com/CSEIT195113

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


69
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1838110

Enhancing Privacy Preserving Query Retrieval Using Locality


Sensitive Hashing
Archana M.S., K. Deepa
1Assistant Professor, Department of Computer Science, Nilgiri College of Arts and Science, Thaloor, The
Nilgiris, Tamil Nadu, India
2Assistant Professor, Department of Computer Science, Providence College for Women, Coonoor, Ooty, Tamil
Nadu, India

ABSTRACT

The usage of smart phones is tremendously increasing day by day. Due to this, Location Based Services (LBS)
attracted considerably and becomes more popular and vital in the area of mobile applications. On the other
hand, the usage of LBS leads to potential threat to user’s location privacy. In this paper, the famous LBS provide
information about points of interest (POI) in spatial range query within a given distance. For that, a more
efficient and an enhanced privacy-preserving query solution for location based, Efficient Privacy-Location
Query (EPLQ) is proposed along with Locality Sensitive Hashing (LSH) reduces the dimensionality of high
dimensional data. Experiments are conducted extensively and the results show the efficiency of the proposed
algorithm EPLQ in privacy preserving over outsourced encrypted data in spatial range queries. The proposed
method performs in spatial range queries and similarity queries of privacy preserving.
Keywords: Location Based Services, points of interest, Efficient Privacy-Location Query, NICT

I. INTRODUCTION combination of various technologies such as


Geographical Information System (GIS) and other
Location Based Services (LBSs) are progressively in spatial and situating advances, the Internet and the
style in today’s society. It's reportable that up to a Web, and new data and correspondence advances
4.77 billion users are enjoying LBSs as of 2018 (NICTs).
compared to 150 million during 2014 [1]. Users can
get the required location information from LBS A framework is context-aware in the event that it
provider to get personalized services like restaurant utilizes context to give pertinent information as well
recommendations, taxi reservations and map as services to the client, where pertinence relies upon
directions. the client's task. Information or services can be
activated with context information with several
The idea of Location Based Services depends on long parameters mirroring the context of the client task.
periods of advancement and intermingling of various These parameters might be subdivided into spatial,
advances and in addition a development of personal, social, technical and physical context. The
information society where context and customization issueconnected with context- aware services is
of information is one of the principal needs for displaying those parameters in quantifiable and
clients. Considering the background of LBS, it is a computable way. The spatiotemporal information

CSEIT1838110 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 70-73 ]
70
Archana M.S., K. Deepa Int J S Res CSE & IT. January-February-2019 ; 5(1) : 70-73

about the client is one of few parameters that are III. DESIGN GOAL
clear and genuinely simple to quantify and utilize.
This is connected to the term Location-Aware It is very challenging to propose a privacy preserving
Services that may be characterized as context-aware query for implementing LBS over the encrypted
services that use the location of the client to embrace location data. Enhancing Processing of Query (EPQ)
the conduct of the services. It is a unique way of enables the cloud server to perform query processing
Location Based Services in a most broad sense as its without exposing client’s location details. EPQ can
services are connected with location data. insist on a client location coordinate query as per
client’s present location. Practically, the LBS
II. RELATED WORKS
application is mainly developed for the user who
used to search his point of interest such as searching
M. Gruteser et al investigates a complimentary
hotels nearest to him. In this work, the user query is
approach that concentrates on the principle of
secured with the keyword and encrypted. This will
minimal collection. In this approach location-based
accurately identify the requested location and the
services collect and use only de-personalized data
distance between the target locations are classified
that is practically anonymous data. But query
and ranking will be given. Our system provides a
encryption is not done and database is not outsourced.
multiuser environment as any number of users can
log at a time. The following are the three main
M. F. Mokbel introduces Casper; a novel framework
processes of our proposed work.
in which mobile users can entertain location based
services without the need to disclose their private
1. To propose a LBS query scheme which is more
location information. Per query privacy and multi-
efficient and privacy preserving to protect the
location query are absent and inefficient in handling
location data and the user’s request. For this, a
frequent attack are some of the drawbacks of this
hybrid encryption technique to encrypt the
work.
outsourced location data is proposed. This scheme
R Shokri et al proposed an approach that protects the
also supports multiuser cloud environment.
location privacy. In this approach, the mainstream
✓ System Construction Module
approach to protecting the location-privacy of mobile
✓ LBS User Module
users in location-based services (LBSs) is to alter the
✓ LBS Provider Module
users' actual locations in order to reduce the location
✓ Dimensionality Reduction Module – Locality
information exposed to the service provider. This
Sensitive Hashing
method is inefficient in terms of handling frequent
✓ Spatial Range Query For Privacy Preserving
attacks. M. E. Andres et al.

1.1 POI Query Processing


G. Ghinita et al employed the Private Information
Retrieval theory to guarantee privacy in location-
POI query processing in privacy preserving is
dependent queries. The main drawback is query
classified into two ways (i) Public LBS and (ii)
encryption is present but lacking in database
Outsourced LBS. In public LBS, POI of spatial
outsourcing.
database is recorded by an LBS provider in plaintext.
These queries are initiated at the provider’s site.
Outsourced LBS consists of three kinds of entities,
LBS provider, Cloud and LBS users.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


71
Archana M.S., K. Deepa Int J S Res CSE & IT. January-February-2019 ; 5(1) : 70-73

1) LBS data is abundant in LBS provider and these


data are nothing but POI records. Authorized users
are allowed by the LBS provider to access these data
by passing location based queries. the LBS provider
provides the query services through the cloud for
getting operational and financial benefits of data
outsourcing. But, the more valuable LBS data is not
disclosed by LBS provider to the cloud. Hence, the
LBS provider will outsource the encrypted LBS data
to the cloud.
Figure 4.1 (a) POI in Coimbatore
2) The cloud has abundant computing resources and
storage. The encrypted LBS data from the LBS
provider are stored and also provides query services
for the users. So, in cloud, the encrypted POI records
are searched in local storage for finding the data that
matches the queries of LBS users.

IV. RESULT DISCUSSION

In the previous section, we have discussed about the


modules for achieving the privacy preserving of Figure 4.2 (b) POI in Calicut
location data. These modules are implemented using
Java and MySQL. For accessing the location, we have V. CONCLUSION AND FUTURE WORK
implemented using J2ME as a mobile application. The
experiments are carried out on Intel Core(TM) I3 In this paper, the famous LBS provide information
Processor (2.27 GHz) about points of interest (POIs) in spatial range query
within a given distance is discussed. For that, a more
efficient and an enhanced privacy-preserving
The system takes input from the users, processes it
location-based query solution, Enhancing Privacy-
and produces an output. Input design is link that ties
Location Query (EPLQ) is proposed.
the information system into the world of its users.
The system should be user-friendly to gain
In specific, to obtain privacy preserving spatial range
appropriate information to the user.
query, the first predicate-only encryption scheme for
inner product range (IPRE) is proposed. This method
The project gives the low time consumption to make
is used to detect a privacy preserving way to find a
the sensitive application made simple. The fund that
position whether it is within a given circular area.
the company is willing to spend into the research and
The IPRE scheme has two vectors, attribute and
development of this system is very limited. The
predicate vectors. The problem with the proposed
design of input covers all the phases of input from the
method is, when the numbers of attributes are
creation of initial data to actual entering of the data
increased, the latency time for retrieving data is also
to the system for processing.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


72
Archana M.S., K. Deepa Int J S Res CSE & IT. January-February-2019 ; 5(1) : 70-73

increased. Here, we apply locality sensitive hashing location based queries,” IEEE Transactions on
to reduce dimensionality thereby supporting k- Knowledge and Data Engineering, vol. 26,no.5,
nearest neighbor search. pp. 1200-1210, 2014.
[8]. L. Sweeney, "Achieving k-anonymity privacy
VI. REFERENCES protection using generalization and
suppression,” International Journal of
[1]. Statista, "Number of location-based service Uncertainty, Fuzziness and Knowledge-Based
users in the United States from2013 to 2018 (in Systems, vol. 10,no. 5, pp. 571-588, 2002.
millions),” Statista; 2017.
https://www.statista.com/statistics/436071/locat Cite this article as :
ion-based- service-users-usa
[2]. K. Xie, X. Ning, X.Wang et al., "Recover Archana M.S, K.Deepa , "Enhancing Privacy
corrupted data in sensornetworks: a matrix Preserving Query Retrieval Using Locality Sensitive
completion solution,” IEEE Transactions on Hashing", International Journal of Scientific Research
Mobile Computing, vol. 16, no. 5, pp. 1434- in Computer Science, Engineering and Information
1448, 2017. Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
[3]. M. Li, H. Zhu, Z. Gao et al., "All your Location 5 Issue 1, pp. 70-73 , January-February 2019.
are belong to us: breaking mobile social Available at doi :
networks for automated user location tracking,” https://doi.org/10.32628/CSEIT1838110
in Proceedings of the 15th ACM International Journal URL : http://ijsrcseit.com/CSEIT1838110
Symposium on Mobile Ad Hoc Networking and
Computing, MobiHoc 2014, pp. 43-52, USA,
August 2014.
[4]. J. Shao, R. Lu, and X. Lin, "FINE: a fine-grained
privacy preserving location-based
serviceframework for mobile devices,” in
Proceedings of the IEEE INFOCOM, pp. 244-
252, IEEE, Ontario, Canada, May 2014.
[5]. G. Ghinita, P. Kalnis, A. Khoshgozaran, C.
Shahabi, and K.-Tan, "Private queries in
location based services: anonymizers are not
necessary,” in Proceedings of the
ACMSIGMOD International Conference on
Management of Data (SIGMOD ’08), pp. 121-
132, ACM, 2008.
[6]. B. Gedik and L. Liu, "Protecting location
privacy with personalized k-anonymity:
architecture and algorithms,” IEEE
Transactions on Mobile Computing, vol. 7, no.
1, pp. 1- 18, 2008.
[7]. R. Paulet, M. G. Kaosar, X. Yi, and E. Bertino,
"Privacy preserving and content-protecting

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


73
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1838111

Enhanced Security for Dynamic Multi Keyword Ranked Search


Using Greedy Best First Search
Bibin Baby1, Sharmila Banu2
1
Assistant Professor, Department of Computer Science, Nilgiri College of Arts and Science, Thaloor,
The Nilgris, Tamil Nadu, India
2
Assistant Professor, Department of Computer Science, Providence College for Women, Coonor,
The Nilgiris, Tamil Nadu, India

ABSTRACT

Today, due to the enormous growth of data technology in cloud computing, the data owners are stimulated to
outsource their data in data management to reduce cost and for the convenient. Data confidentiality, in general,
can be obtained by encrypting the data before it is outsourced. The client stores the encrypted data to the cloud
using Searchable encryption schemes and applies keyword search techniques over cipher text domain. But the
main problem in outsourcing is the lack of security and privacy for the sensitive data. So, to overcome this, for
privacy requirement, the sensitive data can be encrypted before it is outsourced. Various methods were
proposed to preserve the privacy and to provide security to the cloud data which are encrypted. Here in this
paper, we proposed a tree-based search method over the encrypted datain the cloud that supports dynamic
operation and multi-keyword ranked search. Clearly, the commonly used “inverse document frequency (IDF)
term frequency (TF)” model and the vector space method are joined in the query generation and index creation
to give multi-keyword ranked search. To get high search efficiency, a tree-type index structure, “Greedy Best-
first Search” algorithm is proposed based on the tree- index.
Keywords : Cloud Computing, Security, Multi Keyword Search, Greedy Best First Search, Inverse Document
Frequency, Term Frequency

I. INTRODUCTION themselves. No matter the various blessings of cloud


services, outsourcing touchy information (which
Nowadays, cloud computing has been taken into includes e-mails, non-public health statistics,
consideration as a new edition of IT infrastructure, enterprise business enterprise finance records,
that may put together large beneficial useful resource authority’s files, and plenty of others) to far flung
of computing, storage and programs, and allow users servers brings privacy worries. The Cloud Carrier
to revel in ubiquitous, convenient and on-demand Vendors (CSVS) that preserve the facts for users may
network get admission to a shared pool of also additionally get admission to users’ sensitive
configurable computing sources with splendid overall information without authorization. A fashionable
performance and minimal monetary overhead technique to shield the information confidentiality is
.Attracted by way of those appealing features, each to encrypt the statistics earlier than outsourcing but,
individuals and firms are stimulated to outsource this will motive a big rate in phrases of records
their facts to the cloud, rather of buying software usability. For an example, the present techniques on
program and hardware to manipulate the facts keyword-based totally information retrieval, which

CSEIT1838111 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 74-78 ]
74
Bibin Baby, Sharmila Banu Int J S Res CSE & IT. January-February-2019 ; 5(1) : 74-78

are widely used on the plaintext records, cannot be vector space model supporting solidified and
immediately implemented at the encrypted records. particular document look. The searchable file is
Downloading all of the statistics from the cloud and fabricated utilizing Multidimensional B tree. Its
decrypt regionally is impractical. On the opposite, limitations are the likeness rank score of the record
greater sensible unique motive solutions, including vector completely relies on upon the kind of the
searchable encryption (SE) schemes have made report Jiadi [7] proposed this pursuit utilizing two
unique contributions in phrases of performance, round searchable encryption (TRSE). The withdrawal
capability and safety. and restricting is utilized to diminish figure content
size, still the key size is too extensive. The
II. RELATED WORKS correspondence elevated will be high, if the
scrambled trapdoor's size is too vast. It doesn't make
This section discusses the literature survey of various powerful searchable file redesign. These are some of
papers related to Multi-keyword ranked search over the drawbacks of this method.
encrypted cloud data in cloud. It describes the
research work already done by the authors III. PROPOSED METHODOLOGY
corresponding to searchable encryption in cloud
computing. In this section, the Dynamic Multikeyword Ranked
Search for Plaintext (DMRSP)method is explained for
Qin Liu [3] proposed the pursuit that gives catch implementing our proposed method. Based on the
phrase protection, information protection and vector space model and Best First tree based indexing,
semantic secure by open key encryption. The main DMRSP is constructed. Apart from this, a secure
drawback of this system is the correspondence and search scheme, Simple DMRS (SDMRS) is also
computational expense for encryption and decoding explained.
is more. Cong Wang [4] proposed seek which
unravels preparing overhead, information and 3.1 Index Construction using Best First Search
catchphrase security, least correspondence and .
calculation aeronautical. The information proprietor
The search algorithm, Best-first search explores
assemble list alongside the catchphrase recurrence
a graph by increasing the most likely node chosen as
based importance scores for documents. Butit doesn't
per the specified rule.
play out numerous catch phrase seeks and Minimal
overhead in record building. C Wang et al [5] Judea Pearl explains best-first search as calculating
proposed that searchable encryption idea agree to the the promise of node ‘n’ by a "heuristic evaluation
client to solidly search for over scrambled function. It may depend on the depiction of n, the
information through watchwords without first goal and the information gathered. It also collects
applying decoding on it. But this system is too slow in additional knowledge about the problem domain.
processing the request and is not suitable for huge A priority queue is the efficient way of selecting the
volume of data. best candidate.

The first successor of the parent is expanded using


Wenhai Sun [6] proposed an inquiry that gives
greedy algorithm. Once a successor is generated,
comparability based item positioning, catchphrase
security, Index and Query classification and Query
Unlink capacity. The scrambled record is worked by

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


75
Bibin Baby, Sharmila Banu Int J S Res CSE & IT. January-February-2019 ; 5(1) : 74-78

1. If the heuristic of successor is best, then the For achieving this, we need to visit all nodes before
successor is put at the front of the queue and we reach the destination. Since priority queue uses
restarts the loops. Max or Min heap, insert and delete operations takes
O(log n).So the performance of the algorithm highly
2. or else, insert the successor into the queue
depends on evaluation function or the cost is
designed.
The algorithm for Best first search is explained below
for the following fig. 3.1 IV. RESULT DISCUSSION

These modules are implemented using Java and


MySQL. For accessing the location, we have
implemented using J2ME as a mobile application. The
experiments are carried out on Intel Core(TM) I3
Processor (2.27 GHz).

4.1 Precision
In the proposed method, dummy keywords affects
Fig. 3.1 Best First Search Tree the search precision. Search Precision can be defined
1. We start from source S and search for destination as
Ith node using Best First search with the given cost.
2. P and Q initially contains S Pkd=kd’= k wherekd’ = top k- documents
3. Remove S and process unvisited neighbors of S to To obtain higher precision, the very small standard
pq deviation σ is set as the random variable for ∑ 𝜀𝑣
4. Now pq contains {A, C, B} (C is put before B
because C has lesser cost) Table 4.1 Storage Consumption of Index Tree
5. Remove A from pq and process unvisited neighbors
of A to pq.
6. Now pq now contains {C, B, E, D} Size of the
1000 2000 3000 4000 5000
dictionary
7. Remove C from pq and process unvisited neighbors
of C to pq. GDFS (MB) 78 151 230 335 400
8. Now pq contains {B, H, E, D}
BFS (MB) 72 137 210 297 380
9. Remove B from pq and process unvisited neighbors
of B to pq.
10. Now pq contains {H, E, D, F, G}
From the above table it is clearly noted that, the
11. Remove H from pq.
memory occupied by Best first search is very less
12 Now return I
compared to the greedy depth first search.

Worst case time complexity : Best First Search : O(n


* Log n)
- n is number of nodes.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


76
Bibin Baby, Sharmila Banu Int J S Res CSE & IT. January-February-2019 ; 5(1) : 74-78

Table 4.2 Precision Percentages achieve greater search efficiency by implementing


our “Best first Search” algorithm.Even though our
proposed method works efficiently in terms of multi-
NO Precision NO Precision
1 89% 9 90% keyword search, still there are some serious issues
2 95% 10 96% which we will do in the future work. For instance,
3 96% 11 96% every time data owner generates updating
4 97% 12 87% information and this will be sent to the cloud server.
5 100% 13 100% So he has to store the plaintext tree index and IDF
6 86% 14 100%
values for recalculation.
7 88% 15 95%
8 90% 16 87%
VI. REFERENCES

We have randomly chosen 5 keywords as input for


each test and the precision is noted for the top 100 [1]. Q. Liu, G. Wang, and J. Wu, "Secure and

results. This test is repeated 16 times and the average privacy preserving keyword searching for cloud
precision obtained is 93%. The below table gives the storage services,' Journal of Networks and
precision percentage obtained for multi-keyword Computer Applications, vol. 35, no. 3, pp. 927–

search on searching the documents. 933, May 2012


[2]. C. Wang, N. Cao, K. Ren and W. Lou, -

V. CONCLUSION AND FUTURE WORK Enabling secure and efficient ranked keyword
search over outsourced cloud data, IEEE
A tree-based search method is proposed over the Transactions on Parallel and Distributed
encrypted data in the cloud that supports dynamic Systems, vol. 23, issue 8, (2012) August, pp.

operation and multi-keyword ranked search on the 1467–1479.


document collection. Precisely, the widely-used [3]. C. Wang, N. Cao, J. Li, K. Ren, and W. Lou,
“term frequency (TF) ×inverse document frequency "Secure ranked keyword search over encrypted

(IDF)” model and the vector space model are cloud data,' in Proc. of ICDCS'10, 2010.
combined in the index creation and query generation [4]. W. Sun, B. Wang, N. Cao, M. Li, W. Lou, YT,

to afford multi-keyword ranked search. In order to Hou and H. L, -Privacy-preserving multi-


get high search efficiency, a tree-based index keyword text search in the cloud supporting

structure is constructed and for that a “Best First similarity-based ranking, In Proceedings of the

Search” algorithm is proposed based on this tree 8th ACM SIGSAC symposium on Information,
index. The secure kNN algorithm is utilized to computer and communications security, (2013),

encrypt the index and query vectors, and meanwhile pp. 71-82.

ensure accurate relevance score calculation between [5]. N. Cao, C. Wang, M. Li, K. Ren, and W. Lou,
encrypted index and query vectors. From the results, "Privacy-Preserving Multi-Keyword Ranked

it is clearly states that the special structure, tree- Search over Encrypted Cloud Data,' in IEEE

based index can easily achieve sub-linear search time Transactions on Parallel and Distributed
and compact with the insertion and deletion of Systems, vol. 25, no. 1, pp. 222–233, 2014.

documents. Due to the tree-based index, the search [6]. Yu, Jiadi& Lu, Peng& Zhu, Yanmin&Xue,
complexity of the proposed work is basically kept to Guangtao& Li, Minglu. (2013). Towards Secure
logarithmic. The proposed scheme, in practice, can Multi-Keyword Top-k Retrieval over
Encrypted Cloud Data. IEEE Transactions on

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


77
Bibin Baby, Sharmila Banu Int J S Res CSE & IT. January-February-2019 ; 5(1) : 74-78

Dependable and Secure Computing. Cite this article as :


10.1109/TDSC.2013.9.
[7]. W. Sun, B. Wang, N. Cao, M. Li, W. Lou, YT Bibin Baby, Sharmila Banu, "Enhanced Security for
,Hou and H. L, -Privacy-preserving multi- Dynamic Multi Keyword Ranked Search Using
keyword text search in the cloud supporting Greedy Best First Search", International Journal of
similarity-based ranking, In Proceedings of the Scientific Research in Computer Science, Engineering
8th ACM SIGSAC symposium on Information, and Information Technology (IJSRCSEIT), ISSN :
computer and communications security, (2013), 2456-3307, Volume 5 Issue 1, pp. 74-78, January-
pp. 71-82. February 2019.
[8]. Cengizorencik , ErkaySavaş, An efficient Available at doi :
privacy-preserving multi-keyword search over https://doi.org/10.32628/CSEIT1838111
encrypted cloud data with ranking, Distributed Journal URL : http://ijsrcseit.com/CSEIT1838111
and Parallel Databases, v.32 n.1, p.119-160,
March 2014
[9]. Chen, Lanxiang, et al. "DMRS: an efficient
dynamic multi-keyword ranked search over
encrypted cloud data." Soft Computing 21.16
(2017): 4829-4841
[10]. C Orencik, M Kantarcioglu, E SavasA practical
and secure multi-keyword search method over
encrypted cloud data 2013 IEEE Sixth
International Conference on Cloud Computing,
390-397
[11]. Peng, Tao, Qin Liu, and Guojun Wang. "a
multilevel access Control scheme for Data
security in Transparent Computing."
Computing in Science & Engineering (2016).
[12]. Shen, Jian, et al. "An authorized identity
authentication-based data access control
scheme in cloud." Advanced Communication
Technology (ICACT), 2016 18th International
Conference on. IEEE, 2016.
[13]. Shen, Zhirong, JiwuShu, and Wei Xue.
"Keyword Search With Access Control Over
Encrypted Cloud Data." IEEE Sensors Journal
17.3 (2017): 858-868

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


78
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19519

Cyber Security, Challenges, Some Practical Solutions


Y. V. Sai Bharadwaj, 2Sai Bhageerath Y.V, 3Prof. Y.V.S.S.S.V. Prasada Rao
1

1Central University of Hyderabad, Hyderabad, Telangana, India


2 National Institute of Technology, Warangal, Hanamkonda, Telangana, India
3Pattabhipuram, Guntur. Andhra Pradesh, India

ABSTRACT

Cybercrime continues to surge without a slowdown in sight. The cyber security threat continues to worsen. In
the first half of 2018, the number of cyber breaches soared over 140% from a year earlier, leading to 33 billion
compromised data records worldwide. Cyber Security news such as Marriott hack in Nov 2018 is dominating
headlines and becoming a serious headache for business leaders. Malicious outsiders sparked more than half of
the 944 breaches and accounted for roughly 80% of stolen, compromised or lost records. Identity theft
continues to lead data breach types, but financial access incidents are escalating in severity as well. The United
States continues to be the favorite target, and data breaches at major US enterprises continue to grab the
headlines. In 2018, the most notable breaches have occurred at Adidas, FedEx, Jason’s Deli, Macy’s, Under
Armour, Nordstrom’s and the most popular Facebook. [1].

Keywords : Cyber Security, Quantum Computing, Crypto-Agility, Phishing, Cyber Insurance, Phishing attack,
JDLR, Emotional Blackmail, CCO, TalkTalk Data Breach

I. INTRODUCTION breaches have occurred at Adidas, FedEx, Jason’s Deli,


Macy’s, Under Armour, Nordstrom’s and the most
Cybercrime continues to surge without a slowdown popular Facebook. [1].
in sight. The cyber security threat continues to
worsen. In the first half of 2018, the number of cyber II. Cyber security in small and medium sized
breaches soared over 140% from a year earlier, companies
leading to 33 billion compromised data records 2.1. Issue :
worldwide. Cyber Security news such as Marriott Cyber security in small and medium-sized companies
hack in Nov 2018 is dominating headlines and is a big issue, but businesses can take steps to protect
becoming a serious headache for business leaders. themselves. The issue of cyber security in small and
Malicious outsiders sparked more than half of the 944 medium-sized companies is becoming vital. [2]
breaches and accounted for roughly 80% of stolen,
compromised or lost records. Identity theft continues As for cyber security in small and medium-sized
to lead data breach types, but financial access companies, many are realising that they are viewed as
incidents are escalating in severity as well. The attractive a target as the larger companies. Cisco’s
United States continues to be the favorite target, and 2018 SMB Cyber Security Report found that 53% of
data breaches at major US enterprises continue to mid-market companies in 26 countries experienced a
grab the headlines. In 2018, the most notable breach. For these companies, the top security

CSEIT19519 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 79-87 ]
79
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

concerns are targeted phishing attacks against 2. Ensure you have a proper backup system.
employees, advanced persistent threats, ransom ware, And make sure it is easy to access in case you
denial-of-service attacks and the proliferation of need to restore one piece of the system rather
employees allowed to use their own mobile devices. than the entire system. Enterprise-level cloud
systems can help.
Malware of all types is a huge problem. It is 3. Examine all the entry points into your system
becoming more difficult to combat as cyber-attackers and consider where they are vulnerable.
get more adept at developing software that can evade These include all your workstations,
traditional detection and employ more sophisticated communications and mobile devices as well
malware. For small and medium businesses, one as employee access cards, the internet and
breach often puts a victim out of business. That’s cameras.
because 54% of all cyber-attacks cause financial 4. Assess your system threats. These include
damages exceeding $500,000, the 2018 Cisco SMB client lists, passwords, data logs, backups and
cyber security report shows. That price tag along emails, and anyone who specifically has
with a damaged reputation are hard to survive. If access to the system, including customers and
they do survive, they still face significant system vendors.
downtime that averaged eight hours or more in the 5. Put a prevention system in place to defend
last year. Further, such companies often lack the IT against intruders. Put yourself in the place of
talent, budget and technologies to prevent, uncover the cyber attacker and consider the possible
and respond to an attack. ways the attacker could access your system
and steal your data. If your internal IT staff
2.2. Prevention of Cyber Theft : isn’t experienced enough to handle, entrust a
third-party firm, because the prevention
Unsurprisingly, there is no easy solution – and none
system must cover physical and digital
is likely within the near future – to prevent data
security.
breaches. But all businesses, especially small and
medium sized companies can become better prepared III. Quantum Cryptography: The next-generation of
and more adept at protecting against cyber-crime. secure data transmission
3.1. Issue:
Here are five actions concerning cyber security in
Quantum Computing will render much of today's
small and medium-sized companies that can be taken
encryption unsafe. But Quantum Cryptography could
to become more security-conscious:
be the solution. China has recently successfully tested
– for the first time – its quantum communications
1. Conduct a security audit. Learn how secure
system, which cannot be hacked. [3].
your network and other security systems are,
where vulnerabilities exist and how to
3.2. How does quantum cryptography work ?.
resolve them. If you consider cybersecurity
insurance — currently the fastest-growing
Quantum cryptography provides a secure method for
insurance — or have coverage from a
generating and distributing secret keys between two
business insurer, the insurer can usually refer
parties via an optical network. This is achieved by
you to resources to assist in the audit.
capitalising on the inherent unpredictability in the
state of particles – such as photons or electron – to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


80
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

generate the random numbers needed for be resolved at the click of a button. Crypto-agility has
cryptographic applications. The technology then never been more important to ensure businesses can
harnesses this to create and share a secret digital key confidently protect themselves and their customers
that can be used to encrypt or authenticate from hackers. This is why organisations must invest
information via streams of encoded single photons, in a credible technology to automate the tracking of
which are sent through an optical communication certificates; it is no longer feasible to do this
network. manually, there are simply too many certificates to
track. Companies focusing too much on protecting
3.3. The arrival of quantum cryptography : usernames and passwords and not enough on
machine identities.
The arrival of quantum computing will ultimately
render much of today’s encryption unsafe. The 4.3. A future-proof solution :
current consensus that public key encryption is an
Google’s decision to distrust Symantec certificates
essential part of data security is starting to be
doesn’t need to be the end of the world. If companies
questioned in the face of new attack strategies –
are able to manage all their machine identities
which is subsequently driving uptake of robust
centrally and automate the process, it will enable
quantum cryptography solutions and services to
crypto-agility and ensure they can migrate quickly
deliver better data security. As a result, the global
when a flaw or vulnerability is discovered. By doing
quantum cryptography market is forecast to grow
so, companies can insulate themselves against the
from USD 285.7 Million in 2017 to USD 943.7
volatility of the CA market, protect their reputation
Million by 2022, a Compound Annual Growth Rate and ensure business continuity for online services.
(CAGR) of 27 per cent according to Research and
Markets. In reality, however, as things stand today V. What is phishing ?
quantum cryptography is not quite so close to
mainstream fruition.
The word ‘phishing’ was invented as a homophone of
IV. Crypto-agility: the key to ensuring long term ‘fishing’ as it involves creating a bait to lure
website security victims. Typically it involves an email; although
4.1. Issue : sometimes a telephone call – called Vishing – or a
In the age of the internet, you simply cannot afford text – called Smishing. These emails are often
to do business without being online, and keeping credible enough to deceive the recipient into clicking
your website secure should be a top priority [4]. on a link which could then release malware – viruses,
worms, Trojans or bots – onto the recipient’s
4.2. Crypto-agility to the rescue : computer or take the victim to a fake website.
This is why organisations need crypto-agility – i.e.
the ability to manage machine identities in real-time. 5.1. How to protect against phishing :
Crypto-agility enables businesses to quickly identify
The following are the tips against phishing.
and replace certificates in bulk when security events
or business needs call for it. Currently, many
organisations take days or even weeks to find and
replace certificates, which isn’t conducive to ensuring
5.1.1. Never, ever follow suspect links
security. By automating the process, this process can

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


81
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

There is no 100% guaranteed way to detect phishing If attachments with unknown file extensions (or PDF
but, if there is the slightest suspicion that the email files) suddenly appear as an e-mail attachment, it is a
may be fraudulent, do not click on any links it clear indication that something is wrong – especially
contains. Always enter the sender’s website address if you haven’t had any previous dealings with the
(not the link in the email) directly into your browser. sender.

5.1.2. Check out the sender 5.1.6. Personal salutation

Be warned if the part after the ‘at’ sign @ in an email Most companies address their customers by name.
address doesn’t match the purported sender; for But if the name is missing, misspelt or if there is no
example, if ‘PayPal’ sends you an email name at all and it just says something like ‘Hey’ or
from paypal@emails.com or the URL is misspelled ‘Dear Customer’, it could be an indication that this is
as www.paypa1.com or something similar. This is a a fake email.
(fake) website owned by a cyber squatter. Some of
the most well-known companies in the world have 5.1.7. Trust is good but control is better
website impersonators including Facebook, Google,
DropBox and PayPal. By regularly checking your bank statements, you can
mitigate any potentially serious consequences of a
5.1.3. Don’t give in to emotional blackmail phishing attack. Any suspicious or unknown
transactions should be reported directly to the bank
Phishing mails almost always contain the same kind or credit card company immediately.
of content and requests. Sometimes, they ask you to
update your user account or password. But 5.1.8. Keep yourself up-to-date on current scams
sometimes they use psychology to get you to react:
the notification of a big lottery win, an offer to take Take the time to read up regularly on ways to protect
part in a once-in-a-lifetime business opportunity or, your digital safety. If you hear that a service provider
an appeal for a donation to a charity. has been hacked, be sure to follow their instructions
and change your password.
5.1.4. Banks never want to know this
5.1.9. Only use secure websites
There are some things that your bank will never ever
ask you. They don’t want your passwords or PINs to When conducting online transactions, go directly to
be sent by e-mail or text; they don’t want you to the website. If the special offer is genuine, it will be
authorise the transfer of funds to a new account; and available on the website. Look for a sign that the site
they don’t want you to meet a bank representative at is secure, such as a white padlock icon on the
your home to collect cash, bank cards or anything browser’s status bar or a “https” URL (where the “s”
else. stands for “secure”).

5.1.10. Protect your computer with a firewall, spam


filters, anti-virus and anti-spyware software.

Do some research to ensure you are getting the most


5.1.5. Beware of opening attachments up-to-date software, and update them all regularly to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


82
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

ensure that you are blocking new viruses and The following are the cyber security trends and issues
spyware. (more stringent regulation, creations of new roles
etc.,) the world can expect in 2019: [ 6 ].
5.1.11. Click in haste, repent at leisure
6.1. Cyber security regulations improvement
Many phishing e-mails put pressure on you to act
quickly or else, they threaten, something bad will
happen or you will miss out on something very
We need to see a continuing improvement in the
important. A ‘bank‘ might warn you that your
relevant regulations as apply to cyber security. The
account will be closed unless you act quickly; or a
dynamic and fast-moving nature of cyber security
company might tell you that you have won a major
outpaces regulation which is far too slow and clumsy
cash prize, but only if you can claim it in the next 24
to be of any benefit and might actually hinder
hours. Don’t act in haste. Take your time to satisfy
security by building a culture of compliance with
yourself that the message is genuine.
regulations and a false sense of security against
enemies who are agile, motivated, and clever.
5.1.12. Genuine messages don’t make threats

6.2. Data theft turning into data manipulation


Although most phishing scams involve trying to trick
or persuade people into handing over sensitive We can expect to see attackers changing their
information, some fraudsters use fear and methodology from pure data theft and website
intimidation to scare their victims. For example, hacking to attacking data integrity itself. This type of

threatening to send embarrassing videos or photos to attack, in comparison to a straightforward theft of


contacts unless a ransom is paid. Try not to react data, will serve to cause long-term, reputational
immediately to an email take a few minutes to calm damage to individuals or groups by getting people to
down and think rationally. Why would this person question the integrity of the data in question.
be emailing you, specifically about this, all of a
6.3. Demand will continue to rise for security skills
sudden?.
A global shortage of cyber security skills in the
5.2. The JDLR rule : workplace arguably makes organisations more
desirable targets for hacking. Demand for expertise
Criminals are now able to put together professional- will rise as companies realise that their current IS
looking messages and web pages which can trick strategy is not sufficient. Also, with companies
even the most discerning person into giving away increasingly insourcing their security needs, internal
personal information when they are tired, busy or training and skills growth has to continue to
stressed. Check your privacy settings on popular accelerate. Tailored training programmes are crucial.
social networks to restrict how much personal
information you are making public and above all, 6.4. Cyber security and Internet of Things (IoT)
follow the JDLR rule. If it ‘Just Doesn’t Look Right‘, ‘Secure by design’ will garner much copy, but
then it probably isn’t. [ 5 ]. probably will not deliver until 2019 or beyond. We’ll
have to wait and see with this, as connected devices
are increasing in circulation by the day, and perhaps
it is only a matter of time before the security
VI. Cyber security trends in 2019

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


83
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

vulnerabilities are exposed — could there be a repeat the Deep Web, in order to successfully hide and to
of the Mirai Botnet in 2019?. Indeed, the next communicate with other criminals.
generation of AI-powered attacks will be crafty
enough to emulate the behaviours of specific users to 6.8. Breaches will get more complicated and harder to
fool even skilled security personnel.This might beat
include the ability to craft complex and bespoke Cybercriminals will look to grow their malicious
phishing campaigns that will successfully fool even activities using malicious code in ever more devious
the most threat-conscious among us. ways.Such a ransomware variant has already been
discovered using an innovative system to increase
6.5. Attackers will continue to target consumer infections: the software turns victims into attackers
devices by offering a pyramid scheme-style discount.If the
Ransomware is a recognised problem for companies victim passes on a link to the malware and two or
of all shapes and sizes, epitomised by the large scale more people install this file and pay, the original
WannaCry attack that decimated the UK’s National victim has their files decrypted for free.
Health Service (NHS) and organisations around the
world. In 2019 and beyond, will we start to see 6.9. Cyber risk insurance will become more common
consumers being targeted across a range of connected This type of insurance will increasingly become part
objects? This is a likely scenario, with examples of operational risk strategy however, the insurance
coming out of child predators targeting IoT devices in industry needs to tailor products specific to client
toys (designed for children). Attackers might even needs and not just provide blanket cover as
target the smart TV in your house via a ransomware extensions to existing risks. As the industry evolves
attack that would require you to pay a fee to unlock we might see cyber insurance covering for loss of
it. reputation and trust with their customers, loss of
future revenue from negative media or other
6.6. Attackers will become bolder, more commercial exposure, and improvement costs for security
less traceable infrastructure or system upgrades.

6.10. New job titles appearing – CCO (Chief


Hackers will look to become more organised and Cybercrime Officer)
more commercialised, perhaps even having their own In the aftermath of the TalkTalk data breach, MPs
call centres – something already seen with fraudulent recommended appointing an officer with day-to-day
dating sites. They will look to base themselves in responsibility for protecting computer systems from
countries where cybercrime is barely regarded as a attack.
crime and thereby placing themselves outside their
victims’ police jurisdictions. 6.11. Will 2019 see organisations looking to appoint a
chief cybercrime officer?
6.7. Attackers will get smarter
The CCO would be responsible for ensuring that an
Attackers capability to write bespoke targeted code organisation is cyber-ready, would bear the
will continue to improve faster than the defenders responsibility for preventing breaches, would take
ability to counter or get ahead of it.They will the lead if a breach did occur and provide a robust
continue to exploit the Dark Web, a small portion of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


84
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

connection between the board and the rest of the 7.3. Raise awareness, stay vigilant :
company. Awareness should always underpin your prevention
plan. SMEs are often targeted as the way in for
financial attacks through phishing or impersonation
with the aim to extract financial data or currency.
VII. Cyber security prevention plan :
This is often due to it being easier to take advantage
of human behaviour to target specific individuals
While there is no simple, singular solution to combat
who may not have had the right training. Whether
hack attacks and deter criminals, a layered
it’s lack of awareness or just lazy decisions, they need
prevention plan is crucial. There are five main steps
to be properly trained on the risks and repercussions,
that a company should take to prevent serious
along with potential hacking tricks as the majority of
damage from a cyber attack. They are as follows. [ 7 ]
malware still requires a human action to initiate it.
7.1. Assess your current cyber hygiene :

7.4. Layered prevention approach :


The UK Government backs an innovative scheme
called Cyber Essentials which is designed to help all There are business-grade security solutions that can

companies improve their cyber hygiene. It’s an help protect the network and users from would be

essential requirement for any business bidding on attacks such as anti-virus, anti-spam and business
government contracts, but it’s beneficial for any grade firewalls. Businesses should always invest in a
company, as it’s believed to reduce the risk of an layered defence strategy, as the more comprehensive
attack by up to 80%. It results in a certification which the set up is, the less chance an attack will succeed as
demonstrates a commitment to protecting business it has to pass through the various layers. Avoid free
and stakeholder data from threats – crucial for products and solutions which claim that they can
building customer trust. The scheme covers all keep malware off your PC. Windows Defender, for
essential controls such as firewalls, malware instance, doesn’t stop adware or Potentially
prevention and up-to-date software. Unwanted Programs (PUP) and doesn’t possess the
accuracy and effectiveness that more sophisticated
7. 2. Good housekeeping : prevention tools do. Paid-for prevention tools are a
small price to pay for the reassurance of digital safety.
There’s no replacement for the basics. Updates and
The cost of a hack to a business would be much more
patching should be performed regularly —
than the cost of the prevention technology.
WannaCry, the ransomware that caused chaos across
the world, exploited unpatched Windows systems to
7.5. Better safe than sorry :
spread malware. Businesses should steer away from
legacy systems like Windows XP, as these no longer While it’s important to be proactive, businesses need
receive updates and are especially vulnerable to a recovery plan in place in the event of a disaster or
attack. Get in the habit of conducting regular routine any downtime. A staggering 60% of businesses that
maintenance and audits, and seek a service provider encounter an attack go out of business in their first

if necessary, which can use specialist software to year, because of attacks to their network and users.
block ransomware strains. While the big dogs like the NHS, Sony and Equifax
dominate the headlines when they suffer attacks, it’s
the Small & Medium Enterprises (SMEs) that are
secretly suffering the worst due to factors such as the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


85
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

human element and not prioritising investments, as claim while operations are down and sales are halted.
mentioned above. There could be consequential damages if this causes
supply chain issues with its partners who are not
Under General Data Protection Regulation (GDPR) receiving necessary component parts, thereby leading
legislation, you must have a plan in place to be able to third party claims. That's why cyber insurance can
to restore data, whether it’s a cyber attack, file be a smart precaution for any size business.
corruption or simple data loss, otherwise you risk
non-compliance. IT continuity is the bread and 8.1. What is cyber insurance?
butter of so many businesses, and you must make
Cyber insurance generally covers your business'
sure that you have a backup and business continuity
liability for a data breach involving sensitive
plan to prepare, should the worst happen. Seek
customer information, such as Social Security
Disaster Recovery as a Service (DRaaS) to protect
numbers, credit card numbers, account numbers,
critical business data and get operational again after a
health records etc., General liability insurance covers
disaster.
bodily injuries and property damage resulting from
your products, services or operations. Cyber
While there is not a simple, singular solution to
insurance is often excluded from a general liability
combat hack attacks and deter criminals, a layered
policy.
prevention plan is crucial. Follow these steps, and
data misuse and abuse can be minimised, while still
8.2. Cyber insurance coverage :
enabling your business to take advantage of the
The cyber insurance covers the following.
growing opportunities the internet can harness.

VIII. Cyber Insurance • Legal fee and expenses


• Notifying customers about a data breach
Unfortunately, data breaches and other cyber crimes • Restoring personal identities of affected
are becoming way too common. In the past couple of customers
years, data breaches have resulted in major fines and • Recovering compromised data
legal fees – not to mention headaches – for a discount • Repairing damaged computer systems
retail chain, one of the nation’s largest banks, a well-
known health insurer, an entertainment network and
IX. CONCLUSION
the government. But it’s not just large organizations
that are susceptible to being hacked or getting a virus.
Throughout the year 2018, many cyber breaches
Nearly 55% of small businesses have experienced a
were reported. With a different issue almost every
data breach and that 53% have had multiple
week throughout the year, cyber security is at the top
breaches?. A data breach can damage more than just
of bysiness owners’ minds. Also, this is not something
your small-business computer system – it also can
that experts foresee ending any time soon. As such
damage your reputation and put your customers
there are some of the things you shall have to pay
and/or employees at risk. A bad actor deploys ransom
attention in the year 2019.Latrobe University claims
ware into a manufacturing company’s network
that it takes a combination of education and
environment which freezes up operations and delete
awareness to prevent cyber attacks from occurring.
files until it receives ransom payment. This could
Standard defense practices are only as strong as their
lead to direct revenue loss in a business interruption
weakest link, which is why it’s so important to train

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


86
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

your employees regarding phishing and spam emails. Cite this article as :
These act as an entryway for malware, so one should:
Y. V. Sai Bharadwaj, Sai Bhageerath Y.V, Prof.
• Use reliable, well-known security and antivirus Y.V.S.S.S.V. Prasada Rao, "Cyber Security,
software Challenges, Some Practical Solutions", International
• Make sure devices update automatically Journal of Scientific Research in Computer Science,
• Don’t click on email attachments from an Engineering and Information Technology
unknown source (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
• Don’t add people to your social media network 79-87, January-February 2019.
unless you know them Available at doi :
• Don’t use apps or software from free sites – Buy https://doi.org/10.32628/CSEIT19519
them from the official website instead Journal URL : http://ijsrcseit.com/CSEIT19519
• Routinely change passwords, making sure they
contain a unique combination of numbers and
letters (in both upper and lower case)
• Don’t use old, unsupported browsers

With these practices in place, we can expect a higher


level of cyber security in the year 2019.

X. REFERENCES

[1]. Gemalto, International data security company


[2]. Michael Fitzgibbon, Slice Insurance
Technologies
[3]. Dr. Andrew Shields, Cambridge Research
Laboratory of Toshiba Research Europe,
[4]. Scott Carter, Senior Manager – US, Venafi
[5]. Jan Oetjen, CEO of GMX
[6]. Andy Taylor, lead assessor, APMG
International
[7]. James Healey, Managing Director, Air IT

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


87
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195110

Prediction of Fish Yields in Lakes and Reservoirs from simple


Empirical Models using Artificial Neural Network (ANN) : An Review
*1D. Karunakaran, M. Balakrishnan2
1Research Scholar, Research and Development Centre, Bharathiar University, Coimbatore, Tamilnadu, India
2Principal Scientist, National Academy of Agricultural Research and Management, Hyderabad, Telangana, India

ABSTRACT

Prediction of reservoir yield is an important for fisheries managers to use appropriate scientific management
practices to increase the fishery production. Many mathematical or applied mathematics and Artificial Neural
Networks models were developed to predict fish production forecast of reservoirs. Ecology of reservoirs is
dynamic, extraordinarily advanced and nonlinear in nature. There are several drivers have an effect on the
fisheries, both internal and external environmental parameters. Many researchers have assessed fish yield
potential based on leaner models using multiple linear regressions. Accurate modelling to predict fish yield of
the reservoirs and lakes helps to understand behaviour of the system and managers can formulate appropriate
management practices to improve fish yield. This paper provides an in-depth review on existing model
developed from simple empirical estimation to high-level non-linear model for assessing fishery potential of
lakes and reservoirs.
Keywords : Fish Yields, Artificial Neural Network, MEI, CPUE, EBP, Automata Networks

I. INTRODUCTION foreign exchange earnings to the country. Most


important, the fisheries sector plays a great role in
Fish is nutritious food and rich in protein, minerals, the nation building by providing nutritional health
vitamins and essential nutrients. As fishes are very security.
cheap comparatively other animal proteins, many
low-income groups prefer fish in their diet. Most In India, reservoirs are prime resource in terms of
important, the fisheries sectors play a great role in both surface area and production. It offers immense
the nation building by providing nutritional health scope for increasing fish production. More than 3
security and livelihood support to many people in the million ha of manmade reservoirs available in the
country particularly in rural area and give many country can increase the production. As populations
ecosystem services to the mankind. India is richly are increasing rapidly, another blue revolution is
endowed with vast inland open waters in form of necessary to double the fish production from these
rivers and canals (1,71,334 km), reservoirs (3.15 resources to meet the demand. Accurately predicting
million ha), floodplain wetlands (0.24 million ha), reservoir yield will help fisheries managers to use
estuaries (0.27 million ha) and ponds and tanks (2.25 available various scientific management practices to
million ha) (Sinha 1999). These resources provide enhance the fishery production in the reservoir.
employment and livelihood support to many people
as well as contribute more to Indian economy by

CSEIT195110 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 88-100 ]
88
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

Ecology of reservoirs is dynamic, very complex and calculated (Wijayaratne and Amarasinghe, 1984;
nonlinear in nature. There are many drivers Nissanke et al., 2000) and (Hasan et al., 2001)
influencing the fisheries both internal and external predicted fish yield based on chlorophyll-a, Secchi
environmental parameters. Many researchers have depth and morpho-edaphic index in Bangaladesh.
tried to assess fish yield potential based on leaner
models by using multiple linear regression analysis
techniques (Jowett, 1993). These empirical models Others have found that phytoplankton production
are capable of solving many linear problems, but can predict fish yield more accurately (Oglesby,
some time it cannot give accurate results when 1977) or total phosphorus and macro benthos
relationships among the variables are non-linear in biomass/mean depth found to be the best predictors
nature. Advance tools like neural network, fuzzy of fish yield better than MEI (Hanson and
logic, wavelet transform and genetic algorithm will Legget,1982). Surface area alone or Lake Shoreline
give better and accurate results for non-linear development is useful predictor variables of fishyield
problems. in North American lakes (Youngs and Heimbuch,
1982). Hrbacek (1969) found that significant
II. PREDICTION OF FISH YIELD USING correlation between primary production and carp
EMPIRICAL MODEL yield in European ecosystems. Melack (1976) and
Toews and Griffith (1979) reported that primary
Many researchers predicated fish yields in lakes and production is an estimator of fish yield. Stocking
reservoirs for the last 70 years using simple empirical densities of fingerlings in inland reservoirs can be
estimation. Rawson (1952) demonstrated the first calculated based on estimation of potential fish
model to estimate fish yields of lakes based on mean production using morpho-edaphic index (MEI)
depth. Some others also tried to assess fish yield (Welcomme, 1976). Surface area can alone be the
potential from biotic and morphometric parameters best predictor of fish yield reported by researchers
(Hayes, 1957; Northcote and Larkin, 1956). Ryder from North American lakes (Youngs and Heimbuch,
(1965) improves existing estimation further by 1982).
developing morpho-edapic index (MEI) by using total
dissolved solids or conductivity/depth using 23 Models have been developed using reservoir depth,
temperate lakes. Later many researchers used other volume and area for estimation of fishery yield
dependent variables such as lake surface area, (Rawson, 1952; Jenkins and Morais, 1971; Moreau
temperature and other parameters along with MEI and De Silva, 1991). Downing et al., (1990) reported
for prediction of fish yields in lakes and reservoirs in annual phytoplankton production correlated (r2=
many countries. (e.g. Toews & Griffith, 1979; Jenkins, 0.79) than MEI, total phosphorus concentration and
1982; Schlesinger & Reglier, 1982; Machena & Fair, macro benthos biomass/mean depth, were the best
1986). Ryder (1965) applied this method in Canada. univariate predictors of fish yield (r2= 0.84 and r2=
Matuszek (1978) used mean depth and total dissolved 0.48, respectively). Hanson and Leggett (1982)
solids concentration of large North America lakes. reported catchment land use patterns are key
Morpho-edaphic index along with soci-economic parameters for fish yield prediction of reservoir (De
variables such as numbers of fisher men, boat and Silva et al., 2001). Rawson (1938) grouped lake
effort are used to predict yield of African lakes productivity parameters in three groups i.e climatic,
(Henderson and Welcomme, 1974). Maximum morphometric, and edaphic
Sustainable fish yield of Sri Lanka reservoirs were

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


89
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

and fishing effort. Crul)1992( correlated with catch )t


A. Potential Fish Yield from Morphological y-1( and area )km2( for 46 lakes and 25 reservoirs in
parameters Africa.

Reservoir area, volume, depth, and shoreline


Brämick, U and Roland (2003) estimated fish yield
development or gradient is some of the key
potential based on data of 786 lakes in north-east
morphological parameters to assess the productivity
Germany using primary production and total
(Ryder, 1978(. The mean depth alone a single most
phosphorus. Relationship between fish yield
important parameter to assess the fish yield )Rawson,
potential and total phosphorus or chlorophyll a or
1952; Henderson and Welcomme, 1974; Ryder et al.,
primary production of phytoplankton have been
1974; Mehner et al. 2007). Henderson & Welcomme
published and reviewed (Nurnberg 1996; Knösche, R.,
(1974) first applied Ryder's morphoedaphic index to
& Barthelmes, D. 1998). John A. Downing and Céline
tropical fisheries in African lakes and derived
Plante(2011) estimated annual fish production using
relationship from 17 fully exploited lakes where more
independent variables temperature, phosphorus
than 1 fisherman fishing km-2 lake area. Oglesby
concentration, chlorophyll a concentration, primary
(1977) reported that mean depth was not relevant
production, and pH .Knösche, R., & Barthelmes, D.
variable for assessing fish yield, where lake having
(1998) estimated lake fisheries yield from primary
more than 25m deep. Schlesinger & Regier (1982)
plankton production or total phosphorus. John Mark
found that fish yield - MEI relationships accurate
HansonandWilliam Leggett (2011) reported that total
only to lakes within regional level and generalizing
phosphorus concentration and macro benthos
the relationship need to include other dependent
biomass/mean depth best predictor than
variable such as mean annual air temperature into the
morphoedaphic index. Kolding, J., & Van Zwieten, P.
model. Hanson & Legget (1982) developed many
A. (2012) reported water-level fluctuation and mean
equations based on total dissolved solids (TDS)and
depthis a simple estimatorfor fish productivity in
total phosphorus concentration (TP( which was
tropical lakes and reservoirs.
highly correlated and much better predictor of fish
yield. Downing et al. (1990) has reported based on
Wijeyaratne and Amarasinghe )1987( showed that
his study that fish production is closely correlated
maximum sustainable yield )MSY( in several
with annual phytoplankton production (R2 = 0.79),
reservoirs of Sri lanka were correlated with the
mean total phosphorus concentration (R2 = 0.67), and
morphoedaphic index. Amarasinghe et al. )2002(
annual average fish standing stock (R2 = 0.67) but
have shown catchment features like ratio of
least correlated with the morphoedaphic index
catchment, land use patterns and reservoir capacity
(p>0.05). Schneider & Hadrich, (1989) reported that
are important predictor variables in Sri Lankan.
fish landings varied proportion to lake area.
Ramakrishna )1990( reported that catchment has a
Moreau & De Silva )1991( developed fish yield model positive impact on the Indian reservoirs productivity,
using multiple regression for lakes and reservoirs of provided the catchment is moderately fertile
Sri Lanka, Thailand and Philippines using area and
(Natarajan, 1976, 1977, 1979; Jhingran, 1986(.
effort. Moreau & De Silva )1991( also tested a number
of models using predictor variables catchment/lake Vollenweider (1969( reported that Flushing rate
area ratio, mean depth, transparency, total alkalinity, )inflow/storage capacity( is important variable for fish
chlorophyll a concentration, primary productivity production as it regulates nutrient loading.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


90
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

B. Potential Fish Production from Primary Rawson,1952; Depth, Models have


Production Jenkins and Morais, Volume and been
1971; Moreau and Area developed
Oglesby (1977) has derived relationships between De Silva, 1991; based on
fish and summer standing crop of phytoplankton. Bernascek, 1997 morphometric
Jones & Hoyer (1982), correlated sport fish harvests features
with summer chlorophyll a in US reservoirs and Ryder (1965) Total Morpho-
dissolved edaphic Index
lakes. Liang et al., (1981) derived relationship
solids or )MEI(
between net fish yield and gross primary production conductivity predictive
in sub-tropical Chinese lakes and ponds. Relative fish divided by yield model
biomasses expressed as gillnet CPUE were found mean depth for lakes and
correlated to chlorophyll a, total phosphorus, total reservoir
nitrogen and total organic matter in a large number D. R. Toews and J. S. Morphoeda Predicted fish
Griffith (1979) phic- yield using 31
of Argentinean lakes and reservoirs (Quiros, 1990).
index)MEI( African lakes.
Schlesinger & Regier Air Prediction of
Xiong )1996(, shown that 15 factors such as mean (1982) temperature fish yield from
depth, catchment area, water temperature, and MEI reservoirs
precipitation, dissolved oxygen content, total having less
phosphorus, phytoplankton biomass and the number than 25 m
of stocked fingerlings responsible for fish yield. depth
C. Fish Yields from Socio-economic Variables Oglesby (1977) and Chlorophyll Predictive
Biro & Vörös (1988) -a yield model
Fishing effort, number of fishermen or number of
boats or units of fishing gear are important predictor Ryder (1978) Reservoir Predict the
area, reservoir or
socio-economic variables for predict fish yield
Volume, lake
(Henderson & Welcomme, 1974; Bayley, 1988). Depth, and productivity
Shoreline
developmen
TABLE – I. LIST OF MODELS DEVELOPED TO
t or gradient
PREDICT FISH YIELDS IN LAKES AND Biro&Vörös (1988) Ratio of Predicting the
RESERVOIR and Moreau & De surface area fish yields of
Silva (1991) to lakes and
Author(s) Predictor Remarks catchment reservoirs in
variable area Sri Lanka
Rawson (1952) Mean depth Fish yield )CA(
prediction of
Lakes Nissanka, Ratio of the Better
Crul)1992( Area Establishedrela Amarasinghe & De CA to predictor
tion-ship Silva (2000) reservoir variable for
between catch capacity fish yield
and area from
46 lakes and Hrbacek (1969) Primary Reported
25 reservoirs production highly
in Africa significant

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


91
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

relationship cover )FC( Sri Lanka


between and shrub-
primary land )SL( to
production reservoir
and carp yield surface area
from European )RA( and/or
ecosystems. reservoir
capacity
Melack (1976), Gross Useful )RC(
Toewsand Griffith primary estimator of Amarasingheet al. Ratios of Prediction of
(1979) and production fish yield )2002( forest cover fish yield
Plante&Lalonde based on data and/or
(1990) of African shrub cover
lakes. to reservoir
Hanson and Total Better capacity or
Legget(1982) phosphorus predictors of reservoir
fish yields area
than the MEI Downing et al. Phyto- Prediction of
Jenkins and Morais Surface area Opined that )1990( plankton fish yield
(1971); Youngs and alone surface area production,
Heimbuch (1982) can alone be Total
the predictor Phosphorus
of fish yield of concentratio
North n , Annual
American average fish
lakes standing
Henderson and Electrical Studies based stock
Welcomme(1974) conductivity on Africa and Xiong)1996( Phyto- Methodology
Srilanka plankton developed to
primary predict silver
Morpau and De Silva Watershed Asian reservoir production carp and
(1991) area and bighead carp
Mean depth production(Ch
Ramakrishniah(1990 C/A radio Modified MEI ina)
), Ramakrishniah el along with model Xiong (1996) Mean depth, Reported
al. 1998) MEI incorporating Catchment major factors
the drainage area, Water influencing
parameter, the temperature fish yield
ratio of ,
catchment to Precipitatio
reservoir area n, Dissolved
(C/A), (based oxygen
on 19 Indian content,
reservoir) Total
phosphorus,
S. S. De Silvaet. Catchment GIS Model Phyto-
al.(2001) parameters using 11 plankton
i.e forest reservoirs of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


92
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

biomass and (1981) Primary )FYP(or Fish


Number of Hanson&Leggett production Biomass)FB(.
stocked (1982) of
fingerlings Peters (1986) phytoplankt
Henderson & Fishing Socio- Leach et al. (1987) on )PP(
Welcomme, 1974; effort economic Downinget al. (1990)
Bayley, 1988 ( number of variable to Qutros (1990 &
fishermen) predict fish 1991)
or number yield Barthelmes (1992)
of boats or Downing&plante
units of (1993)
fishing gear Nurnberg (1996)
John A. Downing Temperatur Estimated Knosche&Barthelme
and Céline Plante e, Lake annual s(1998)
(2011) phosphorus production of Reiner Knösche and Primary The most
concen- fish D. Barthelmes (1998)
plankton promising
tration, populations production limnological
Chlorophyll )PP( or parameters for
-a concen- Total fish yield
tration, phosphorus estimation
Primary JeppeKoldingandPau Water-level Reported
production, l A.M.van fluctuation simple and
and pH. Zwieten)2012( and mean robust
depth indicator of
Water-level Simple robust fish
Jeppe Kolding and fluctuation indicator of productivity in
PaulA.M.van and mean fish tropical lakes
Zwieten(2012) depth productivity in and reservoirs
tropical lakes Hoyer (1982) chlorophyll Correlated
Chlorophyll All these a sport fish
C Nissanka; U S a, Dissolved parameters harvests in US
Amarasinghe; and S phosphorus were found reservoirs
S De Silva and Total positively
phosphorus, influence with III. ARTIFICIAL NEURAL NETWORKS (ANNs)
Alkalinity to fish yield in
mean depth reservoirs
Artificial Neural Networks (ANNs) are mathematical
)MEIa( and
models designed to mimic the information processing
Conductivit
y to mean functions of a network of neurons in the brain
depth and (Hinton1992; Jensen 1994). Humans and animals are
CA/RC processing information by neurons. Computer
ratios algorithms mimic the way biological systems are
Dillon &Rigler Total Empirically functioning called artificial neural networks. The
(1974), phosphorus derived
brain is a highly complex, nonlinear and parallel
Oglesby (1977) )TP(, relations
Lianget al. (1981) Chlorophyll between Fish processing system computing many times faster than
Bulon&Vinberg -a or Yield Potential digital computer. It is widely used in many
disciplines for modelling. An artificial neural

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


93
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

network is highly popular because of similarity in Mediterranean Sea using ANN. The input variables
biological systems and has anability to learn from are biomass/abundance ratio, depth of the water
experiences and improving its performance. It is a column, latitude and longitude, and
very powerful tool for modelling non-linear and monthlysampling of675 sample stations of North
complex, imprecise and noisy data. It has an ability to Aegean Sea.Error back-propagation (EBP) algorithm
manipulate large amounts of data and generalize is used to learn the network.
results.
Baran, Pet.al.(1996) developed stochastic models to
A. Prediction of Fish Yields in Lakes and Reservoirs predict trout population density or biomass on a
using Artificial Neural Network mesohabitat scale using neural networks.
Backpropagation algorithm was used to train habitat
ANN have been used in many ecological modeling, variables width, gradient, mean depth, coefficient of
phytoplankton production (Scardi, 1996), fish species variation of depth, mean bottom velocity, coefficient
richness prediction (Brosse, Set al., 1998), and of variation of bottom velocity, froude number, area
prediction of density and biomass of various fish of cover, area of shelter, pool deep water area and
populations (Baran et al., 1996; Lek et al., 1996a,b; elevation. The result shows that the back propagation
Mastrorilloet. al., 1997). Many authors have reported neural network estimatedmore accuratelythan the
that ANN predicts more accurate than multi leaner multiple regression.
regression (Ehrman et al .1996; Leket. al. 1996b;
Scardi1996). Brosse, S. et. al. (1999) predicted fish spatial
occupancy and abundance in a mesotrophic reservoir
Laë, R., Lek, S., & Moreau, J. (1999) predicted fish using Artificial Neural Network.The input databased
yield of African lakes using Artificial neural networks on 306 observations collected using electrofishing.
using six input variables such as catchment area, The relationships between physical parameters and
maximum area, fishing effort, conductivity, depth, the abundance fish species are studied. Eight
altitude and latitude. The structure of the feed independent environmental variables depth, distance
forward neural network is six input neuron with one from the bank, slope of the bottom, flooded
hidden layer of five neurons, sigmoid functions, and vegetation cover, percentage of boulders, percentage
backpropagation algorithm for the training of the of pebbles, percentage of gravel and percentage of
ANNs. The result shows that correlation coefficients mud are used to quantify fish density. Structure of
between the estimated and observed values were the network is feed-forward with eight input neurons,
significantly very high. one hidden layer with 10 nodes with bias and
Zhang, H., & Zimba, P. V. (2017) developed multi- aoutput layer (8-10-1). The network was trained
layer neural networks to assess effects of estuarine using back-propagation algorithm to predict
freshwater fluxes on fish abundanceusing artificial accurately with minimum error.
neural network. The predictor variable consists of
inflow, evaporation, precipitation and annual catch Scardi, M., et.al (2008) evaluated ecological integrity
rate of fish species of Nueces Estuary. The network of streams and rivers using an expert system based
was trained using Levenberg-Marquardt back fish assemblage model. Twenty-seven environmental
propagation algorithm. Maravelias, C. D., Haralabous, parameters along with fish assemblage composition of
J., & Papaconstantinou, C. (2003) predicted 63 locations in Latium streams and rivers were used
distributions of demersal fish species in the in this study. Neural network was trained using most

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


94
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

popular error back-propagation algorithm. Aoki, I., maximum depth, shoreline perimeter, elevation, total
and Komatsu, T. (1997) predicted the winter catch of dissolved solids, pH, summer stratification (0, 1) and
young Japanese sardine (Sardinopsmelanostictus) littoral-zone predator (0, 1). Artificial neural
from climatic, hydrological and biological parameters networks predicted accurately abundance as well as
in the Joban-Boso Seas of the Pacific coast of central occurrence of fish species.
Japan using neural network. A feed-forward three
layers an input, hidden and an output layer is used. Kılıç, H., et. al (2007) developed ANN models to
Predictor variables are hydrological and biotic predict primary production of reservoir by using
parameters. Hydrological parameters consist of preprocessing technique of an Automata Networks
southern limit of the Oyashio, path type of Kuroshio, (AN) to find suitable variables for subsequent ANN
northern limit of the Kuroshio Extension, sea surface modeling. The AN based preprocessing followed by a
temperature in the northeastern sea area of Japanand ANN application predicted primary productivity
sea surface temperature in the fishing ground. accurately using Chl-a. The correlation coefficient as
Zooplankton, climatic southern oscillation index, far high as 0.83 and RMSE was as low as 2.69g/l was
east zonal index and east sea index are biological achieved withdouble hidden layer structure with 10
parameters. The network trained by back- neurons.
propagation algorithm to predictchanges in the
sardine abundance. Kuo, J. T., Hsieh, M. H., Lung, W. S., & She, N. (2007)
predict reservoir water quality using Artificial Neural
Joy, M. K., & Death, R. G. (2004) developed Network with back-propagation algorithm. The input
predictive model and spatial mapping of decapod variables are dissolved oxygen (DO), total phosphorus
assemblages using combination of GIS and neural (TP), chlorophyll-a (Chl-a), and secchi disk depth
networks. Single hidden-layer feed forward multi- (SD). Results show that correlation coefficients
layer perceptron trained by back propagation error between predicted values and measured data are 0.7
algorithm are used in this network. Catchment area, with reasonable accuracy.
average catchment elevation, elevation upstream end
of reach, average catchment slope, average annual IV. CONCLUSION
catchment rainfall, average catchment air
temperature, estimated river flow, reach length, Prediction of reservoir and lakes fish yield is the
latitude, stream order, catchment rainfall, distance important factor for the fishery managers to improve
from the coast, lake catchment area, catchment land the decision-making processes of reservoir fishery
use proportion and catchment geology proportions management and use appropriate management
(surface rock) are input variables. The decapod was practices to enhance the fish production. There is
predicted with high degree of accuracy from extensive literature available over the past decade on
geospatial landscapepredictor variables. prediction of fish yield in reservoir and lakes both in
linear and non-linear models. Many authors reported
Olden, J. D., & Jackson, D. A. (2001). Developed fish– that Artificial Neural Network model is giving high
habitat models for nine fish species using Artificial level of accuracy than Multilinear Regression Model.
Neural Networks. The study was conducted using
data of 128 lakes from the Madawaska river drainage
and 32 lakes from the Oxtongue river drainage of
Canada. The habitat input variables are area,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


95
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

II. REFERENCES fish production in different waters. The basis of


freshwater ecosystem investigation, 5-10.
[1]. Amarasinghe, U. S., De Silva, S. S., & Nissanka, [10]. Crul, R. C. (1992). Models for estimating
C. (2002). Evaluation of the robustness of potential fish yields of African inland waters.
predictive yield models based on catchment CIFA Occasional Paper (FAO). no. 16.
characteristics using GIS for reservoir fisheries [11]. D'Angelo, D. J., Meyer, J. L., Howard, L. M.,
in Sri Lanka. Fisheries Management and Gregory, S. V., & Ashkenas, L. R. (1995).
Ecology, 9(5), 293-302. Ecological uses for genetic algorithms:
[2]. Aoki, I., & Komatsu, T. (1997). Analysis and predicting fish distributions in complex physical
prediction of the fluctuation of sardine habitats. Canadian Journal of Fisheries and
abundance using a neural network. Oceanolica Aquatic Sciences, 52(9), 1893-1908.
Acta, 20(1), 81-88. [12]. De Silva, S. S., Amarasinghe, U. S., Nissanka, C.,
[3]. Aoki, I., & Komatsu, T. (1997). Use of a neural Wijesooriya, W. A. D. D., & Fernando, M. J. J.
network to analyse and predict the winter catch (2001). Use of geographical information systems
in the Joban-Boso Seas off the Pacific coast of as a tool for predicting fish yield in tropical
central Japan of young Japanese sardine reservoirs: case study on Sri Lankan reservoirs.
(Sardinopsmelanostictus) Fisheries Management and Ecology, 8(1), 47-60.
[4]. Baran, P., Lek, S., Delacoste, M., & Belaud, A. [13]. De Silva, S.S. and Amarasinghe, U.S. (eds.).
(1996). Stochastic models that predict trout (2009) Status of reservoir fisheries in five Asian
population density or biomass on a mesohabitat countries. NACA Monograph No.2. Network of
scale. Hydrobiologia, 337(1-3), 1-9. Aquaculture Centres in Asia-Pacific, Bangkok,
[5]. Bayley, P. B. (1988). Accounting for effort when Thailand. 116p.
comparing tropical fisheries in lakes, river‐ [14]. Dillon, P. J., &Rigler, F. H. (1974). The
floodplains, and lagoons. Limnology and phosphorus‐chlorophyll relationship in lakes 1,
Oceanography, 33(4part2), 963-972. 2. Limnology and oceanography, 19(5), 767-773.
[6]. Bramick, U., & Lemcke, R. (2003). Regional [15]. Downing, J. A., & Plante, C. (1993). Production
application of a fish yield estimation procedure of fish populations in lakes. Canadian Journal of
to lakes in north-east Germany. Limnologica- Fisheries and Aquatic Sciences, 50(1), 110-120.
Ecology and Management of Inland Waters, [16]. Downing, J. A., Plante, C., & Lalonde, S. (1990).
33(3), 205-213. Fish production correlated with primary
[7]. Brosse, S., Guegan, J. F., Tourenq, J. N., &Lek, S. productivity, not the morphoedaphic index.
(1999). The use of artificial neural networks to Canadian Journal of Fisheries and Aquatic
assess fish abundance and spatial occupancy in Sciences, 47(10), 1929-1936.
the littoral zone of a mesotrophic lake. [17]. Fausch, K. D., Hawkes, C. L., & Parsons, M. G.
Ecological modelling, 120(2-3), 299-311. (1988). Models that predict standing crop of
[8]. Brosse, S., Lek, S., & Dauba, F. (1999). Predicting stream fish from habitat variables: 1950-85. Gen.
fish distribution in a mesotrophic lake by Tech. Rep. PNW-GTR-213. Portland, OR: US
hydroacoustic survey and artificial neural Department of Agriculture, Forest Service,
networks. Limnology and Oceanography, 44(5), Pacific Northwest Research Station. 52 p, 213.
1293-1303. [18]. Hanson, J. M., & Leggett, W. C. (1982).
[9]. Bulon, V. V., &Vinberg, G. G. (1981). Empirical prediction of fish biomass and yield.
Proportions between primary production and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


96
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

Canadian Journal of Fisheries and Aquatic reservoirs. Transactions of the American


Sciences, 39(2), 257-263. Fisheries Society, 111(2), 176-179.
[19]. Hasan, M. R., Mondal, M. A. W., Miah, M. I., & [29]. Jowett, I. G. (1993). A method for objectively
Kibria, M. G. (2001). Water quality study of identifying pool, run, and riffle habitats from
some selected Oxbow Lakes with special physical measurements. New Zealand journal of
emphasis on chlorophyll-a. In. Reservoir and marine and freshwater research, 27(2), 241-248.
culture-based fisheries: Biology and [30]. Joy, M. K., & Death, R. G. (2004). Predictive
management. De Siva, S. S. (Ed.)2001. 384 pp. modelling and spatial mapping of freshwater fish
[20]. Hayes, F. R. (1957). On the variation in bottom and decapod assemblages using GIS and neural
fauna and fish yield in relation to trophic level networks. Freshwater Biology, 49(8), 1036-1052.
and lake dimensions. Journal of the Fisheries [31]. Kılıç, H., Soyupak, S., Tüzün, İ., İnce, o.,&
Board of Canada, 14(1), 1-32. Başaran, G. (2007). An automata networks based
[21]. Henderson, H. F., & Welcomme, R. L. (1974). pre-processing technique for artificial neural
The relationship of Morphoedaphic Index and network modelling of primary production levels
numbers of fishermen in African inland in reservoirs. Ecological modelling, 201(3-4),
fisheries. Committee for Inland Fisheries of 359-368.
Africa. Occasional Paper I, FAO, Rome, Italy. [32]. Knosche, R., & Barthelmes, D. (1998). A new
[22]. Henderson, H. F., &Welcomme, R. L. (1974). approach to estimate lake fisheries yield from
The relationship of yield to morpho-edaphic limnological basic parameters and first results.
index and numbers of fishermen in African Limnologica, 28(2), 133-144.
inland fisheries. Food and Agriculture [33]. Kolding, J., & van Zwieten, P. A. (2012).
Organization of the United Nations. Relative lake level fluctuations and their
[23]. Hinton, G. E. (1992). How neural networks influence on productivity and resilience in
learn from experience. Scientific American, tropical lakes and reservoirs. Fisheries Research,
267(3), 144-151. 115, 99-109.
[24]. http://www.fao.org/docrep/005/y3610e/y3610E3 [34]. Kuo, J. T., Hsieh, M. H., Lung, W. S., & She, N.
0.htm (2007). Using artificial neural network for
[25]. Jenkins, R. M., & Oglesby, R. T. (1982). The reservoir eutrophication prediction. Ecological
morphoedaphic index - concepts and practices: modelling, 200(1-2), 171-177.
The morphoedaphic index and reservoir fish [35]. Lae, R., Lek, S., & Moreau, J. (1999). Predicting
production. Transactions of the American fish yield of African lakes using neural
Fisheries Society, 111(2), 133-140. networks. Ecological modelling, 120(2-3), 325-
[26]. Jensen, M. C., & Meckling, W. H. (1994). The 335.
nature of man. Journal of applied corporate [36]. Leach, J. H., Dickie, L. M., Shuter, B. J.,
finance, 7(2), 4-19. Borgmann, U., Hyman, J., &Lysack, W. (1987).
[27]. Jhingran, A.G., 1986. Reservoir Fisheries A review of methods for prediction of potential
Management in India. Bulletin No. 45, Central fish production with application to the Great
Inland Fisheries Research Institute, Barrackpore. Lakes and Lake Winnipeg. Canadian Journal of
[28]. Jones, J. R., & Hoyer, M. V. (1982). Sportfish Fisheries and Aquatic Sciences, 44(S2), s471-
harvest predicted by summer chlorophyll-α s485.
concentration in Midwestern lakes and [37]. Lek, S., Delacoste, M., Baran, P., Dimopoulos, I.,
Lauga, J., &Aulagnier, S. (1996). Application of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


97
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

neural networks to modelling nonlinear [47]. Natarajan, A.V., 1977.Fifth Workshop of AICP
relationships in ecology. Ecological modelling, on Ecology and Fisheries of Freshwater
90(1), 39-52. Reservoirs. Ukai, 7-8 September, 1977.
[38]. Liang, Y., Melack, J. M., & Wang, J. (1981). [48]. Natarajan, A.V., 1979. Ecosystem oriented
Primary production and fish yields in Chinese approach for reservoir fisheries development in
ponds and lakes. Transactions of the American India. Souvenir on the ICAR Golden Jubilee
Fisheries Society, 110(3), 346-350. Year, CIFRI, Barrackpore, West Bengal, Part I,
[39]. Machena, C., & Fair, P. (1986). Comparison of 107-111.
fish yields from prediction models between [49]. Nissanka, C., Amarasinghe, U. S., & De Silva, S.
Lakes Tanganyika and Kariba. Hydrobiologia, S. (2000). Yield predictive models for Sri Lankan
137(1), 29-32. reservoir fisheries. Fisheries Management and
[40]. Maravelias, C. D., Haralabous, J., & Ecology, 7(5), 425-436.
Papaconstantinou, C. (2003). Predicting [50]. Northcote, T. G., & Larkin, P. A. (1956). Indices
demersal fish species distributions in the of productivity in British Columbia lakes.
Mediterranean Sea using artificial neural Journal of the Fisheries Board of Canada, 13(4),
networks. Marine ecology progress series, 255, 515-540.
249-258. [51]. Nürnberg, G. K. (1996). Trophic state of clear
[41]. Matuszek, J. E. (1978). Empirical predictions of and coloured, soft-and hard water lakes with
fish yields of large North American lakes. special consideration of nutrients, anoxia,
Transactions of the American Fisheries Society, phytoplankton and fish. Lake and Reservoir
107(3), 385-394. Management, 12(4), 432-447
[42]. McConnell, W. J., Lewis, S., & Olson, J. E. [52]. Oglesby, R. T. (1977). Relationships of fish yield
(1977). Gross photosynthesis as an estimator of to lake phytoplankton standing crop,
potential fish production. Transactions of the production, and morphoedaphic factors. Journal
American Fisheries Society, 106(5), 417-423. of the Fisheries Board of Canada, 34(12), 2271-
[43]. Mehrotra, S. N., &Jhingran, A. G. (1986). 2279.
Evaluation of limnological characteristics and [53]. Olden, J. D., & Jackson, D. A. (2001). Fish–
trophic status of Gulariya Reservoir. Proc. Nat. habitat relationships in lakes: gaining predictive
Acad. Sci. India, 53, 204-210 and explanatory insight by using artificial neural
[44]. Melack, J. M. (1976). Primary productivity and networks. Transactions of the American
fish yields in tropical lakes. Transactions of the Fisheries Society, 130(5), 878-897.
American Fisheries Society, 105(5), 575-580. [54]. Peters, R.H. (1986). The role of prediction in
[45]. Moreau, J., & De Silva, S. S. (1991). Predictive limnology.Limnol.Oceanogr. 31: 1143-1159.
fish yield models for lakes and reservoirs of the [55]. Quiros, R. (1990). Predictors of relative fish
Philippines, Sri Lanka and Thailand (No. 319). biomass in lakes and reservoirs of Argentina.
Food & Agriculture Org. Canadian Journal of Fisheries and Aquatic
[46]. Natarajan, A.V., 1976. Ecology and state of Sciences, 47(5), 928-939.
fishery development in some of the man made [56]. Raju, M. M., Srivastava, R. K., Bisht, D., Sharma,
reservoirs of India. In: Symposium on the H. C., & Kumar, A. (2011). Development of
development and utilisationof Inland Fishery artificial neural-network-based models for the
resources, Colombo, 27-29 October 1976. simulation of spring discharge. Advances in
Artificial Intelligence, 2011.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


98
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

[57]. Ramakrishnaiah, M., & Sarkar, S. K. (1982). S.H. Abidi, N.K. Thakur, R.S. Birader and L.
Plankton productivity in relation to certain Shenoy.(eds.) Vision on Indian Fisheries of 21st
hydrobiological factors in Konar reservoir Century. Central Institute of Fisheries
(Bihar). Journal of the Inland Fisheries Society Education, Bombay.
of India, 14(1), 58-68. [68]. Toews, D. R., & Griffith, J. S. (1979). Empirical
[58]. Ramakrishniah, M. (1990, October). estimates of potential fish yield for the Lake
Morphodrainage index, a fish yield predictor for Bangweulu system, Zambia, Central Africa.
Indian reservoirs. In 2nd Asian Reservoir Transactions of the American Fisheries Society,
Workshop, Hangzhou, PR China (pp. 15-19). 108(3), 241-252.
[59]. Rawson, D. S. (1952). Mean depth and the fish [69]. Unni, K.S., 1993.Limnology of Central Indian
production of large lakes. Ecology, 33(4), 513- lakes, reservoirs and ponds – a review. In:
521. Mishra, P.C. and Trivedy, R.K. (eds.). Ecology
[60]. Rempel, R. S., & Colby, P. J. (1991). A and pollution of Indianlakes and reservoirs.
statistically valid model of the morphoedaphic Ashish Publishing House, New Delhi.
index. Canadian Journal of Fisheries and Aquatic [70]. Vollenweider, R.A., 1969.A manual on methods
Sciences, 48(10), 1937-1943. for measuring primary productivity in aquatic
[61]. Rigler, F. H. (1982). Recognition of the possible: environment. IBP Hand Book No. 12. Blackwell
an advantage of empiricism in ecology. Canadian ScientificPub., U.K. 213 pp.
Journal of Fisheries and Aquatic Sciences, 39(9), [71]. Weimin, M., Liu, J., Li, Z., Xie, S., Vass, K.K.,
1323-1331. Sugunan, V.V., Pradhan, G.B., Amarasinghe,
[62]. Ryder, R. A. (1965). A method for estimating the U.S., Weerakoon, D.E.M. and Jutagate, T., 2006.
potential fish production of north‐temperate Status of Reservoir Fisheries in Five Asian
lakes. Transactions of the American Fisheries Countries.
Society, 94(3), 214-218 [72]. Welcomme, R. L. (1976). Approaches to
[63]. Ryder, R. A. (1982). The morphoedaphic resource evaluation and management in tropical
index—use, abuse, and fundamental concepts. inland waters. In FAO, Rome (Italy). Fisheries
Transactions of the American Fisheries Society, Dept. Indo-Pacific Fisheries Council.
111(2), 154-164. Symposium on the Development and Utilization
[64]. Ryder, R. A., Kerr, S. R., Loftus, K. H., &Regier, of Inland Fishery Resources. Colombo (Sri
H. A. (1974). The morphoedaphic index, a fish Lanka). 27 Oct 1976.
yield estimator—review and evaluation. Journal [73]. Welcomme, R.L.,1972. The inland waters of
of the Fisheries Board of Canada, 31(5), 663-688. Africa. FAO, CIFA Technical Paper, 1, 117pp.
[65]. Scardi, M. (1996). Artificial neural networks as Publ. FAO, Rome.
empirical models for estimating phytoplankton [74]. Wijeyaratne, M. J. S., &Amarasinghe, U. S.
production. Marine Ecology Progress Series, 139, (1987). Estimation of maximum sustainable fish
289-299. yield and stocking densities of fish fingerlings in
[66]. Schlesinger, D. A., & Regier, H. A. (1982). fresh water lakes and reservoirs.
Climatic and morphoedaphic indices of fish [75]. Wijeyaratne, M. R. S., &Amarasinghe, U. S.
yields from natural lakes. Transactions of the (1984). Estimations of maximum sustainable fish
American Fisheries Society, 111(2), 141-150. yields and stocking densities of inland reservoirs
[67]. Sinha, M. 1999. Vision of inland fisheries of of Sri Lanka. Journal of the National Aquatic
India of twenty first century. p. 154-168. In: Resources Agency, 31, 65-72.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


99
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

[76]. Youngs, W. D., & Heimbuch, D. G. (1982).


Another consideration of the morphoedaphic
index. Transactions of the American Fisheries
Society, 111(2), 151-153.
[77]. Zhang, H., & Zimba, P. V. (2017). Analysing the
effects of estuarine freshwater fluxes on fish
abundance using artificial neural network
ensembles. Ecological modelling, 359, 103-116.

Cite this article as :

D. Karunakaran, M. Balakrishnan, "Prediction of Fish


Yields in Lakes and Reservoirs from simple Empirical
Models using Artificial Neural Network (ANN) : An
Review", International Journal of Scientific Research
in Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 88-100, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195110
Journal URL : http://ijsrcseit.com/CSEIT195110

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


100
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195111

Evolving Cryptographic Approach for Enhancing Security of Resource


Constrained Mobile Device Outsourced Data in Cloud Computing
K. Kaviya 1, K. K. Shanthini 1, Dr. M. Sujithra2
1MSc. Software Systems, Coimbatore Institute of Technology, Coimbatore, Tamil Nadu, India
2Assistant Professor, Coimbatore Institute of Technology, Coimbatore, Tamil Nadu, India

ABSTRACT

Mobile cloud computing in the increasing popularity among users of mobile device enables to store their data in
cloud. Security is the major concern when the sensitive information is stored and transferred across the internet.
It is essential to make sure that the data is secured and protected. In this paper, we discussed the problem of
privacy of data with reducing the resources usage. Moreover, mobile cloud computing has limitations in
resources such as power energy, processor, Memory and storage. Cryptography ensures the confidentiality,
authentication, availability, and integrity of the data. This is done through cryptographic algorithms such as
Data Encryption Standard (DES), Blowfish and Advanced Encryption Standard (AES). The experimental results
evaluated and compared the performance of the encryption algorithms. The performance metrics used are
encryption and decryption time, CPU and Memory utilization. Evaluation results showed a significant
improvement in reducing the resources, amongst all the techniques, choosing a suitable encryption algorithm
based on different parameters that are best fit to the future user requirements is considered.

Keywords : Mobile Cloud Computing, Blowfish, AES, Security, Privacy, Mobile Device.

I. INTRODUCTION

Mobile cloud computing uses cloud computing to


carry out the resource intensive tasks over the
internet to provide higher scope of functionality with
minimal pressure on mobile resources. Cloud
computing is a modern era computing technique that
has a greater future and bringing a lot of benefit to
Figure 1.1 Cloud Environments
the information technology. The major advantage of
using cloud is that it offers the cloud services to
There are various layered architectures available for
people by Pay-as-You-Go manner. Cloud computing
cloud computing to provide the services as a utility.
applies a virtualized platform with elastic resources
Cloud’s backbone layer consists of physical servers
on demand by provisioning hardware, software, and
and switches. The cloud service provider is
data sets dynamically.
responsible to run, manage, and upgrade cloud
hardware resources according to the requirements of
users. The backbone layer is also responsible to

CSEIT195111 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 101-106 ]
101
K. Kaviya Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 101-106

allocate hardware resources to users in an efficient, can/cannot access the data. The principle of
quick, and smooth way. The supervisor software confidentiality specifies that only the sender and
layer contains the system software to manage the the intended recipient should be able to access
cloud hardware resources. The system software the contents of a message.
permits application software to run and utilize  Integrity: The integrity mechanism ensures that
underlying resources in an efficient way.In the the contents of the message remain the same
context of mobile cloud security, cloud providers when it reaches the intended recipient as sent by
should ensure reliability and availability by the sender. It is defined as the rightness of data
integrating security technologies MCC satisfies stored in the cloud. The alterations between two
limitations of resource constrained devices by updates of a record violate the data integrity.
allocating cloud services for the purpose of  Security: In the traditional file systems data was
application executions and data storage. So, whenever stored within boundaries, but cloud data is stored
we talk about the security in MCC, it is necessary to outside the boundaries of an organization, say,
consider the security and privacy issues. and third-party storage using strong encryption
techniques.
To secure the mobile cloud environment, several  Authentication: Authentication mechanisms help
issues need to be considered regarding the data to establish proof of identities. This process
security, data integrity, data confidentiality, ensures that the origin of the message is correctly
authentication, authorization, network security, data identified.
violation issues etc. A secure framework is essential  Non- repudiation: Non-repudiation does not
to protect sensitive data of mobile users with minimal allow the sender of a message to refute the claim
performance degradation. The proposed scheme uses of not sending the message.
data encryption to protect sensitive data leakage.  Access Control: Access Control specifies and
When user’s data is transmitted to and stored on controls who can access the process.
cloud, it creates many possibilities of unauthorized  Availability: The principle of availability states
access of data either during transmission or from that resources should be available to authorized
storage devices. There are many cryptographic parties all the times.
algorithms to deal with security, but the selection
must be taken by considering security as well as III. CRYPTOGRAPHY
performance improvement, especially in the case of
resource constrained devices. Cryptography is a method of protecting information
and communications with codes so that only those
II. SECURITY ISSUES for whom the information is intended can read and
process it. Cryptography refers to secure information
 Data Storage: Cloud storage providers manage the and communication techniques derived from
data in multiple copies across many independent mathematical concepts and a set of rule-based
locations calculations called algorithms to transform messages
 Confidentiality: Confidentiality can be defined as in ways that are hard to decipher. Security has always
the sensitive data not being disclosed to been an important term in all the fields. As the paper
unauthorized process, devices and person. A is discussing on cloud storage systems, the proposed
cloud service provider knows where the user’s methodology considers the security required to the
public or private data is located and who data that is stored on the cloud. Cryptography has

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


102
K. Kaviya Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 101-106

come up as a solution to this security issue. block size. It includes 64 bits key that contains 56 bits
Cryptography is classified into symmetric are directly utilized by the algorithm as key bits and
cryptography, asymmetric cryptography and hashing. are randomly generated. The DES algorithm
processes the 64 bits input with an initial
Types of Cryptography permutation, 16 rounds of the key and the final
permutation. The DES was initially considered as a
 Secret Key Cryptography: When the same key is strong algorithm, but today the large amount of data
used for both encryption and decryption, DES, and short key length of DES limits its use.
Triple DES, AES, Blow Fish RC5 etc., may be the
examples of such encryption, then that Advanced Encryption Standard (AES)
mechanism is known as secret key cryptography.
 Public Key Cryptography: When two different The algorithm explains about by AES is a secret-key
keys are used, that is one key for encryption and algorithm which means of the same key is used for
another key for decryption, RSA, Elliptic Curve both encrypting and decrypting the data. This is one
etc., may be the examples of such encryption, reason why it has a comparably small number of
then that mechanism is known as public key rounds. AES encryption is fast and flexible. It can be
cryptography. implemented on various platforms especially in small
 Hash Algorithms where the input data (message) devices. All the operations in this algorithm involve
is recreated from the hash value (message complete bytes for effective implementation. Three
digest/digest) examples include: MD5, SHA, different key lengths such as 128, 192 and 256 block
MD2, MD4, MD6.SHA-256, SHA-512, SHA-1, size are supported by AES.
Whirlpool etc.
Table 4.1 Comparison of Symmetric Algorithms

Algorith DES AES BLOWFIS


ms/Metr H
ics
Structure Feistel Substitution- Feistel
Permutation

Key 56 bits 32-448 128,192 or


Length 256
Rounds 16 10, 12, 14 16

Block 64 bits 128 bits/192 0r 256 64 bits


Size
Figure 3.1 Types of cryptographic algorithms Through < AES < Blowfish High
put
Security Adequat Excellent Excellent
IV. PROPOSED METHODOLOGY e
Speed Slow Fast Fast

Data Encryption Standard (DES)


Blowfish

DES utilized the one secret key for encryption and


Blowfish is fast, license free, unpatented, freely
decryption process and key length is 56 bits and
available and alternative for existing encryption
performs the encryption of message using the 64 bits

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


103
K. Kaviya Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 101-106

algorithms. It uses the key length range up to 32-448- CPU Time = I * CPI * T, where I = number of
and 64-bits block. Blowfish algorithm employed 16 instructions in program, CPI = average cycles per
rounds for the encryption process. Each round instruction and T = clock cycle time.
contains a key dependent permutation and key- and CPU Time = I * CPI / R, where R = 1/T the clock rate,
data dependent substitution. Data Encryption T or R are usually published as performance measures
involves the iteration of a simple function of 16 times. for a processor, I requires special profiling software
Each round contains a key dependent permutation and CPI depends on many factors (including
and data dependent substitution. Sub key Generation memory).
involves converts the key up to 448 bits long to 4168
bits. Blowfish is very suitable algorithm for the Performance Calculation
platform of smart phones because of its high security
level and high speed. Seconds/Program = (Instructions/Program) x
(Clocks/Instruction) x (Seconds/Clock)
V. PERFORMANCE EVALUATION METRICS
Memory Consumption Calculation
CPU Time Calculation Total Memory - (Free + Buffers + Cached) = current
total memory usage

Table 5.1. Experimental Results: Time Comparison

Time Encryption Time Decryption


Algorithm File Size
(milliseconds) (milliseconds)
10 KB 300 299
15 KB 303 304
Blowfish 20 KB 310 310
25 KB 315 313
30 KB 320 317
10 KB 303 303
15 KB 306 307
AES 20 KB 316 313
25 KB 320 317
30 KB 325 320
10 KB 308 303
15 KB 309 309
DES 20 KB 314 317
25 KB 318 318
30 KB 321 324

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


104
K. Kaviya Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 101-106

Table 2. Experimental Results: CPU and Memory Consumption

E- CPU D- CPU E-Memory D-Memory


Algorithm File Size
(%) (%) (KB) (KB)
10 KB 21.1 16.2 17.6 8
15 KB 22.7 18.1 17.9 8.3
BlowFish 20 KB 17.4 18.1 17.9 17.4
25 KB 20.2 18.8 17.6 17.4
30 KB 20.8 24.2 17.9 17.6
10 KB 21.4 22.2 20.4 20.3
15 KB 25.3 23 20.5 20.3
AES 20 KB 20.4 34 20.5 20.6
25 KB 25 23.7 20.6 20.1
30 KB 25.5 31.7 20.7 20.7
10 KB 22.4 22.2 21.7 20.4
15 KB 23.3 23.6 21.9 21.3
DES 20 KB 24.4 25.2 22.9 22.8
25 KB 25.3 26.7 23.9 22.9
30 KB 25.7 27.7 24.6 22.5

Figure 5.1 Time Comparison for Encryption and Decryption

Figure 5.2 CPU and Memory comparisons for Encryption and Decryption

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


105
K. Kaviya Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 101-106

VI.CONCLUSION RSA algorithms over cloud storage", Journal of


Theoretical and Applied Information
The demonstration of results and discussion about Technology, Vol.80. No.2 ISSN: 1992-8645,
these algorithms are mainly focused on evaluation October 2015.
parameter like encryption and decryption time, [6]. HealeyM(2010) Why IT needs to push data
memory and CPU utilization which has more impact sharing efforts. Information Week. Source:
on the security, confidentiality, integrity, and http://www.informationweek.com/services/inte
reliability for secure communication. Security as gration/why-it-needs-to-push-data-sharing-
earlier discussed is the main challenge faced while effort/225700544. Accessed on Oct 2012
storing data in the cloud, the proposed system [7]. D. Elminaam, H. M. A. Kader, and M. M.
provides security for the data stored in the cloud Hadhoud, "Performance evaluation of
computing model. Based on the performance symmetric encryption algorithms,"
evaluation, the results of Blowfish, AES and DES International Journal of Computer Science and
provide more security based on the resources Network Security, vol. 8, no. 12, pp. 280-286,
availability. In future we can use encryption 2008.
techniques in such a way that it can consume less [8]. Shanthni KK., Kaviya K and Sujithra M.2018, A
time and minimum energy consumption. Survey on Cloud Computing: Data Security
Challenges and Their DefensiveMechanisms.
VII. REFERENCES Int J Recent Sci Res. 9(5), pp. 26497-26500.
DOI:
[1]. Gurpreet Kaur and Manish Mahajan (2013), http://dx.doi.org/10.24327/ijrsr.2018.0905.2070
Analyzing Data Security for Cloud Computing [9]. D. Salama, A. Minaam, H. M. Abdual-kader,
Using Cryptography Algorithms‖, International and M. M. Hadhoud, "Evaluating the effects of
Journal of Engineering Research and symmetric cryptography algorithms on power
Application, Vol.-3,782-786. consumption for different data types,"
[2]. Sujithra, M., G. Padmavathi, and Sathya International Journal of Network Security, vol.
Narayanan. Mobile Device Data Security: A 11, no. 2, pp. 78-87, 2010.
Cryptographic Approach by Outsourcing [10]. L Krithikashree ; S. Manisha ; M Sujithra,"
Mobile Data to Cloud. In: Procedia Computer Audit Cloud: Ensuring Data Integrity for
Science 47; 2015.p. 480-485. Mobile Devices in Cloud Storage",Published in:
[3]. Paresh D.Sharma, Prof. Hitesh Gupta(February 2018 9th International conference on
2014) An Implementation for Conserving Computing, Communication and Networking
Privacy based on Encryption Process to Secured Technologies (ICCCNT), IEEE , DOI:
Cloud Computing Environment‖ IJESRT 10.1109/ICCCNT.2018.8493963
Sharma, 3(2).
[4]. D. S. A. Elminaam, H. M. A. Kader, and M. M. Cite this article as :
Hadhoud, "Evaluating the performance of K. Kaviya, K. K. Shanthini, Dr. M. Sujithra, "Evolving
Cryptographic Approach for Enhancing Security of Resource
symmetric encryption algorithms," Constrained Mobile Device Outsourced Data in Cloud
International Journal of Computer Theory and Computing", International Journal of Scientific Research in
Computer Science, Engineering and Information Technology
Engineering, vol. 10, no. 3, pp. 343-351, 2009.
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp. 101-106,
[5]. M. Sujithra, and G. Padmavathi, "Ensuring January-February 2019. Available at doi :
Security on mobile device data with two phase https://doi.org/10.32628/CSEIT195111

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


106
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT183857

Semantic Web Search Engines : A Comparative Survey


Anita Kumari, Dr. Jawahar Thakur
Department of Computer Science, Himachal Pradesh University, Shimla, Himachal Pradesh, India

ABSTRACT

Search engines play important role in the success of the Web. Search engine helps the users to find the relevant
information on the internet. Due to many problems in traditional search engines has led to the development of
semantic web. Semantic web technologies are playing a crucial role in enhancing traditional search, as it work
to create machines readable data and focus on metadata. However, it will not replace traditional search engines.
In the environment of semantic web, search engine should be more useful and efficient for searching the
relevant web information. It is a way to increase the accuracy of information retrieval system. This is possible
because semantic web uses software agents; these agents collect the information, perform relevant transactions
and interact with physical devices. This paper includes the survey on the prevalent Semantic Search Engines
based on their advantages, working and disadvantages and presents a comparative study based on techniques,
type of results, crawling, and indexing.
Keywords : Semantic Web, Semantic Search Engines, Intelligent Search

I. INTRODUCTION which synchronies and arranges all the data over the
web in a disciplined manner.
The web is a medium for accessing a great variety of So moving from current web to semantic web
information stored in different parts of the world. particularly depends on enhancement of knowledge
Information is mostly in the form of unstructured and overcome the current web issues. Semantic search
data. As the data on the web grows at explosive rates, plays an important role because it intelligently
it has lead to several problems such as increased understands the content and give smart and relevant
difficulty of finding relevant information, extracting results according to the quires. Traditional search
useful knowledge. engines get the answers syntactically accurate but
larger in amount. However, semantic engines provide
Conventional web search engines like Google and accurate information in terms of well-defined
yahoo return thousands of results for users quires expressions. The main aim of semantic web is to lift
because they do not understand the exact meaning of the capacity of users and software agents to find
quires. They uses keyword based search approach, documents, information and knowledge from the web
thus new technology “Semantic Web” comes into and make the content to machine and human
existence, which works on the semantics (meaning) of readable. This is possible by adding the semantic layer
queries. Semantic web offers a smarter web services to the current web.

CSEIT195115 | Received : 10 Jan 2019 | Accepted : 22 Jan 2019 | January-February -2019 [ 5 (1) : 107-115 ]
107
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

The organization of this document is as follows. In pyramid. The research provides deep
Section 2 (Literature Survey), literature survey is Understanding of five main semantic search
explained with comparative study. In Section 3 engines based on comparative analysis that may
(Keyword Based Search), keyword based search help for future work for semantic web in general
engines and the limitations are discussed. Semantic and for semantic search engines in particular.
web, approaches to semantic web and few search 4) Nikhil CHITRE [8], analyses the semantic search
engines (Hakia, Falcons, Swoogle, DuckDuckGO and engines generation into the intelligent web. Their
Knigne) and comparison on these search engines are focus is on finding the difference between the
discussed in Section 4(Semantic Web Search) based traditional and semantic search engines. Explains
on few parameters. Section 5 (Conclusion) concludes the various techniques and advantages of few
the paper by providing a summary of all reported important semantic web search engines that has
work in semantic web search. been developed so far.
5) Yi Jin Zhuying Lin, Hongwei Lin [9], states that
II. LITERATURE SURVEY search engines plays important role in success of
web; search engine helps any internet user to
There is significant literature work that has been rapidly find relevant information. Architecture
done on semantic web area. Various authors and of semantic search engine is explained and how
researchers have contributed very well in this area. elements of semantic search engine are used in
performing the fundamental task of information
1) Vidhi Shah1, Akshat Shah [5], concluded that retrieval is explained. An improved algorithm
searching the internet today is a challenge and based on TFIDF algorithm is also proposed.
semantic search engine enhances the traditional 6) Kyumars Sheykh Esmaili, Hassan Abolhassani
web search engines. They discuss various [10], presents a categorization scheme for web
semantic search engines and present a search engines. For each category, components
comparison based on the approaches. are explained according to a proposed general
2) G.Anuradha, G.Sudeepthi et al [6], presents a brief architecture and various approaches are used in
survey of the existing literature regarding these components also represented. It should note
intelligent semantic web search engine that some of these engines are constructed
technologies. Also, review their characteristics through a portal and therefore they have not
respectively. They concluded that existing individual names.
techniques have limitations particularly in terms 7) Ranjana Jain, Neelam Duhan el at [11], states that
of time response, accuracy of results, importance web 2.0 search engines are unable to present
of the results and relevancy of results. An direct answers against the users fired query. Web
efficient semantic web search engines should 3.0 is semantic web using RDF format
meet out these challenges efficiently and organization information in more structure form,
compatibility with global standards of web which helps the semantic search engines to
technology. present results in a more direct way. Also include
3) Maliha Majid Qureshi, Bibi Assma [7], presents these semantic search engines, review their
the semantic search engines and broad survey to working, and compare them based on few factors.
analyze the semantic search engines has done. 8) Jyoti Chaurasia, Jagdish Raikwal[12], states that
Comparative analysis of the semantic search semantic web technology plays an crucial role in
engines has done based on Factors cited in the enhancing traditional web search. They proposed

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


108
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

a semantic search engine based on ontology. An difference between the keyword based and
algorithm is also introduced to overcome the semantic based search engines.
various challenges in the fields of unstructured 10) G.Madhu1 and Dr. A. Govardhan [14], presents a
data using semantic web technology and provide brief survey of the existing literature regarding
the complete description. intelligent semantic search technologies. Review
9) Ankita Malve , P.M. Chawan [13], concluded that their characteristics respectively. In addition, the
semantic based search engine have more issues within the reviewed intelligent semantic
advantages over keyword based search engine in search methods and engines are concluded based
terms of accuracy of getting results. Search on four perspectives differentiations between
process in semantic search engine is based on designers and users’ perceptions, static knowledge
semantics of queries. It provides assurance to user structure, low precision and high recall and lack
to get accurate and relevant results based on of experimental tests.
meaning of word. They also include the basic
TABLE I
COMPARISONS BASED ON LITERATURE SURVEY

Reference Semantic Search Language Objects Conclusion


no. Engines
[6] Falcons, Kosmix, RDF, OWL, OIL, Differentiate Combining semantic
Sunbot, Hakia, DAML+OIL,ODEX. between semantic search with current
Swoogle, XCDSearch, search engines and search engines to make
XSearch, Kngine, presents their content understandable
Powerset, advantages and by human as well as
DuckDuckGO. disadvantages. machine.
[7] Hakia, Sensebot, QDEX, NLP. Research is carried This paper presents
Powerset, Cognition, out to explore the pyramid based
Lexxe. different dimensions comparison. Values of
of semantic search pyramid includes: search
engines. environment, intrinsic
problem, query type, and
iterative and exploratory.
[8] DuckDuckGO, Hakia, RDF, DAML+OIL, Analyse semantic The semantic search
Kngine. OIL, OWL, search engines engine improves the
QDEXing. generation and the limitation of traditional
part of search search engine and
technologies and provide 100% relevant
intelligent web. result.
[9] Architecture of OWL, XML, TFIDF How fundament Shows that how the
semantic search engine algorithm. task of retrieval is fundamental elements of
and improved performed. semantic search engines
algorithm based on perform the information
TFIDF are presented. retrieval.
[10] Content meta semantic RDF,OWL. Differentiate Provide a categorization
engine, crawler based between ontology scheme for different type
ontology search engine, semantic search of search engines on
evolutionary search engines and semantic web.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


109
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

engines, semantic semantic search


association discovery eninge.
engine.
[11] Architecture of search RDF, OWL, CSV, Why we need web Comparative study of
engines, meta-search XML. search engines. Google, Falcons,
engines, semantic Swoogle, Hakia, SWSE.
search engines
(Swoogle, Falcons,
Hakia, SWSE).
[13] Google, DuckDuckGO OWL, RDF, XML, Comparison Search process in
HTML. between keyword semantic search engine is
based search based on semantic of
engines and query and provide
semantic search assurance to get 100%
engine. accurate data.
[12] Google, a new RDF, OWL, Propose a semantic Proposed an algorithm
algorithm proposed in DAML+OIL, OIL, search engine based with ontology’s to give
addition with Google. QDEXing. on domain better search experience.
ontology.

III. KEYWORD BASED SEARCH • Low value of Recall and current Precision
parameters.
Search Engines are used to find the information on • More network congestion because multiple
the internet and getting results in less time but it crawlers try to crawl the same web pages form
lacks the existence of the semantic structure so it web servers.
becomes difficult for the machine to understand the • Polysemy words:- engines do not differentiate
meaning of the terms and expressions provided by between the same spelled words having different
the users. As the results, search engines returns the meaning hence give irrelevant pages.
ambiguous data sets. The survey indicate that user do • Stemming: - also these search engines don not
not find the accurate results in the first set of URLs differentiate between the singular and plural
returned because of the increased size of the links on words. Hence may produce the confusion while
the web pages. Because sometime one word has searching a query.
several meanings and several words have same • Synonymy: - it is the problem of understanding
meaning, in this case user will not get the result what the different spelled words with the same
he wanted to search [13]. Semantic web is developed meaning and it produce the confusion while
to overcome the following limitations of the current searching.
web search:- • Lack of semantic understanding [6][13][12].

• These engines are based on the occurrence of IV. SEMANTIC WEB SEARCH
words in the documents, so it is difficult to get
A. Semantic Web
the relevant results.
• Lack of proper semantic structure according to
It is an extension to WWW and international
the representation of information.
certified standard by W3C, which aims in handling

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


110
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

mostly unstructured data. It describes the 3) The Third approach focus on natural language
relationship between things and properties of things. understanding.
The main idea of semantic web is to add the metadata 4) The Fourth approach makes use of ontology to
into the current web pages to improve the represent the knowledge about a Particular
description of online resources. domain and expand quires [6][8].

C. Semantic Based Search Engines


Tim Berners-Lee originally expressed the vision of
the semantic web as follows ―I have a dream for the The semantic search describes that search engine will
Web [in which computers] become capable of give most accurate search results possibly by
analysing all the data on the Web – the content, links, understanding searcher intent, query content and the
and transactions between people and computers. A relation between words. Alternatively, we can say
Semantic Web, which should make this possible, has that semantic search engines will understand the
yet to emerge, but when it does, machines talking to natural language as the way human can understand
machines will handle the day-to-day mechanisms of and it improves the search accuracy of query related
trade, bureaucracy and our daily lives. The intelligent data and delivers exact and smart results. Semantic
agent’s people have touted for ages will finally search engines uses ontology to get the relevant and
materialize [1]. It presents the easier way to share, accurate results in less time. Ontology provides the
reuse and discover the online resources [2]. Also related association between the contents. RDF
represent the information in the way that human and (Resources Description Framework) and OWL (Web
machines can understand using semantics [2]. Ontology Language) are used for semantic web
documents. Rather than using ranking algorithms
(PageRank to predict relevancy), Semantics are used.
It provides the guaranty that most relevant results
will retrieve depending upon the meaning and
relations of the words not a specific keyword. The
goal of semantic search is to use the meaning
(semantic) to improve the search experience of the
users.

In order to access structured data, a variety of search


engine has been introduced which works on
semantics(meaning) of data and provide the accurate
Figure 1. Architecture of Semantic Web results in less time as compare to the traditional
search engine. The broad categories of semantic
B. Approaches to Semantic Web
search engine according to their workings are [10]
There is not only one approach in the semantic
search. Every approach has a lot to contribute. These
1) Crawler Base Search Engines,
main four approaches are used; different search
2) Human powered directories,
engine may use one or more approaches’.
3) Hybrid Search Engines,
4) Meta Search Engines.
1) The First approach uses contextual analysis to
disambiguate the search queries.
Some of the important semantic search engines are:
2) The Second approach focus on reasoning.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


111
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

1) SWOOGLE: - It is a crawler-based indexing and 4) KNGINE: - It is intelligent search engine and the
retrieval system for semantic web documents result is divided into either web results or image
(SWD). It uses the crawler to discover the RDF results. They are preceded by the information
and HTML documents. It extracts metadata and about the “concepts”. For instance, attempting to
computes relations between documents. Swoogle find a town can be preceded by using statistics
is also a content-based search engine that used to about the metropolis, nearby points of interest,
analyse, discover, and index the knowledge from activities, climate and hotels. It provides the high
the web. It is based on the page rank algorithm degree of related data to the user’s queries. It
and OWL language [11]. analyze the main keywords from the users
2) HAKIA:- It is a meaning based search engine that queries and form the relations between them and
works on matching the meaning rather than return results. It is a question-answer engine that
keyword match. The searching capacity of this understands the questions and tries to answer
engine is not limited to single keyword; you can with the help of natural language processing and
search sentences, a phrases or questions. This machine learning [8][6].
engine gets the results using equivalent terms. 5) DUCKDUCKGO: -“The search engine that does
Hakia engine essentially builds around mainly not track you” it is a feature rich search engine
three evolving technologies i.e., Sense that gives many reasons to put Google away. It
Repository, Query Indexing Technique and focuses on maintaining user privacy and
Semantic Ranking algorithm [8][6][11]. therefore does not store any personalized search
3) FALCONS: - It is a keyword based ontology results. This engine guaranty that they does not
search engines. It is develop around the concept store users IP address, does not users login
sharing and ontology reusing. It display the information. This engine only uses cookies to
required information on snippets, therefore user maintain the privacy of the users. It is a semantic
does not need to explore the pages. This engine search technology based search engine that
provides the best user interface. This engine presents relevant match or concept match for
integrates these two terms i.e., Concept-Level- example if a user searches for a keyword that has
Search and Ontology- Level-Search by more than one connotation, it gives you the
recommending ontology and allows filtering the options to choose what you were originally
concepts using ontology’s. Through this simple searching for along with the disambiguate
process user can easily find the ontology’s that outcome. This engine has the most important
satisfy their need within a few seconds [11]. property that it provides users privacy which
Google does not give [8][6][10].
TABLE II
COMPARISONS BETWEEN DIFFERENT SEMANTIC SEARCH ENGINES

SWOOGLE FALCONS HAKIA KNGINE DUCKDUCKGO

APPROACH Content based Keyword Content match, knowledge cluster search,


approach. Based NLP based , NLP.
approach. Statistical

TECHNIQUE 1.Rational Surfer Object OntoSem(Sense knowledge- Cluster search,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


112
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

Model, Ranked based Repository), Based- NLP


2.It uses crawler on QDEX( Query Approach,
based indexing combination Indexing Statistical
semantic search of their Method), Approach.
technique that relevance to Semantic
search ontology the query and Algorithm.
and instance data. their
popularity.
Uses cosine
similarity
measures
between the
query and
virtual
documents of
objects.

TYPE OF Entity(OWL,RDF) Entity(gives Documents Web Summaries(provide


RESULTS the result in (Links and free Results or Privacy)
snippet itself) Text) Image
Results
INDEXING Swangling Inverted Query Provide Instant
TECHNIQUE technique. Indexing detection and Answers
extraction
IT CRAWEL RDF Data RDF Data HTML Web Web pages RDF Data
Pages
FEATURES 1.Find exact 1.It integrate 1.Easily 1.Display 1.Zero click
ontology’s, 2.Finds concepts level identify results in information,
exact instance data search and information the form of 2.Provide privacy,
structure of the ontology level from credible image, 3.Does not record
semantic web. search, 2.User sites, 2.It is users information,
can find 2.Save time, multi- 4.Produces the result
results 3.Excellent lingual, based on many
quickly. response time. 3.It allows source and also from
the user to its own crawler.
search in
parallel
manner.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


113
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

LIMITATIONS 1.Poor indexing of 1.This engine 1.URL 1.Does not 1.It provides itself an
documents, does not rank Canonization, provides option of privacy but
2.Poor response the objects 2.Privacy silent mode they are just as
time for find according to Session IDS, option. opaque as Google is,
queries. query. 3.Does not since they don’t
index open source their
everything, code, how do you
4.Need other know they are not
search engines tracking you?
too,
5.Contains
virtual content
and dynamic
content.

V. CONCLUSION [2]. Shadbolt, Nigel, Tim Berners-Lee, and Wendy


Hall. "The semantic web revisited." IEEE
This paper includes the survey on the prevalent intelligent systems 21.3 (2006): 96-101.
Semantic Search Engines based on their advantages, [3]. Berendt, Bettina, Andreas Hotho, and Gerd
working and disadvantages and presents a Stumme. "Towards semantic web mining."
comparative study based on techniques, type of International Semantic Web Conference.
results, crawling, and indexing. It concluded that Springer, Berlin, Heidelberg, 2002.
semantic based search has more advantages over the [4]. Al-Feel, Haytham T., Magdy Koutb, and Hoda
keyword-based search and provide the relevant and Suoror. "semantic web on Scope: A New
accurate results in first URL. Semantic Web organizes Architectural Model for the semantic web."
the information in more structure form by using RDF (2008).
or OWL format, which helps the semantic search [5]. Shah, Vidhi, Akshat Shah, and Asst Prof
engines to, presents the results in way that is more Khushali Deulkar. "Comparative Study Of
accurate. Semantic Search Engines." International
Journal Of Engineering And Computer Science
Future work can be extended to the development of 4.11 (2015).
new semantic search engine technologies that [6]. Sudeepthi, G., G. Anuradha, and M. Surendra
overcomes their limitations. Prasad Babu. "A survey on semantic web search
engine." International Journal of Computer
VI. REFERENCES Science Issues (IJCSI) 9.2 (2012): 241.
[7]. Qureshi, Maliha Majid, Bibi Asma, and Hikmat
[1]. Berners-Lee, Tim, James Hendler, and Ora Ullah Khan. "Comparative analysis of semantic
Lassila. "The semantic web." Scientific american search engines based on requirement space
284.5 (2001): 34-43. pyramid." International Journal of Future
Computer and Communication 2.6 (2013): 562.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


114
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

[8]. Chitre, Nikhil. "Semantic Web Search Engine." Cite this article as :
International Journal 4.7 (2016).
[9]. Jin, Yi, Zhuying Lin, and Hongwei Lin. "The Anita Kumari, Dr. Jawahar Thakur, "Semantic Web
research of search engine based on semantic Search Engines : A Comparative Survey",
web." International Symposium on Intelligent International Journal of Scientific Research in
Information Technology Application Computer Science, Engineering and Information
Workshops. IEEE, 2008. Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
[10]. Esmaili, Kyumars Sheykh, and Hassan 5 Issue 1, pp. 107-115, January-February 2019.
Abolhassani. "A categorization scheme for Available at doi :
semantic web search engines." Computer https://doi.org/10.32628/CSEIT195115
Systems and Applications, 2006. IEEE Journal URL : http://ijsrcseit.com/CSEIT195115
International Conference on.. IEEE, 2006.
[11]. Jain, Ranjna, Neelam Duhan, and A. K. Sharma.
"Comparative study on semantic search
engines." International Journal of Computer
Applications, ISSN (2015): 0975-8887.
[12]. Chaurasia, Jyoti, and Jagdish Raikwal.
"Enhanced Semantic Web Retrieval Technique:
Domain Ontology." International Journal of
Computer Applications 119.16 (2015).
[13]. Malve, Ankita, and P. M. Chawan. "A
Comparative Study of Keyword and Semantic
based Search Engine." (2015): 11156-11161.
[14]. Madhu, G., Dr A. Govardhan, and Dr TV
Rajinikanth. "Intelligent semantic web search
engines: a brief survey." arXiv preprint
arXiv:1102.0831 (2011).

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


115
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195121

Image Dehazing Technique Based On DWT Decomposition and


Intensity Retinex Algorithm
1Sunita Shukla, 2Prof. Silky Pareyani
M. Tech Scholar, Gyan Ganga College of Technology Jabalpur, Madhya Pradesh, India
1

2Assistant Professor, Gyan Ganga College of Technology Jabalpur, Madhya Pradesh, India

ABSTRACT

Conventional designs use multiple image or single image to deal with haze removal. The presented paper uses
median filer with modified co-efficient (16 adjacent pixel median) and estimate the transmission map and
remove haze from a single input image. The median filter prior(co-efficient) is developed based on the idea that
the outdoor visibility of images taken under hazy weather conditions seriously reduced when the distance
increases. The thickness of the haze can be estimated effectively and a haze-free image can be recovered by
adopting the median filter prior and the new haze imaging model. Our method is stable to image local regions
containing objects in different depths. Our experiments showed that the proposed method achieved better
results than several state-of-the-art methods, and it can be implemented very quickly. Our method due to its
fast speed and the good visual effect is suitable for real-time applications. This work confirms that estimating
the transmission map using the distance information instead the color information is a crucial point in image
enhancement and especially single image haze removal.
Keywords : AF: Adaptive filter, AHE: Adaptive Histogram Equalization, LOE: Lightness Order Error

I. INTRODUCTION captured under bad visibility often has a contrast and


many of its features are difficult to see.
Restoration of hazy image is an important issue in
outdoor vision system. Image enhancement and II. METHODOLOGY
dehazing remain a challenging problem as well as an
important task in image processing. Image In this paper, we proposed a new method for single
enhancement is really an important issue in image image dehazing using the NAM (Non-symmetry and
processing applications such as digital photography, Anti-packing Model)-based decomposition and
medical image analysis, remote sensing and scientific contextual regularization. We estimated the airlight
visualization [1]. Image dehazing and enhancement is by decomposing the image using non-symmetry and
the process by which the appearance and the anti-packing model [11] to eliminate the false
visibility of an image are improved such that the estimation at the boundary or the over-bright object.
obtained image is suitable for visual perception of Then, the scene transmission was calculated using the
human beings or for machine analysis. It is useful not combination of the boundary constraints, the
only from an aesthetic point of view but also helps in contextual regularization and the optimization
image analysis and object recognition, etc. Image proposed by Meng et al. [12]. The proposed method
had better color visual and haze-free image when it

CSEIT195121 | Received : 10 Jan 2019 | Accepted : 23 Jan 2019 | January-February -2019 [ 5 (1) : 116-122 ]
116
Sunita Shukla, Prof. Silky Pareyani Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 116-122

comes to the image dehazing problem, a dehazing Then the haze-free image could be estimated:
model called Atmospheric Scattering Model is widely ( )
( )
used in ( ( ) )
I (x) =J (x)t(x) + A(1-t(x))
where is usually a small constant (0.001) in order to
where I(x) represents the hazy image, and J(x) avoid division by zero.
denotes the origin haze-free image(also called scene
radiance). A is the airlight which is the global light in Proposed an algorithm which combines the merits of
the atmosphere while t(x) denotes scene transmission transform color space algorithm and wavelet
function (0<t(x)<1). t(x) usually correlates with the transform algorithm.
scene depth. When the observed object is further
away from the camera, it’s more affected by the haze. First, the RGB image is converted to the HSI color
As a result, t(x) is closer to zero when the object space, Then histogram equalization is applied to
depth in the scene is greater. In the dehazing process, intensity component I to enhance the contrast of
the goal is to estimate the A (airlight) and construct image and the segmentation
the most optimal transmission function t(x).

Input Median R
Original Filter G
Image
B

Histogram Equalization

RGB to RGB Stretching


HSI

Hue Saturation Intensity

DWT

HH HL LH LL

Retinex
Algorithm

IDWT

Dehazed
Image

Figure 1. Flow diagram of proposed work

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


117
Sunita Shukla, Prof. Silky Pareyani Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 116-122

Exponential enhancement algorithm is applied to Let input image is x which is a RGB image
saturation component S, and then the intensity First Histogram Equalization need to be done
component I is divided into high and frequency sub-
( )
bands with wavelet transform and then Retinex
algorithm is applied to the low-frequency sub-band
( ) ( )
to reduce the effect of haze and adjust image ( )
√( ) ( )( )
luminance, a fuzzy enhancement algorithm is applied
to high-frequency sub-band to achieve the
enhancement and de-noising for the image details. (
( )
Finally, utilize the inverse wavelet transform to Let ‘img’ is the HSI image and its intensity block is of
reconstruct the I component and then the 3x3 is as below, and the intensity need to enhance
reconstructed component I will be synthesized with with K coefficient
H and S components to get a clear RGB image, and
the proposed algorithm is represented by following
flowchart [ ]

Table 3.1 Histogram equalization algorithm

Pixel intensity A B C D e

Pixel value f1 f2 f3 f4 f5

Probability f1/9 f2/9 f3/9 f4/9 f5/9

Cumulative F1/9
probability

CP*k K*F1/9
{ } { } { } { }

Floor Na=
rounding floor(K*F1/9) [{ } ] [{ } ]

[{ } ] [{ } ]

[ ] ( ) ∑ ( ) ( )

( ) ∑ ( ) ( )
is the intensity frame of HSI image of MxN DWT
applied on ‘I’
Table 1 below shows the symlet type 4 HPF and LPF ( )
filter coefficients. Proposed work use ‘sym4’ type ( )
wavelet for decomposition of Cover image, figure 1
below shows HPF and LPF decomposition using
DWT.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


118
Sunita Shukla, Prof. Silky Pareyani Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 116-122

Retinex let (x,y) are the pixels coordinates of ‘p’ in accelerate the processing speed, a fast single image
space domain then W is the reflection component dehazing method has to be developed in near future
and Z illumination component then which will be based on the atmospheric scattering
model and gray projection. our future method will
( ) ( ) ( ) have transmission map is estimated roughly using
minimum filtering which is based on dark channel
( ) ∑ ∑ ( ) ( )
prior and refined with fast average filtering. The gray
( ) projection method is adopted to obtain the
( )
atmospheric light. The subsection mechanism is
designed to avoid the high light of sky region in
Where c is Gaussian scale and is a constant that
recovered image, and Weber-Fechner Law is
makes F(x, y) equal to 1.
employed for color compensation in restored image. a
( ) ( )* ( ) ( )+
optimized approach is planned to be develop in near
( ) ( ( ))
future, in which the transmission map is roughly
( ( )) ( ( )
obtained with DCP and processed with a fast average
( ))
( ) ( ) filtering to improve the running speed. By gray
projection, the atmospheric light is obtained. And
Let (u,v) are the pixels coordinates of ‘q’ in space after the restoration of haze-free image, the intensity
domain compensation is employed to increase the visual
effect. The designed algorithm must achieve satisfied
( ) result with a computation cost.
( )

III. RESULTS AND DISCUSSION

( ) ( ( ) ) LOE measure is based on the lightness order error


( ) ( )( ) between original image X and enhanced image Y
( ) ( ) .The LOE measure is defined as
Mq is the final enhanced high frequency component
q
∑∑

∑ { ( ) ( ) }
is the relative order difference

Outdoor images are degraded by haze. The degraded


images have image contrast. This seriously affects ∑∑( ( ) ( ))
many computer vision applications such as video
surveillance systems and vehicle visual systems.
Dehazing methods have attracted much interest in The lightness L of an image is the maximum of its
recent years. Generally, it is difficult to prevent the three color channel.
reflection of haze. To restore the image contrast, ( )( )

dehazing methods need to be performed. To


Simulation is been taken for Crop image

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


119
Sunita Shukla, Prof. Silky Pareyani Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 116-122

Figure 2. GUI for the test image Crop

Figure 5. Test image Crop with histogram


enhancement

Figure 3. Histogram analysis of test image Crop

Figure 6. Enhanced test image Crop

Tables below shows the observed results of the


proposed simulation for test images 'house’,
‘building’, ‘Man’, Crop’ and 'Ground’.

Table 1 Observed results of MSE, PSNR and LOE

SN Test MSE PSNR LOE


Image
1 House 0.237564 48.97 0.890866
Figure 4. Test image Crop RGB to HSI with haze 2 Building 0.223472 48.4388 0.838019
enhancement 3 Man 0.300008 50.997 1.12503
4 Crop 0.135195 44.0736 0.506982
5 Ground 0.202547 47.5849 0.759551
Average 0.19652 47.234 0.862502

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


120
Sunita Shukla, Prof. Silky Pareyani Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 116-122

Table 2 LOE comparison for Building image changing when the homogeneity of RGB channels
have a big difference. The results obtains for the
Author/ Journal/Year LOE for Image of thesis work are better in terms of LOE and SNR then
Building available works.
Yuhei Kudo / IEEE 2018 1.932
Proposed 0.838019
V. REFERENCES

Table 3 Mean Square Error comparison of Crop


[1]. Yuhei Kudo, Akira Kubota,Image Dehazing
Image
Method by Fusing Weighted Near-Infrared
Image, 978-1-5386-2615-3/18/ ©2018 IEEE
Work MSE for image of
[2]. Elisee A Kponou, Zhengning Wang, Ping wei,
Crop
Efficient Real-time Single Image Dehazing
Proposed 0.135195
Based on Color Cube Constraint, 2017 IEEE
Yunping Zheng/ IEEE 2017 0.1393 2nd International Conference on Signal and
Image Processing, 978-1-5386-0969-
9/17/©2017 IEEE
IV.CONCLUSION [3]. Yunping Zheng, Zhenfeng Xie, Changting Cai,
Single Image Dehazing Using Non-symmetry
The proposed work studies different types of methods and Anti-packing Model Based Decomposition
and technologies that have been used for image and Contextual Regularization, 2017 13th
dehazing and observed that the low contrast and International Conference on Natural
noise remains a barrier to visually pleasing images in Computation, Fuzzy Systems and Knowledge
dehazing conditions. In that condition, to find out a Discovery (ICNC-FSKD 2017), 978-1-5386-
more accuracy in image enhancement process there is 2165-3/17/©2017 IEEE
need to detect and measure the intensity level of [4]. Takuya Mikami, Daisuke Sugimura and
individual pixel channel as well as have to present an Takayuki Hamamo to, CAPTURING COLOR
appropriate enhancement factor for enhancement AND NEAR-INFRARED IMAGESWITH
purpose, so that effective and efficient image DIFFERENT EXPOSURE TIMES FOR IMAGE
enhancement process will be created. A new method ENHANCEMENT UNDER EXTREMELY
for image dehazing based on DWT Decomposition HAZE SCENE, 978-1-4799-5751-4/14/2014
and intensity Retinex. The air-light of image is IEEE
estimated by decomposing the image using non- [5]. Xiankun Sun, Huijie Liu, Shiqian Wu, Zhijun
symmetry and anti-packing model to refine the Fang, Chengfan Li, and Jingyuan Yin, Haze
illumination value. Next, the scene transmission Image Enhancement Based on Guided Image
function is calculated using the combination of the Filtering in Gradient Domain, Hindawi,
boundary constraints and the contextual International Journal of Digital Multimedia
regularization. The proposed method produces high Broadcasting, Volume 2017, Article ID
quality dehazed picture in most cases and decrease 9029315, 13 pages,
artifact. What’s more, the tone of the image is https://doi.org/10.1155/2017/9029315
natural. But there’s still some points needing to be [6]. Pixel Binning Yoonjong Yoo , Jaehyun Im and
improve such as the time consuming and the tone’s Joonki Paik, Haze Image Enhancement Using

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


121
Sunita Shukla, Prof. Silky Pareyani Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 116-122

Adaptive Digital, Sensors 2015, 15, 14917-


14931; doi:10.3390/s150714917, ISSN 1424-
8220, www.mdpi.com/journal/sensors
[7]. Zhenqiang Ying, Ge Li, Yurui Ren, Ronggang
Wang, and Wenmin Wang, A New Haze
Image Enhancement Algorithm using Camera
Response Model National Science Foundation
of China (No.U1611461), Shenzhen Peacock
Plan (20130408183003656), and Science and
Technology Planning Project of Guangdong
Province, China (No. 2014B090910001 and No.
2014B010117007).
[8]. Zhuang Feng, Haze Image Enhancement by
Refining Illumination Map with Self-guided
Filtering, 2017 IEEE International Conference
on Big Knowledge, 978-1-5386-3120-1/17 2017
IEEE, DOI 10.1109/ICBK.2017.37

Cite this article as :

Sunita Shukla, Prof. Silky Pareyani, "Image Dehazing


Technique Based On DWT Decomposition and
Intensity Retinex Algorithm", International Journal
of Scientific Research in Computer Science,
Engineering and Information Technology
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
116-122, January-February 2019. Available at doi
: https://doi.org/10.32628/CSEIT195121
Journal URL : http://ijsrcseit.com/CSEIT195121

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


122
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195116
Data Extraction in Malignant Environments
Amith Mandal1, Dr. R. P. Ramkumar2
1M. Tech Scholar, Department of CSE, Malla Reddy Engineering College (A), Telangana, India
2Professor, Department of CSE, Malla Reddy Engineering College (A) Telangana, India

ABSTRACT

Deliberate or inadvertent escape of classified data is undoubtedly one among the premier extreme security
dangers that associations look inside the advanced period. The risk right now stretches out to the private lives:
an abundance of private information is out there to interpersonal organizations and great telephone providers
and is in a roundabout way exchanged to undependable outsider and fourth gathering applications. amid this
work, a bland data genealogy structure LIME (Data Lineage in the Malicious Environment), is utilized for data
stream over different elements that take two trademark, guideline parts (i.e., proprietor and buyer). It
characterizes the exact security ensures required by such a data heredity component toward recognizable proof
of a blameworthy substance, and decide the disentangling non-denial and trustworthiness suspicions. At that
point create and break down a totally special dependable data exchange convention between two substances
among noxious surroundings by expanding upon unmindful exchange, solid watermarking, and mark natives.
At long last, an exploratory investigation to exhibit the helpfulness of convention what's more, apply structure
to the important data run projections of learning outsourcing and the informal organizations. In general,
LIME(Data Lineage in the Malicious Environment),lineage structure for information exchange, to be a key
advance towards accomplishing responsibility by plan.
Keywords: Classified Data, LIME, Trademark, Malicious Environment, Micro Benchmarking, Pairing Based
Cryptography, GNU

I. INTRODUCTION In the past few decades there has been an exponential


growth in technology in which the major part was
Data Extraction avoidance is the type of solutions the revolution of computers. This revolution has
which can help an institute to apply control for the taken upon every aspect both resourcefully and
prevention of unwanted, unintentional or leakage of individually, this growth has surely changed our lives
specific information to illegal entities inside or by making it much easier. In the process of this
outside the institute. Here specific information may growth there has been a severe risk which everyone
refer to the institute’s internal processing of may not be conscious of which is data extraction.
documents, tactical business plans, academic property, Data extraction is basically a scenario where any
economic statements, security policy, network intruder will access our personal data and make the
architecture, blueprints etc. extracted data public or the intruder might use it for
any other intentions. Today we have a huge access of
smartphones and gadgets which makes data

CSEIT195116 | Received : 15 Jan 2019 | Accepted : 25 Jan 2019 | January-February -2019 [ 5 (1) : 123-130 ]
123
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

extraction much easy as the impostor can now easily outcome .Amid this distribution the prerequisite for
transfer the data. general answerability instrument in learning
exchange is demonstrated which may specially go
We as a consumer find many apps on internet and we with confirmable measures and control the condemn
knowingly handle our data to the respective blameworthy. This philosophy of identification is
organizations but is our data really safe in the hands named as learning root. Endeavors are made by
of the organization, what if any employee of the numerous associations to utilize learning source
fellow organization decides to make a step that ruins inside the type of watermarking or including of false
the organizations name, what if any employee information or data in specially appointed way has
wantedly access your data and misuses it. The answer not been that flourishing.
to that problem is represented in the following work
which will determine the guilty entity with proof II. EXISTING SYSTEM
and held for explanation if any data extraction occurs.
There are few safety mechanisms such associate In the digital era, data outflow during intentional
secret writing of valuable data, however it's a exposures, or un-intentional damage by untruthful
proactive measure that cannot verify the info leak staff and harmful external entities, always provide
once it happens. Once the intruder decrypts the serious threats to organizations. it's not laborious to
information then there's no potential live which will believe that this is simply the edge of an iceberg, as in
prevent the info leak of the personal data and most cases information of knowledge outflow go
business data by any desired measures. Following are unreported thanks to concern of loss of client
few measures that we often see and there are a confidence or restraining penalties huge amount of
number of situations where the event is determined. digital data can be derived at nearly no price and
might be unfold through the net in terribly short
1. We should take facebook for instance wherever we time to boot, the risk of obtaining or getting caught
can see assortment of outsiders that approaches our for information outflow or the extraction is very low,
insight, if any outsider is experienced in learning spill as there square measure presently nearly no
it winds up impractical to call attention to or see that answerableness mechanisms. For these reasons, the
wherever extremely the data was misused as there matter of information loss has reached a replacement
square measure a few outsiders advertised. dimension these days.
2. Consultancies have all the valuable data of
representatives/applicants caning to seek out work 2.1 Disadvantages
which outsider contracts another outsider then all Even with access management mechanisms,
the three organizations will right now approach the wherever access to sensitive information is restricted,
individual information then it may turn out to be a malicious approved user will publish the valuable
horrendously difficult to undeniably relate one the information as soon as he receives it. Primitives like
three partnerships for learning spill, in each the coding provide protection solely as long because the
circumstances witnesses can either don't target the info of interest is encrypted, however once the
exact learning safety or the intruder may attentively recipient decrypts a message, nothing will stop him
reveal classified data which cannot be tied to any of from publication the decrypted content. Therefore it
the handlers. At times there comes experience the looks not possible to stop information outpouring
explanatory procedures however it's horribly proactively.
expensive and it also doesn't create a terribly sensible

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


124
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

2.2 Proposed System The key advantage of this representation is that it


In this work, the requirement for a general applies answerableness right from design; i.e., it
answerability mechanism in information transfers is drives the system designer to contemplate doable
got wind. This answerability are often directly information leakages and also the corresponding
related to demonstrably detective work a answerableness constraints at the look stage. This
transmission history of information across multiple helps to beat the prevailing state of affairs wherever
entities ranging from its origin. This can be called most lineage mechanisms ar applied solely once an
information place of origin, information lineage or escape went on.
supply tracing.
III. MODULES
Here the matter is dignified of practically associating
the problem to the extraction, and work on the 3.1 Data Owner
information lineage methodologies to unravel the In this module owner uploads their pictures with
matter of knowledge leak in numerous leak situations. their contents data to the net server. For the safety
This system defines LIME, a generic information purpose to the information, owner assigns the digital
lineage framework for information flow across sign and stores it inside the internet and in addition
multiple entities within the malicious atmosphere. performs the subsequent operations to transfer image
with its digital signature with supported title,
Here it's observe that entities in information flows description
assume one amongst 2 roles: owner or shopper and Below are the few operations an Owner can perform
introduce an extra role within the sort of auditor, 1 .Upload image or a file,
whose task is to see a problem for any data extraction, 2. Verify image or file details
and summarize the accurate properties for 3. View Consumer requests and
communication between these roles.In the process 4. Deleted image details
Associate in nursing non-mandatory non-repudiation
assumption created between two house owners, 3.2 Web Server
Associate in nursing and non-mandatory trust The Web service supplier manages an internet to
(honesty) assumption created by the auditor supply information storage service. And performs the
concerning the house owners. subsequent operations reminiscent of Storing all
image files with their signature, read all image Files
The second contribution, associate in nursing with its details, view all image comments, and view
responsible for information transfer protocol to all information house owners and Users, read all
provably and securely transfer valuable information attackers.
between the two parties is bestowed. It deals with
associate in nursing and untrusted sender, associate in A web server will be monitoring the system held
nursing and untrusted receiver situation related to responsible for a secure login including analyzing of
data transfer between any two customers, the all the data that has been transferred between two
protocol we use is an interesting combination of the entities
sturdy watermarking, oblivious transfer, and
signature primitives. The basic operations that are performed by the web
server are as follows
2.2.1 Advantages of Proposed System 1. View Owner and consumer Details

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


125
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

2. View all the files details usage of the Cox calculation is utilized for vigorous
3. View the attacked files picture and set the negative factor, which decides the
4. View Download history quality of the watermark, to a estimation of 0.1.Now
3.3 Auditor execute the explore different avenues regarding
Auditor has capability to access, manage or monitor diverse parameters to examine the execution. The
the transferred knowledge or data beneath the sender and beneficiary piece of the convention will
entrustment of information of owner, auditor has be executed in a similar program, i.e., don't break
experience and capabilities that a standard user down system sending, yet just computational
doesn't have, for occasionally auditing the outsourced execution.
knowledge. This audit service is considerably
necessary for digital forensics and credibleness in 4.2 PBC (Pairing Based Cryptography)
Web and performs the subsequent operations
1.View image/file details Blending based cryptography is a generally youthful
2. View Attacked files territory of cryptography that rotates around a
3. Block or unblock an attacker certain capacity with exceptional properties.
4. View messages from a consumer
The PBC (Pairing-Based-Cryptography) library is a
3.4 Data Consumer free C library based on the GMP library that plays
The net user of an Organization encompasses a great out the numeric activities and fundamental matching
quantity of data to be stored in servers and have the based on cryptosystems.
permissions to access and manipulate and keep image
and its data. The PBC library is planned to be the foundation of
executions of matching based cryptosystems, in this
The consumer can search information and accessing manner speed and portability are key objectives. It
the image data if he's approved and performs the provides schedules by comparing it to the elliptic
subsequent operations bend era, elliptic bend number juggling and blending
1.View all the files calculation. Due to the GMP library, in spite of being
2. Download a file composed in C, pairings times are shabby. On a 1GHz
3. Send download request Pentium III:
4. Send message to the auditor
• Fastest matching is 11ms
IV. TECHNIQUES USED • Short matching is 31ms

4.1 Micro Benchmarking The API is sufficiently dynamic that the PBC library
The execution of the convention is done in as a will be used despite of the fact that the associated
proof-of-idea and to determine its result. For the researcher has exclusively a rudimentary
unaware exchange sub-convention is executed to the comprehension of pairings. There’s no must be
convention utilizing the PBC library, which itself constrained to get some answers concerning elliptic
makes utilization of the GMP library. For marks BLS bends or a significant measure of collection
plot is executed, likewise utilizing the PBC library. hypothesis. (The base necessity is a few information
For symmetric encryption a usage of AES from the of cyclic gatherings and properties of the blending.)
Crypto++ library is utilized. For watermarking a

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


126
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

The PBC library can likewise be utilized to construct concerning once every year. Since variant about six,
customary cryptosystems. GMP is appropriated beneath the twin licenses,
wildebeest LGPL v3 and wildebeest GPL v2. These
licenses make the library freed to utilize, share, and
enhance, and allow you to kick the bucket the
outcome. The wildebeest licenses offer flexibilities,
Fig 1. Two party one round key agreement protocol however conjointly set firm limitations on the use
with without non programs. GMP's primary target
3.3 The GNU Multiple Precision Arithmatic Library stages are Unix-type frameworks, reminiscent of
GMP is attentively designed to be fast for small GNU/Linux, Solaris, HP-UX, waterproof OS
operands as well as for large operands. The speed is X/Darwin, BSD, AIX, and so on. It is additionally
accumulated by use of adequate words as basic celebrated internationally to figure on Windows in
arithmatic type, by use of quick algorithms, with each 32-bit furthermore, 64-bit mode.
immensely advanced assembly code for the most
familiar inner loops for a lot of CPUs, by a general 3.4 BLS Scheme
insistence on speed. In cryptography, the Boneh– Lynn– Shacham (BLS)
signature topic allows a client to confirm that an
GMP could be a free library for subjective accuracy endorser is bona fide. The subject uses a straight
number juggling, in task on marked whole numbers, blending for confirmation, Associate in nursing
objective numbers, and gliding point numbers. marks square measure segments of an elliptic bend
there's no sensible cutoff to the accuracy aside from group. working in Associate in Nursing elliptic bend
those verifiable by the available memory inside the bunch gives some safeguard against list math assaults
machine GMP keeps running on. GMP highlights a (with the proviso that such assaults square measure
made arrangement of capacities, and furthermore the still achievable inside the objective bunch of the
capacities have a regular interface. The fundamental matching), allowing shorter marks than FDH marks
target applications for GMP are cryptography for an indistinguishable level of security. Marks made
applications and investigation, web security by the BLS signature subject square measure
applications, unadulterated science frameworks, normally talked as short marks, BLS short marks, or
process unadulterated arithmetic examination, and so just BLS marks. The mark subject is undeniably
forth. secure.

GMP is thoroughly intended to be as brisk as feasible, 3.5 Cox Algorithm


each for little operands and for huge operands. The The Cox– Zucker machine is a calculation made by
speed is accomplished by exploitation full words in David A. Cox and Steven Zucker. This calculation
light of the fact that the fundamental number decides whether a given arrangement of areas gives a
juggling sort, by exploitation brisk calculations, with premise (up to torsion) for the Mordell– Weil
to a great degree enhanced gathering code for the gathering of an elliptic surface E → S where S is
principal regular inward circles for loads of CPUs, isomorphic to the projective line. The recipe was
and by a general weight on speed. The primary GMP introductory uncovered inside the 1979 paper
unharness was made in 1991. it's consistently created "Crossing point quantities of segments of elliptic
and kept up, with a substitution unharness surfaces" by Cox and Zucker and it totally was later

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


127
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

named the "Cox– Zucker machine" by Charles inside the archive, this can be not an inconsequential
Schwartz in 1984. assignment.
V. RESULTS

Fig 3. file for auditor

Fig 2.Working of cox algorithm

3.6 Possible Data Distortion


Here a simple cacophonic algorithmic program is
utilized. Here the picture is spilt into n similarly
estimated squares. In any case, once utilized a
powerful watermark for the modest components
(that will be that the - issue used by the Cox
algorithmic program is zero.5), varieties between Fig 4.web server
neighboring components ended up noticeable in spite
of the way that the one watermarks zone unit in
cognoscible.This outcome turns out to be
considerably more grounded when numerous
emphases as decided. Now and again, this bending
would potentially affect the ease of use of the record.
The Cox algorithmic program is utilized with relates
alpha issue of zero.1 and no mutilation is noticeable.
It would entrance to examine if this drawback are
regularly dodged by exploitation extra detailed
cacophonic calculations. As most watermarking plans
assemble utilization of the nearness of information Fig 5. consumer download page

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


128
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

create the predefined assurance for touchy data.


LIME is adaptable as we have a tendency to separate
between beyond any doubt senders (typically
proprietors) and untrusted senders (generally buyers).
Inside the instance of the beyond any doubt sender, a
dreadfully simple convention with next to no
overhead is achievable. The untrusted sender needs a
considerable measure of modern convention,
however the outcomes don't appear to be bolstered
put stock in presumptions furthermore, along these
Fig 6. Attacker being detected and blocked
lines they should be prepared to change over an
unbiased substance (e.g. a judge). Our work
conjointly rouses any investigation on data release
recognition procedures for differed report sorts
furthermore, projections. To Illustrate, it'll be a
propelling future examination heading to style a
certain ancestry convention for determined data.

VII.REFERENCES

Fig 2. A message request from a non guilty consumer [1]. "Chronology of data breaches,"
http://www.pri"acyrights.org/data-breach.
VI. CONCLUSION AND FUTURE ENHANCEMENT [2]. "Data breach cost,"
http://www.symantec.com/about/news/release/
We are blessing LIME, a model for capable data article.jsp?prid=2011030801
exchange over different elements. We tend to plot [3]. "Pri"acy rights clearinghouse,"
teaming up parties, their interrelationships and give a http://www.pri"acyrights.org.
solid inner portrayal for a data exchange convention [4]. "Electronic Pri"acy Information Center (EPIC),"
utilizing a novel blend of negligent exchange, strong http://epic.org, 1994.
watermarking and computerized marks. We have a [5]. "Facebook in Pri"acy Breach,"
tendency to demonstrate its rightness and show that http://online.wsj.com/article/SB1000142405270
it's feasible by giving little seat checking comes about. 2304772804575558484075236968.html.
By introducing a general relevant structure, we have [6]. "Offshore outsourcing,"
a tendency to present answerableness as right on time http://www.computerworld.com/s/article/1099
as inside the style area of a data exchange foundation. 38/Offshore outsourcing cited in Florida data
Despite the fact that LIME doesn't effectively prevent leak.
data release, it presents responsive answerableness. [7]. A. Mascher-Kampfer, H. Stogner, and A. Uhl,
Along these lines, it'll deflect vindictive gatherings "Multiple re-watermarking scenarios," in
from unseaworthy individual reports and can Proceedings of the 13th International
energize legitimate (however reckless) gatherings to Conference on Systems, Signals, and Image

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


129
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

Processing (IWSSIP 2006). Citeseer, 2006, pp. Cite this article as :


53-56.
[8]. P. Papadimitriou and H. Garcia-Molina, "Data Amith Manda, Dr. R. P. Ramkumar, "Data Extraction
leakage detection," Knowledge and Data in Malignant Environments", International Journal of
Engineering, IEEE Transactions on, "ol. 23, no. Scientific Research in Computer Science, Engineering
1, pp. 51-63, 2011. and Information Technology (IJSRCSEIT), ISSN :
[9]. "Pairing-Based Cryptography Library (PBC)," 2456-3307, Volume 5 Issue 1, pp. 123-130, January-
http://crypto.stanford.edu/pbc. February 2019. Available at doi :
[10]. I. J. Cox, J. Kilian, F. T. Leighton, and T. https://doi.org/10.32628/CSEIT195116
Shamoon, "Secure spread spectrum Journal URL : http://ijsrcseit.com/CSEIT195116
watermarking for multimedia," Image
Processing, IEEE Transactions on, "ol. 6, no. 12,
pp. 1673-1687, 1997.
[11]. B. Pfitzmann and M. Waidner, "Asymmetric
fingerprinting for larger collusions," in
Proceedings of the 4th ACM conference on
Computer and communications security, ser.
CCS ’97, 1997, pp. 151-160.
[12]. S. Goldwasser, S. Micali, and R. L. Ri"est, "A
digital signature scheme secure against adapti"e
chosen-message attacks," SIAM J. Comput., "ol.
17, no. 2, pp. 281-308, 1988.
[13]. A. Adelsbach, S. Katzenbeisser, and A.-R.
Sadeghi, "A computational model for
watermark robustness," in Information Hiding.
Springer, 2007, pp. 145-160.
[14]. J. Kilian, F. T. Leighton, L. R. Matheson, T. G.
Shamoon, R. E. Tarjan, and F. Zane, "Resistance
of digital watermarks to collusi"e attacks," in
IEEE International Symposium on Information
Theory, 1998, pp. 271-271.
[15]. M. Naor and B. Pinkas, "Efficient obli"ious
transfer protocols," in Proceedings of the
Twelfth Annual ACM-SIAM Symposium on
Discrete Algorithms, 2001, pp. 448-457

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


130
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1| ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195119

Diagnosis of Various Thyroid Ailments using Data Mining


Classification Techniques
Umar Sidiq1, Dr. Syed Mutahar Aaqib2, Dr. Rafi Ahmad Khan3
1 Research Scholar, Department of Computer Science, Mewar University, Rajasthan India
2Assistant Professor, Department of Computer Science, Amar Singh College, Srinagar, Jammu and Kashmir,
India
3Assistant Professor, University of Kashmir Srinagar, Jammu and Kashmir, India

ABSTRACT

Classification is one of the most considerable supervised learning data mining technique used to classify
predefined data sets the classification is mainly used in healthcare sectors for making decisions, diagnosis
system and giving better treatment to the patients. In this work, the data set used is taken from one of
recognized lab of Kashmir. The entire research work is to be carried out with ANACONDA3-5.2.0 an open
source platform under Windows 10 environment. An experimental study is to be carried out using classification
techniques such as k nearest neighbors, Support vector machine, Decision tree and Naïve bayes. The Decision
Tree obtained highest accuracy of 98.89% over other classification techniques.
Keywords : Thyroid disease, K-Nearest Neighbor, Support Vector Machine, Decision Tree, Naïve Bayes.

I. INTRODUCTION hormones released from the thyroid gland which is


one of the important organlocated in the front of the
Classification techniques play a vital as well as major neck and below the Adam's apple of our body. The
role in analyzing survivability of diseases and secretion of thyroid hormones from the thyroid gland
providing facilities to reduce the cost to the patients. are of two types i.e. levothyroxine or T4 and
Now-a-days, Disease diagnosis has become very triiodothyronine or T3. These hormones help in
crucial because of occurrence of so many diseases production of balanced amount of proteins,
every year. People from all over the world have been regulating the temperature of body, and maintaining
suffering from various health issues like diabetes, overall production of energy [18]. Thyroid disease
heart disease, typhoid, tuberculosis, kidney disease occurs when thyroid gland stop to functioning
etc [16] [17].Beside these health issues , thyroid properly and are mainly divided into hypothyroidism
disease have also been detected worldwide and thus and hyperthyroidism [4].
become a serious endocrine health problem and an
issue of concern. It is expected that in India about 42 The excess and deficient secretion amount of thyroid
million people suffer from thyroid disorders [2]. As hormone causes hyperthyroidism and
per resent studies, women are 5 to 8 times more hypothyroidism respectively. The common
prone to thyroid disorders than men worldwide. It is symptoms of hyperthyroidism are sudden weight loss,
caused by the improper secretion of thyroid rapid heartbeat, nervousness, etc. and

CSEIT195119| Received :15Jan 2019| Accepted : 25Jan2019 | January-February-2019[ 5 (1) : 131-136]


131
Umar Sidiqet al. Int J Sci Res CSE & IT. January-February-2019 ;5(1):131-136

hypothyroidism has weight gain, tiredness, weakness, neuro fuzzy method emplaced in ESTDD system for
feeling cold etc. One of the most common cause thyroid diseases diagnosis and achieved 95.33% of
occurred due to hyperthyroidism is graves’ disease [3]. accuracy. ESTDD is one of significant tool for
The underestimated thyroid disease causes thyroid endocrinologists or students studying endocrinology
storm and myxedema which may lead to death [12]. for testing their knowledge by comparing their
In this research work, a classification model is predictions with ESTDD[21].
trained using classification algorithms like K-nearest
neighbor (KNN),support vector machine (SVM), In year 2009, Temurtas performed diagnosis of
decision tree (DT) and Naïve bayes (NB) for the thyroid diseases with the help of artificial neural
diagnosis of thyroid diseases. The Decision Tree network with 3-fold and 10-fold cross-validation
outperformed over other techniques. approaches were used to estimate the performance of
the used neural networks. In this work highest
The rest of the paper is followed as: - Section 2 accuracy of 94.81% was achieved [12].
represents related work in diagnosis of thyroid
diseases. Section 3 contains dataset and methods. In year 2011, EsinDogantekin et al. worked on
Section 4 represents the results and discussion. Wavelet Support Vector Machine (WSVM) with
Section 5 contains conclusion and at last references Generalized Discriminant Analysis (GDA) methods
are mentioned. for diagnosis of thyroid diseases and achieved 91.86%
classification accuracy [22].
II. RELATED WORK
K.Saravana Kumar et al in 2014, proposed KNN and
By studying literature, it seems that several SVM classification methods for the diagnosis of
classification methods have been developed for thyroid disease. They showed that the prediction
thyroid diseases diagnosis. In year 2002, Ozyilmaz et accuracy of KNN and SVM are 96.34% and 94.43%
al in [14] showed that feed forward neural network respectively [1].
could be successfully used for diagnosis of thyroid
diseases. In this work, three architectures of MLP, In year 2017 MP Gopinath proposed SVM, FKNN
RBF and CSFNN were used and achieved accuracy and Decision tree for the diagnosis of thyroid diseases.
89.80%, 79.08% and 91.14% respectively. This work The SVM obtained highest classification accuracy of
shows CSFNN gives the best classification accuracy 96.30% over other techniques [23].
and takes less training times than all other algorithms.
III. DATASET AND METHODS
In 2007, Polat et al. studied on the artificial immune-
recognition system (AIRS) classification method for The dataset used in this investigate work is a clinical
diagnosis of thyroid diseases and achieved 81% dataset. The dataset was taken from one of the
classification accuracy. On the other hand AIRS with leading diagnostic lab in Kashmir. The dataset
Fuzzy weighted pre-processing was classified and contains the record of 807 patients of almost all age
showed the accuracy of 85.00% .This means AIRS groups. Out of 807 patients (224 Males and 583
with fuzzy outperformed than AIRS [15]. Females) 553 belongs to normal, 218 belongs to
hypothyroidism and 36 belongs to hyperthyroidism.
In year 2008, keles et al. proposed an expert system The dataset has 6 attributes as: age, gender, TSH, T3,
for thyroid disease diagnosis (ESTDD) and using T4 and added classification attribute for indication of

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


132
Umar Sidiqet al. Int J Sci Res CSE & IT. January-February-2019 ;5(1):131-136

normal or hyperthyroidism or hypothyroidism. Table hyper plane is generally done, that has largest
1 shows the description of dataset. distance to the closest training data point of any class
(so-called functional margin), since in general the
greater the margin the smaller the generalization
Table 1. Attributes used in our dataset. error of the classifier [19].

Serial Attribute description Value Decision tree


No. name Decision Tree (DT) is tree like graph known as one of
1 Age Age in years Numeric the most admired classification data mining
technique that splits the dataset into parts on the
2 Gender M-Male Nominal
decisions [5]. In decision tree, each internal node or
F- Female
non-leaf node represents a test on a particular
3 TSH Continuous Numeric
attribute, each branch denotes the outcome of that
4 T3 Continuous Numeric test, and each leaf node has a class label. The paths
through which a particular test data is to classify
5 T4 Continuous Numeric from root to leaf represent classification rules based
6 Results Normal Nominal on maximum information gain [6].
Hyperthyroidism
Hypothyroidism Naïve Bayes
Naïve Bayes (NB) is a simple classification algorithm
K Nearest Neighbour for predictive modeling with clear semantics,
K-nearest neighbour (KNN) is a supervised technique representing and the probabilistic learning method
as well as non-parametric in nature. The input of K- based on Bayesian theorem [13]. Naive Bayes
NN depends on the K closest instances present in the classifier assumes that the value of one attribute is
feature space. The generated output depends on not dependent on the value of any other attribute,
whether KNN is Classification or regression methods and it assumes that the presence or absence of
[7] [8].When prediction is required for undetected particular attribute doesn’t affect the prediction
data instances, the KNN algorithm will search process. Suppose there are m classes say C1, C2….Cn
through the training data instances for the k-most having a unidentified data sample X, Naive Bayesian
similar instances. The prediction attribute of the most classifier will predict an unknown sample X to the
similar instances is summarized and returned as the class Ci on the basis of class having highest
prediction for the undetected instance [19]. probability [20].
P( Ci |X > P( Cj |X ) for 1≤j≤m , j ≠ i
Support Vector Machine
Support vector machine (SVM) is a supervised Anaconda
learning classification technique that are used to For implementation of our methods we used
analyzethe data for regression and classification ANACONDA3-5.2.0 64 bit a free and open source
methods [9].It constructs an optimal hyper plane in platform distribution of python and R programming
a high- or infinite-dimensional space in which new language with number of modules, packages and
examples are assigned to one group or the other libraries that provides multiple ways of achieving
one[10].The separation of data is achieved by the classification problems. ANACONDA can be
downloaded from the website [11].

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


133
Umar Sidiqet al. Int J Sci Res CSE & IT. January-February-2019 ;5(1):131-136

In this work, Out of four classifiers, The Decision


Tree shows highest test accuracy of 98.89% over
other classifiers. The accuracy comparison among
K- Fold cross- validation classifiers is shown in figure 2. The highest accuracy
In k-fold validation, the whole dataset is divided into of our method is higher than some of the methods
K equal size subsets and one of the subset K is taken available in literature as shown in table 2.
as test data and the remaining K-1 folds acts as
training data. Thus different test results exist on each
iteration and at last average of these results gives the
test accuracy of the algorithm [12]. In this study 10-
fold cross-validation is used to find out the
classification accuracy by the classification methods.

System implementation
The thyroid dataset has three output categories. First
one is Normal, second one is hypothyroidism and
third one is hyperthyroidism. The thyroid dataset of
five input attributes Age, Gender, TSH, T3 and T4 is
Figure 2. Accuracy comparison of classifiers
supplied to the classifiers of KNN, SVM, Decision
Tree and Naïve bayes in PYTHON to classify the data.
Study Method Highest
The performance of each classifier is evaluated in
Accuracy
terms of accuracy as shown in figure 1.
Ozyılmaz and MLPNN with BP 91.14%
Yıldırım [14 ] (3×FC)
Data mining MLPNN with FBP
classification Performance
techniques measure (3×FC)
RBF
KNN (3×FC)
CSFNN
Thyroid Accuracy
dataset SVM (3×FC)
Polat et al. [15 ] AIRS 85.00%
Decision
Tree
(10×FC)
AIRS with Fuzzy
(10×FC)
Naïve
Bayes weighted
preprocessing
keles et al. [21] ESTDD 95.33%
F. Temurtas [12 ] MLNN with LM 94.81%
Figure 1. Thyroid disease prediction process (3×FC)
PNN (3×FC)
IV. RESULTS AND DISCUSSION LVQ (3×FC)

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


134
Umar Sidiqet al. Int J Sci Res CSE & IT. January-February-2019 ;5(1):131-136

MLNN with LM
(10×FC) [1].K.Saravana Kumar, Dr. R. ManickaChezian,
PNN "Support Vector Machine and K- Nearest Neighbor
(10×FC) Based Analysis for the Prediction of Hypothyroid.
LVQ International Journal of Pharma and Bio
(10×FC) Sciences",volume - 2,Issue - 5,page no-(447-453),2014
EsinDogantekinet GDA-WSVM 91.86% .
al.[22] [2].http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3
MP SVM 96.30% 169866/(accessed dec 2015)
Gopinath.[23] FKNN [3].G. Zhang, L.V. Berardi, An investigation of neural
Decision Tree networks in thyroid function diagnosis, Health Care
Our study KNN 98.89% Manage. Sci. (1998)
(10×FC) [4].Xia C, Hsu W (2006) BORDER: efficient
computation of boundary points. In: IEEE, 2006
SVM
(10×FC) [5].Available from: http://en.wikipedia.org.Last
DT accessed on Dec24].
(10×FC) [6].Apte & S.M. Weiss, Data Mining with Decision
NB Trees and Decision Rules, T.J. Watson Research
(10×FC) Center,
http://www.research.ibm.com/dar/papers/pdf/fgcsapt
Table-2 shows the highest classification accuracy ewe issue_with_cover.pdf, (1997).
obtained by our method over other methods used for [7].Roychowdhury S (2014) DREAM: diabetic
the diagnosis of thyroid diseases. retinopathy analysis using machine learning. In:
IEEE, 2014
V. CONCLUSION [8].Chetty N, Vaisla KS, Patil N (2015) An improved
method for disease prediction using fuzzy approach.
The work has been done using classification data In: IEEE, 2015
mining techniques for the diagnosis of thyroid disease. [9].S. Sathiya Keerthi, Olivier Chapelle, Dennis
For this purpose, K nearest neighbor, Support vector DeCoste "Building Support Vector Machines with
machine, Decision tree and Naive Bayes classifiers Reduced Classifier Complexity" Journal of Machine
have been used. The Decision Tree classifier Learning Research, Vol: 7, PP 1493- 515, January -
outperformed over other classifiers. However, if we (2006).
merge it with any other classification technique such [10].Shen X, Lin Y (2004) Gene expression data
as neural network, then the result might be even classification using SVM-KNN classifier". In: IEEE,
better as compared to what we got with the current 2004
study. [11].www.anaconda.com.
[12].F. Temurtas, "A comparative study on thyroid
disease diagnosis using neural networks," Expert
Systems with Applications, vol. 36, 2009, pp. 944-
949.
VI. REFERENCES

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


135
Umar Sidiqet al. Int J Sci Res CSE & IT. January-February-2019 ;5(1):131-136

[13].G. H. John and P. Langley, "Estimating International Journal of Pure and Applied
Continuous Dis-tributions in Bayesian Classifiers," Mathematics Volume 117 No. 7 2017, 53-63.
Proceedings of the 11th Conference on Uncertainty Cite this article as :
in Artificial Intelligence, San Francisco, 1995, pp.
338-345. Umar Sidiq, Dr. Syed Mutahar Aaqib, Dr. Rafi
[14].Ozyılmaz, L., Yıldırım, T. (2002). Diagnosis of Ahmad Khan, "Diagnosis of Various Thyroid
thyroid disease using artificial neural network Ailments using Data Mining Classification
methods. In Proceedings of ICONIP’02 9th Techniques", International Journal of Scientific
international conference on neural information Research in Computer Science, Engineering and
processing (pp. 2033-2036). Singapore: Orchid Information Technology (IJSRCSEIT), ISSN : 2456-
Country Club. 3307, Volume 5 Issue 1, pp. 131-136, January-
[15].Polat, K., Sahan, S., & Gunes, S. (2007). A novel February 2019.
hybrid method based on artificial immune Available at doi :
recognition system (AIRS) with fuzzy weighted https://doi.org/10.32628/CSEIT195119
preprocessing for thyroid disease diagnosis. Expert Journal URL :http://ijsrcseit.com/CSEIT195119
Systems with Applications, 32, 1141-1147.
[16].Sehgal MSB, Gondal I (2014) K-ranked
covariance based missing values estimation for
microarray data classification. In: IEEE, 2004
[17].Bonner A (2004) Comparison of discrimination
methods for peptide classification in tandem mass
spectrometry. In: IEEE, 2004
[18].HalifeKodaz et al. Medical application of
information gain based artificial immune recognition
system (AIRS): Diagnosis of thyroid disease.
[19].Joel Jacob et al. "Diagnosis of Liver Disease Using
Machine Learning Techniques". (IRJET) Volume: 05
Issue: 04 | Apr-2018
[20].K. Pavya et al. "Diagnosis of Thyroid Disease
Using Data Mining Techniques: A Study". (IRJET)
Volume: 03 Issue: 11 | Nov -2016.
[21].Keles, A., and Keles, A., ESTDD: Expert system
for thyroid diseases diagnosis. Expert Syst. Appl.
34(1):242-246, 2008.
[22]. Dogantekin, E., Dogantekin, A., and Avci, D.,
An expert system based on generalized discriminant
analysis and wavelet support vector machine for
diagnosis of thyroid diseases. Expert Syst. Appl.
38(1):146-150, 2011.
[23].M.P.Gopinath "Comparative Study on
Classification Algorithm for Thyroid Data Set".

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


136
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195120
Analysis of An Indoor Air Quality Antenna for Asthma
Management in Children
N. Vijayalakshmi1, J. G. Rohini2
1 Assistant Professor, Department of Computer Science, Bon Secours College for Women, Bharathidasan
University, Thanjavur, Tamil Nadu, India
2M. Sc Computer Science, Bon Secours College for Women, Bharathidasan University, Thanjavur, Tamil Nadu,
India
ABSTRACT

The extensive of diseases related with Indoor Air Quality (IAQ) usually referred to as Sick Building Syndrome
(SBS) and Building Related Illness (BRI) has generated concern for technological approach in monitoring and
analysis of the quality of air in confine environments. In line with “prevention is better than cure” mantra, such
monitoring system is necessary for timely awareness of quality of surrounding air so as to take preventive
action not only at the individual level, but also aid prompt diagnosis by the medical personnel. Motivated by
the need for an affordable, simple and efficient indoor air quality device for general use at homes, offices and
other related confine environment, a sensor node with capability for wireless sensor networking is presented in
this study. The proposed device incorporates: sensors to measure major air quality indices: odor, carbon
monoxide, dust, humidity, and temperature, LED indicators for prompt visual status alert, memory device (SD
card) for off-line data access, and wireless capability for remote data transfer in applications such as web and
mobile-apps interfacing. The performance of the system is evaluated in a laboratory scale indoor environment.
The results show effectiveness of the system in capturing the air-quality status and providing timely
information for proactive action.
Keywords : Indoor Air Quality, Sick Building Syndrome, Building Related Illness, wireless capability, mobile-
apps.
I. INTRODUCTION has called for a pro-active solution proportionate to
the advances in the related technological. SBS and
The alarming rate of indoor-air quality related BRI are related and can be defined as sickness and
diseases commonly referred to as Sick Building discomfort experience by occupants due to poor
Syndrome (SBS) and Building Related Illness (BRI) indoor air environment with no specific causes

CSEIT195120 | Received : 01 Feb 2019 | Accepted : 10 Feb 2019 | January-February -2019 [ 5 (1) : 137-142]
137
N. Vijayalakshmi, J. G. Rohini Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 137-142

identify but related to amount of time spent inside. was explored and analyzed in comparison to
Symptoms such as headaches, sore throats, itchy eyes, acceptable standard set by ASHRAE [7]. In [8], the
breathing difficulties, lung cancer, asbestosis, indoor air parameters variation with respect to
mesothelioma, asthma, skin irritation and allergic microclimate condition and impact of outdoor
reactions are among the reported health effect of environment was investigated and reported with
poor air environment according to medical report reference to large museum exhibition halls for a
and study [1-6]. According to the World Health period of one year. The study showed the differences
Organization (WHO), 4.3 million people a year die in indoor microclimate in air-conditioned rooms and
from the exposure to household air pollution [4]. The non-air conditioned ones in the same building. In
problems are been compounded with the increasing addition, external conditions were reported to have a
amount of time people spend in a confine significant impact on the indoor air parameters. On
environment - either in home or in offices or in car, the other hand, at the system level, the pace of
or malls or in school-, and the increasing potential of research work on monitoring systems for IAQ is
air pollution due to cooking using solid fuels, driven by the dynamic rate of sensors and computing
smoking in and around buildings, poor ventilation, technologies advancement. Several studies have been
dust, humidity and temperature. For instance, it has reported in this direction [9-15]. A wireless
been reported that people spend a substantial monitoring system incorporating temperature,
proportion of their time in buildings [5]. In the arid humidity, CO and CO2 sensors for HVAC control
region (MENA) most people spend 90% of their time and IAQ monitoring is presented in [13]. An
indoors [1,2]. In response to these problems, research ATMega88 with a transceiver was adopted in the
studies have been directed to (i) qualitative study for the implementation, sample performance
monitoring and analysis of indoor air quality, results was presented to justify the success of the
including guidelines on IAQ [4] and (ii) development study. Sneha and Sandeep [10] presented an
of sensor node (monitoring system) for quantitative embedded system model for air quality monitoring
monitoring of the IAQ for several applications. For incorporating a LCD display unit to help an allergic
instance, assessment of indoor air quality pollutants patient aware of its environment. No real-life data
in the offices and their potential negative effect in results presented in the study. Hybrid sensor system
Dubai were carried out in [1] using both quantitative with both stationary and mobile sensors is reported
and qualitative techniques. In that study, the in [15] for accurate tracking of air pollutant.
concentration levels of specific indoor air quality Meanwhile, apart from the fact that no single system
indicators (e.g. TVOC, Ozone, Carbon Monoxide, exists that addresses all the pollutants nor
Carbon Dioxide, Relative Humidity and Temperature) incorporates all the performance requirements for a

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


138
N. Vijayalakshmi, J. G. Rohini Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 137-142

given situation and environment, most of the IAQ designated as follows: Green: indicates clean, Yellow:
devices are either expensive, thereby hindering indicates moderate, and Red: indicates danger. This is
affordability for general use, or lack simplified data to provide easy and prompt status information for the
presentation for a common end users. Hence, the end users for timely proactive action
current study proposes a wireless sensor node for
IAQ monitoring with simplicity of indication Apart from the cost and power consumption
(situation alert), affordability, easy of modification which are aimed to be low cost and low power,
and maintenance, and yet provides opportunity of respectively, the choice of the sensors stated above is
being fixed located at a place or used as a mobile based on characteristic of the local environment, and
hand-held device (i.e. mobility option). In addition, it common potential pollutant-generator indoor
is aimed to serve as a base for further research on the activities such as
ongoing IAQ data monitoring and analysis, and IAQ • Tobacco smoking,
wireless sensor networks applications suitable for the • Traditional charcoal-based air freshener
emerging Internet of Things (IoT). • Dust (penetration from outside dusty air, poorly
maintained HVAC systems, etc.)
The subsequent sections provide the essential of IAQ • Cooking gas leakage
followed by the design and development of the • Sewage pipe leakage
proposed system, result, discussion and further study. • Ambient thermal influence
The conclusion of the study is presented at the last
section.

II. METHODOLOGY

The simplified block and components interfacing


diagrams of the proposed IAQ Wireless Sensor Node
(WSN) are shown in Fig. 1. The system comprises of
five air parameter sensors - Carbon monoxide (CO),
Methane (CH4), dust, humidity and temperature Fig 1 : Simplified Block and Components Interfacing

sensors-, microcontroller board for processing, LEDs Diagrams.

display for status alert, micro SD card for data


capturing, Wi-Fi for wireless interface capability and Several compounds constitute air pollutants once

power supply. The three LEDs status alert are their concentration exceeded the Occupational

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


139
N. Vijayalakshmi, J. G. Rohini Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 137-142

Exposure Level (OEL) specified by appropriate Thermal anemometer, Direct-reading dust sampler,
regulations/standards. The types and composition of and microbial sampling [16]. It is noted that, (i) there
the indoor air pollutants depends on ambient outside is no single instrument to sample all the pollutants
air condition and nature of the confined environment which necessitates the need for localized device for a
(be it residential, office, markets/malls, sport center given situation, (ii) some of the testing and sampling
etc.). methods are only interpretable by IAQ professionals,
thereby limits their direct adoption and benefits to
Essentially for a given indoor environment, the IAQ the occupants (general users) and (iii) affordability in
issues involved identification of potential pollutants, terms of cost still remains an obstacle towards
their sources, health concerns, standards/regulations widespread adoption of IAQ monitoring system for
in respect of their OEL, monitoring and preventive preventive maintenance and safety. These issues have
measures to minimize human exposure, and initiating led to dynamic research and development of different
corrective action when necessary. Among the major types of air testing/sampling instruments to meet the
and common parameters in IAQ are Carbon demands for timely and informative monitoring of
monoxide, Carbon dioxide, Dust/particulates, Volatile the ambient indoor air condition in line with the
Organic Compounds (VOCs), Formaldehyde, state-of-the-art technology.
Methane, and temperature/Humidity for thermal
comfort. Others are Benzene, Nitrogen dioxide, III. RESULTS, DISCUSSION AND FUTURE STUDY
Sulfur dioxide, Radon etc. [16]. Table 1 gives the
summary of the sources, effects and standards of the The results of both web data transmission and real-
common IAQ parameters while the detailed can be time data monitoring show the successful
obtained in the literatures [4, 16]. implementation of the proposed IAQ-WSN. The web
data interface of Fig. 1 provides the detail
With respect to the management of IAQ issues (such information on the status of the air parameters and
as monitoring, prevention and control) the core working condition of the device. This is aimed at
requirement is air quality sampling and testing with providing remote data access for occupants and
the aid of appropriate instruments. Specific health officers for timely preventive action. The
procedures have been proposed and reported in the results of Fig. 5 show the effectiveness of the node in
sampling and measurements of different types of air monitoring the pollutant in real-time. All the
contaminants. The common types of instrument are: pollutants data were captured almost at the same time,
direct-reading Colorimetric Tubes, Direct-reading (around 6000seconds) based on the simultaneous
Portable monitors, Air samplers, Smoke tube, injection of the materials in the enclosure. This

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


140
N. Vijayalakshmi, J. G. Rohini Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 137-142

provides a qualitative performance of the design and [2]. Jennifer Bell, “Sick buildings are leading to
development of the proposed IAQ-WSN. Future sick UAE office workers, doctors say”, The
study is expected to conduct quantitative analysis and Nation, UAE, Spetember 12, 2016.
benchmark the system performance with standard / http://www.thenational.ae/uae/health/sick-
industrial IAQ sensors. In addition, network of the buildings-are-leading-to-sickuae- office-
nodes is to be developed with appropriate routing to workers-doctors-say. accessed date: 1/10/2017.
cover large indoor environment or complete building. [3]. Khawaja Mohammad Hassan, Sick Building
Syndrome: The killer within, The Khaleej
IV. CONCLUSION times,
http://www.khaleejtimes.com/article/2011011
The design and development of a wireless sensor 5/ARTICLE/3011 59995/1002, Filed on
node for Indoor Air Quality monitoring system was January 16, 2011, accessed date: 10/12/2017.
presented in this project. The evaluation results of [4]. World Health Organization (WHO),
both web data transmission and real-time data “Guidelines for Indoor Air Quality: Selected
monitoring show the successful implementation of Pollutants”.WHO,2010.http://www.euro.who.
the proposed IAQ-WSN. Future study is expected to int/__data/assets/pdf_file/0009/128169/e94535.
conduct quantitative analysis and benchmark the pdf.
system performance with standard / industrial IAQ [5]. HS. Brightman and N. Moss. Sick building
sensors. In addition, network of the nodes is to be syndrome studies and the compilation of
developed with appropriate routing to cover large normative and comparative values. Indoor Air
indoor environment or complete building. Quality Handbook. Editors: J. D. Spengler, J. F.
McCarthy, and J. M. Samet, New York:
V. REFERENCES McGraw-Hill, 2001.
[6]. US Environmental Protection Agency
[1]. Hanan Ahmad Ibrahim, Indoor Air Quality in (USEPA). Indoor air quality tools for schools
UAE Office Buildings and Their Effects on Communications guide. http://www.epa.gov,
occupants’ Health, Comfort, Productivity and accessed date: 10/12/2017
Performance”, Master thesis, Faculty of [7]. ASHRAE: The American Society of Heating,
Engineering and IT MSc Sustainable Design of Refrigerating and Air- Conditioning
the Built Environment, The British University Engineers. https://www.ashrae.org. accessed
in Dubai, May 2015. date: 10/12/2017.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


141
N. Vijayalakshmi, J. G. Rohini Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 137-142

[8]. Joanna Ferdyn-Grygierek Monitoring of Electrical, Electronics and System Engineering


indoor air parameters in large museum (ICEESE), Pages: 60 - 64.
exhibition halls with and without air- [13]. R. du Plessis, A. Kumar; GP Hancke and BJ
conditioning systems, Building and Silva, (2016). “A wireless system for indoor air
Environment Volume 107, October 2016, quality monitoring”, IECON 2016 - 42nd
Pages 113-126, Elsevier 2016. Annual Conference of the IEEE Industrial
[9]. Shaharil Mad Saad, Abdul Rahman Mohd Electronics Society, Pages: 5409 - 5414.
Saad, Azman Muhamad Yusof Kamarudin, [14]. Jung-Yoon Kim, Chao-Hsien Chu, and Sang-
Ammar Zakaria, and Ali Yeon Md Shakaff, Moon Shin, (2014). “ISSAQ: An Integrated
“Indoor Air Quality Monitoring System using Sensing Systems for Real- Time Indoor Air
Wireless Sensor Network (WSN) with Web QualityMonitoring”, IEEE Sensors Journal ,
Interface” proceeding International Volume: 14, Issue: 12,Pages: 4230 - 4244.
conference of electrical, electronics and [15]. Yun Xiang, Ricardo Piedrahita, Robert P.
system engineering, (ICEESE) Malaysia, 2014. Dick, Michael Hannigan, Qin Lv, and Li
[10]. Sneha Jangid and Sandeep Sharma, (2016). “An Shang, (2013). “A Hybrid Sensor System for
embedded system model for air quality Indoor Air Quality Monitoring”,2013 IEEE
monitoring”, 2016 3r International Conference International Conference on Distributed
on Computing for Sustainable Global Computing in Sensor Systems, Pages: 96 - 104.
Development, Pages: 3003 - 3008. [16]. Indoor Air Quality Tool Kit, (2009).
[11]. Tsang-Chu Yu, Chung-Chih Lin,Chun-Chang Government of Alberta, Employment and
Chen, Wei-Lun Lee, Ren- Guey Lee, Chao- Immigration.
Heng Tseng, and Shi-Ping Liu, “Wireless
sensor networks for indoor air quality
monitoring” Medical Engineering & Physics, Cite this article as :
Volume 35, Issue 2, February 2013, Pages 231-
235. N. Vijayalakshmi, J. G. Rohini, "Analysis of An
[12]. Shaharil Mad Saad, Abdul Rahman Mohd Indoor Air Quality Antenna for Asthma
Saad, Azman Muhamad Yusof Kamarudin, Management in Children", International Journal
Ammar Zakaria, Ali Yeon Md Shakaff, (2013).
of Scientific Research in Computer Science,
“Indoor air quality monitoring system using
Engineering and Information Technology
wireless sensor network (WSN) with web
interface”,2013 International Conference on

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


142
N. Vijayalakshmi, J. G. Rohini Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 137-142

(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1,


pp. 137-142, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195120
Journal URL : http://ijsrcseit.com/CSEIT195120

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


143
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195132
Home Automation using Google Voice Assistance with
Raspberry Pi
Undra Narasimha Babu1, S. N. Sreekanth2
1PG Scholar, Department of ECE, Siddartha Educational Academy Group of Institutions, Gollapalli, Andhra

Pradesh, India
2Associate Professor, Department of ECE, Siddartha Educational Academy Group of Institutions, Gollapalli,
Andhra Pradesh, India

ABSTRACT

Home automation popular in day to day to number of advantages. In previous and current we monitor the
home appliances remotely by emails or text messages. This paper gives an innovative idea of controlling the
home appliance using with Google voice assistance by using the Raspberry Pi. As a server system Internet of
Things (IoT) links smart objects to the internet. It can enable exchange of data in a more secure way. Rapid
adaptations of Smart City and Internet of Things technologies are assisting in urban planning to ensure
sustainable cities and lifestyles. Life is made simpler and more productive through IOT. As the Wi-Fi hotspots
are increasingly becoming common in the recent times, the existing infrastructure can be used to develop a
cost-effective solution to enable the existing appliances with IOT Raspberry Pi is single board computer (SBC)
is proposed in this paper to Raspberry pi known as a Single Board Computer (SBC) is proposed in this paper to
reach these goals. The voice commands will give to Raspberry PI using Google voice assistance. These
commands will control the any appliances or devices at home. This will provide a better communication in
automated home as compared to normal homes.
Keywords: IoT, WI-FI, Raspberry Pi.

I. INTRODUCTION there is a need to robotize everything, so the idea of


brilliant home mechanization framework is a thought
In this cutting-edge time, computerization of which is utilized to make the city shrewd. A Smart
everything is the need of great importance Home is one that gives solace, security and gives the
Automation is the utilization of control frameworks sentiment of home to house individuals. Shrewd
and data innovation to manage gear, modern homes likewise give Energy proficiency (low
hardware, and procedures, limiting the requirement working expense) and comfort consistently, for each
for human inclusion. Computerization assumes an person at home. Home computerization implies the
inexorably imperative job in the worldwide economy checking and control of family unit protests
and in step by step experience. Architects work to insightfully for powerful use. The family questions
connect mechanized gadgets with scientific and ought to be astutely interconnected and also give data
hierarchical devices to make complex frameworks for to better tasks. Home computerization increased with
quickly growing parameters of uses and human the Internet of Things (IoT) gives better adaptability
exercises. For the improvement of keen urban areas, in overseeing and controlling family unit questions in

CSEIT195132 | Received : 15 Jan 2019 | Accepted : 30 Jan 2019 | January-February -2019 [ 5 (1) : 143-146 ]
143
Undra Narasimha Babu, S. N. Sreekanth et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 143-146

a more extensive perspective. This will bolster the used as a piece of equipment wanders, and for an
interconnectivity of countless homes for better asset extensive part of the things that your work zone PC
usage in more extensive zone. does, like spreadsheets, word-getting ready and
diversions. It likewise plays superior quality video.
II. EXISTING SYSTEM We need to see it being utilized by kids everywhere
throughout the world to figure out how PCs
In the existing system, home appliances are function, how to control the electronic world around
controlled through Bluetooth, Zigbee and other them, and how to program.
devices. But they are limited to certain area. To
overcome this problem, we are moving to the
proposed system.

III. PROPOSED SYSTEM

In the proposed system we are control home


appliances using Raspberry Pi with Google assistance
in mobile. We can give command to the Google
assistance according to that command’s loads will
turn on or turn off. The user can have access to Figure 2 : Raspberry Pi3 module
operate the loads from anywhere in the world.

IV. METHODS AND MATERIAL

BLOCK DIAGRAM

Figure 3 : Pin layout of raspberry Pi GPIO used in


system

Figure 1 : block diagram of the system d) Motor


Electrical DC Motors are continuous actuators that
HARDWARE REQUIREMENTS convert electrical energy into mechanical energy.
The DC motor achieves this by producing a
a) Raspberry Pi 3 continuous angular rotation that can be used to rotate
pumps, fans, compressors, wheels, etc.
The Raspberry Pi is an ATM card sized PC that
fittings into your TV. It is a fit little PC which can be

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


144
Undra Narasimha Babu, S. N. Sreekanth et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 143-146

SOFTWARE COMPONENTS

a) Raspbian OS

Raspbian OS is such of the no buts about it operating


systems at hand without electronic commerce a dime
to transform and use. The apparatus is chiefly based
on Debian Linux and is optimized to employment
effectively by the whole of the Raspberry Pi laptop.
Figure 4 : DC motor
As we erstwhile know an OS is a reside of integral
packages and utilities that murmur a indisputable
b) Power Supply
hardware, in this position the Pi. Debian could be
literally lightweight and makes a amazing in a class
all by itself for the Pi. The Raspbian includes tools for
surfing, python programming and a GUI computing
device.

WORKING PRINCIPLE

We are using Raspberry PI to control the appliances


in the home through wireless Communication. The
Appliances in the Home are controlled through
Google assistance through the Wi-Fi Connectivity.
Figure 5 : Flow chart of power supply ON/OFF instructions are given through Wi-Fi
present in the user Mobile, the Data Transfers from
c) Relay mobile phone to Module and then instruction is
further given to the Controller and thus the loads are
Relay is electromagnetic switch that open or close the operated.
switches electrically or electromechanically. Relay is
mostly used to switch smaller circuits. V. RESULTS AND DISCUSSION

APPLICATIONS

➢ Used in homes,
➢ offices,
➢ industrial areas etc.,

ADVANTAGES
.
➢ Smart and secure.
Figure 6 : Relay ➢ Controlling home appliances from anywhere
in the world.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


145
Undra Narasimha Babu, S. N. Sreekanth et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 143-146

RESULTS VI. CONCLUSION

Home automation integrates any electrical devices


and in houses each other. In this paper we
implemented home appliances (fan, electric tubes,
refrigerator and washing machine) control using
Google voice assistance with help of raspberry
Pi.Development of such Smart Home achieve by
using Internet of Things technologies. This system is
low cost and more reliable.

VII.REFERENCES

Figure 7 : Hardware design of the project


[1]. HOME AUTOMATION USING ATmega328
MICROCONTROLLER AND ANDROID
APPLICATION, S.Anusha1, M.Madhavi2,
R.Hemalatha3. International Research Journal
of Engineering and Technology (IRJET),
Volume: 02 Issue: 06 | Sep-2015 www.irjet.net
[2]. Internet of Things: Ubiquitous Home Control
and Monitoring System using Android based
Smart Phone, Rajeev Piyare. International
Journal of Internet of Things 2013, 2(1): 5-11
Figure 8 : Google voice assistance for light on DOI: 10.5923/j.ijit.20130201.02
[3]. Design and Implementation of a WiFi Based
Home Automation System, Ahmed ElShafee,
Karim AlaaHamed. World Academy of Science,
Engineering and Technology International
Journal of Computer, Electrical, Automation,
Control and Information Engineering Vol:6,
No:8,2012.

Cite this article as : Undra Narasimha Babu, S. N.


Figure 9 : Google voice assistance for light off Sreekanth, "Home Automation using Google Voice
Assistance with Raspberry Pi", International Journal
to keep the city clean by informing about the garbage of Scientific Research in Computer Science,
levels of the bins by providing graphical image of the Engineering and Information Technology
bins via IOT Php web development platform. (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
143-146, January-February 2019. Available at doi :
https://doi.org/10.32628/CSEIT195132
Journal URL : http://ijsrcseit.com/CSEIT195132

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


146
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195130

Sentiment Analysis of movie reviews using Microsoft Text


Analytics and Google Cloud Natural Language API
S Jaswanth Sai Chowdary, Sk Waseem Farooq, Sk Abubakar Siddique, S Dinesh
Department of CSE, Vasireddy Venkatadri Institute of Technology, Guntur, Andhra Pradesh, India

ABSTRACT

Sentiment analysis is focused on the extraction of emotions and opinions of the people towards a particular
topic from a structured, semi-structured or unstructured textual data. In this paper we focus our task of
sentiment analysis on The Movie DB (TMDB) database which is an online database and contains information of
movies, television shows and video games including cast, crew and reviews. We examine the sentiment
expression to classify the polarity of the movie review on a scale of 1(highly disliked) to 5(highly liked) and
perform feature extraction and ranking and use these features to train our multilabel classifier to classify the
movie review into its correct label.We will collect the reviews for various movies and perform classification on
them. The classification for the reviews are done based on score obtained for the text. We are using two
approaches to generate scores for the reviews. These approaches are Microsoft cognitive services and Google
cloud natural language API.Finally we are able to generate the scores and can classify them based on movie
reviews.
Keywords : Sentiment Analysis, TMDB database, Microsoft cognitive services, Google cloud natural language
API, classification

I. INTRODUCTION websites are not formal reviews but are rather very
informal and are unstructured form of grammar.
Now a days Internet has become a huge Cyber Opinions expressed in movie reviews give a very true
Database which hosts gigantic amount of data which reflection of the emotion that is being conveyed. The
is created and consumed by the users. The database presence of such a great use of sentiment words to
has been growing at an exponential rate giving rise to express the review inspired us to derive an approach
a new industry filled with it, in which users express to classify the polarity of the movie using these
their opinions across channels such as Rotten sentiment words. Sentiment Analysis is the process of
Tomatoes, Facebook, Twitter, and Foursquare. computationally identifying and categorizing
Opinions which are being expressed in the form of opinions expressed in a piece of text, especially in
reviews provide an opportunity for new explorations order to determine whether the writer's attitude
to find collective likes and dislikes of cyber towards a particular topic, product, etc. is positive,
community. One such domain of reviews is the negative, or neutral. Sentiment Analysis is a
domain of movie reviews which affects everyone technology that will be very important in the next
from writers, users, film critics to the production few years. With opinion mining, we can distinguish
company. The movie reviews being posted on the poor content from high quality content. With

CSEIT195130 | Received : 15 Jan 2019 | Accepted : 30 Jan 2019 | January-February -2019 [ 5 (1) : 147-150 ]
147
S Jaswanth Sai Chowdary et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 147-150

thetechnologies available we can know if a movie has A. Feature Selection


more good opinions than bad opinions and find the Most researchers apply standard feature selection in
reasons why those opinions are positive or negative. their approach to improve computational
Most sentiment analysis studies are now focused on performance with a handful using more sophisticated
social media sources such as TMDB, IMDB [1], approaches. Papers focusing entirely on feature
Twitter [2] and Facebook, requiring the approaches selection to improve sentiment analysis are few. We
be tailored to serve the rising demand of opinions in fetched the review data based on content keyword by
the form of text. Furthermore, performing the eliminating unwanted fields from TMDB [3]
phrase-level analysis of movie reviews proves to be a database. Based on the movie id data related movie is
challenging task. generating which is shown in below figure.

II. METHODS AND MATERIAL

Sentiment analysis aims to determine the attitude of a


speaker or a writer with respect to some topic or the
overall contextual polarity of a document. The
attitude may be their judgment or evaluation,
affective state which is the emotional state while
writing, or the intended emotional communication
which is the emotional effect the writer wishes to
have on the reader. Sentiment is defined as any kind
of emotion and in the context of sentiment analysis is Figure 2 : Movie data send as response by TMDB
the opinion that is being expressed in the form of text database
or speech.
Here details about the author of the movie, id of user
and other unnecessary fields are eliminated. The
review given by user is with key content. The final
reviews are gathered from users by eliminating the
unwanted fields which is shown in the below figure.

Figure 3 : After eliminating unwanted fields reviews


Figure 1 : Process Model Diagram
of each movie

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


148
S Jaswanth Sai Chowdary et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 147-150

B. Microsoft Text Analytics SentiLabel = 4


The Text Analytics is a cloud-based service that Else if SentiScore >=0.4 &&<0.6,then:
provides advanced natural language processing over SentiLabel = 3
raw text, and includes four main functions: sentiment Else if SentiScore >= 0.2 &&<0.4,then:
analysis, key phrase extraction, language detection, SentiLabel = 2
and entity linking. The API is backed by resources Else
in Microsoft Cognitive Services[4], a collection of SentiLabel = 1
machine learning and AI algorithms in the cloud,
readily consumable in your development projects. III.RESULTS AND DISCUSSION

C. Google Cloud Natural Language API


Both the features analyse the reviews data set and
Google Cloud Natural Language [5] reveals the generate the SentiScore. Based on the algorithm
structure and meaning of text both through powerful SentiLabel is determined. Finally the class of the
pre trained machine learning models in an easy to use movie is obtained from the SentiLabel. We use this
REST API and through custom models that are easy scale for defining the class label of the movie reviews
to build with Auto ML Natural Language. You can which is divided as follows:
use Cloud Natural Language to extract information
about people, places, events and much more 1. Sentiment Label 1 belong to class 1
mentioned in text documents, news articles, or blog 2. Sentiment Label 2 belong to class 2
posts. You can use it to understand sentiment about 3. Sentiment Label 3 belong to class 3
your product on social media or parse intent from 4. Sentiment Label 4 belong to class 4
customer conversations happening in a call centre or 5. Sentiment Label 5 belong to class 5
a messaging app. You can analyse text uploaded in
For any movie finally the classification is done based
your request or integrate with your document storage
on the reviews from scale of class 1(highly disliked)
on Google Cloud Storage.
to class 5(highly liked).
D. Classification
We can see that the SentiScore(Sentiment Score) is IV.CONCLUSION
calculated from the two features(Microsoft Text
Analytics and Google Cloud Natural Language API). In this work, we extracted new features that have a
The weight of SentiScore is in the range of 0 to 1. strong impact on determining the polarity of the
Now, once we get the SentiScore[6], we determine movie reviews and applied computation linguistic
the label using the following algorithm: methods for the calculation of sentiment score for the
reviews dataset. We then performed the classification
Algorithm 1. Proposed Algorithm for determining the for the movies by using an algorithm for SentiLabel
label generation.

Initialize SentiScore = 0 In future, we would like to evaluate the effectiveness


SentiScore = value from any of the feature of the proposed sentiment classification features and
If SentiScore >= 0.8 && <=1.0, then: techniques for other tasks, such as sentiment
SentiLabel = 5 classification. We would like to apply in-depth
Elseif,SentiScore>=0.6&&<0.8,then: concepts of Natural Language Processing for better

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


149
S Jaswanth Sai Chowdary et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 147-150

prediction of the polarity of the document. We Cite this article as : S Jaswanth Sai Chowdary, Sk
would also like to extend this technique on other Waseem Farooq, Sk Abubakar Siddique, S Dinesh,
domains of opinion mining like television shows, "Sentiment Analysis of movie reviews using
political discussion forums etc. Microsoft Text Analytics and Google Cloud Natural
Language API", International Journal of Scientific
V. REFERENCES Research in Computer Science, Engineering and
Information Technology (IJSRCSEIT), ISSN : 2456-
[1] Sahu, T. P., & Ahuja, S. (2016). Sentiment 3307, Volume 5 Issue 1, pp. 147-150, January-
analysis of movie reviews: A study on feature February 2019. Available at doi :
selection & classification algorithms. 2016 https://doi.org/10.32628/CSEIT195130
International Conference on Microelectronics, Journal URL : http://ijsrcseit.com/CSEIT195130
Computing and Communications
(MicroCom).DOI:10.1109/microcom.2016.752258
3
[2] Magdum, S. S., &Megha, J. V. (2017). Mining
online reviews and tweets for predicting sales
performance and success of movies. 2017
International Conference on Intelligent
Computing and Control
Systems(ICICCS).DOI:10.1109/iccons.2017.82507
38
[3] https://developers.themoviedb.org/3/movies/get-
movie-reviews
[4] https://azure.microsoft.com/en-
in/services/cognitive-services/text-analytics/
[5] https://cloud.google.com/natural-language/
[6] Topal, K., & Ozsoyoglu, G. (2016). Movie review
analysis: Emotion analysis of IMDb movie
reviews. 2016 IEEE/ACM International
Conference on Advances in Social Networks
Analysis and Mining
(ASONAM).DOI:10.1109/asonam.2016.7752387

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


150
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195137

Blockchain and its Role in the Internet of Things (IoT)


Tanweer Alam
Department of Computer Science, Faculty of Computer and Information Systems, Islamic University of
Madinah, Saudi Arabia
Email: tanweer03@iu.edu.sa

ABSTRACT
Blockchain (BC) in the Internet of Things (IoT) is a novel technology that acts with decentralized, distributed,
public and real-time ledger to store transactions among IoT nodes. A blockchain is a series of blocks, each block
is linked to its previous blocks. Every block has the cryptographic hash code, previous block hash, and its data.
The transactions in BC are the basic units that are used to transfer data between IoT nodes. The IoT nodes are
different kind of physical but smart devices with embedded sensors, actuators, programs and able to communicate
with other IoT nodes. The role of BC in IoT is to provide a procedure to process secured records of data through
IoT nodes. BC is a secured technology that can be used publicly and openly. IoT requires this kind of technology
to allow secure communication among IoT nodes in heterogeneous environment. The transactions in BC could
be traced and explored through anyone who are authenticated to communicate within the IoT. The BC in IoT
may help to improve the communication security. In this paper, I explored this approach, its opportunities and
challenges.
Keywords : Blockchain, Internet of Things (IoT), Cryptography, Security, Communication.

I. INTRODUCTION other and every device has its previous device address.
The blockchain and IoT together work in the
The IoT is growing exponentially year by year with its framework of IoT and Cloud integration. In the future,
aim in 5G technologies, like Smart Homes and Cities, the BC would revolutionize the IoT communication
e-Health, distributed intelligence etc. but it has [1]. The goals of BC and IoT integration could be
challenges in security and privacy. The IoT devices are summarized as follows.
connected in a decentralized approach. So, it is very
complex to use the standard existing security
techniques in the communication among IoT nodes.
The BC is a technology the provide the security in
transactions among the IoT devices. It provides a
decentralize, distribute and publicly available shared
ledger to store the data of the blocks that are processed
and verified in an IoT network. The data stored in the
public ledger is managed automatically by using the
Peer-to-peer topology. The BC is a technology where
transactions fired in the form of a block in the BC
among IoT nodes. The blocks are linked with each
Figure 1 : Blockchains and IoT

CSEIT195137 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 151-157 ]
151
Tanweer Alam Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 151-157

i) Decentralized framework: This approach is similar in exchanging a document with privacy without storing
IoT and BC. It is removed the centralized system and any information on the time-stamping service. The
provide the facility of a decentralized system. It idea of blockchains comes from [3] but the first
improves the failure probability and performance of blockchains were presented by Satoshi Nakamoto in
the overall system. 2008. He presented a paper where the blocks were
ii) Security: In the BC, the transactions among nodes added in a chain and form a blockchain [4]. In the
are secured. It is a very novel approach for secure article [5], the authors presented the “IoTChain” for
communication. BC allows IoT devices to authentication of information exchanged between two
communicate with each other in a secure way. nodes in an IoT network. They have presented an
iii) Identification: In IoT, all the connected devices are algorithm to exchange the information in IoT and
uniquely identified with a unique ID. Every block in blockchains (fig 2) [5]. In this paper, authors are
BC is also uniquely identified. So, BC is a trusted focused on the authorization part of the security in the
technology that provides uniquely identified data IoTChain framework.
stored in public ledger.
iv) Reliability: IoT nodes in BC have the capabilities to
authenticate the information passed in the network.
The data is reliable because it is verified by the miners
before entering in BC. Only verified blocks can enter
in the BC.
v) Autonomous: In BC, all IoT nodes are free to
communicate with any node in the network without
the centralized system.
vi) Scalability: In BC, the IoT devices will Figure 2 : IoT Chain framework
communicate in high-available, a distributed
intelligence network that connects with destination In the article [6], the authors explored the cloud and
device in a real-time and exchange information. MANET framework to connect the smart devices in
the internet of things and provide communication
The rest of the paper is summarized as follows: section security. In the article [7], authors represent a very
1 represents the introduction of the paper, section 2 nice framework called internet-cloud framework, it is
represents the literature survey, section 3 introduces a good idea to provide secure communication to the
the role of BC in IoT, section 4 represents the IoT devices. In the article [8], the authors provide a
opportunities of the integrated approach, section 5 middleware framework in the cloud-MANET
represents the challenges and section 6 represents the architecture for accessing data among the IoT devices.
conclusion. Article [9,10] represents the reliability in the
communication among IoT nodes. The articles
II. LITERATURE SURVEY [11,12,13,14,15] are providing the mobility models for
communication in 5G networks. In the article [16], the
The security and privacy in the communication among fuzzy logic-based mobility framework is explained for
IoT devices paid too much attention in the year of communication security. In the article [17], a nice
2017 and 2018. Several papers are published during the survey on blockchains and IoT done by the researchers.
year 2017 and 2018. In the year of 1990, Stuart Haber They present the idea of the security in the BC-IoT to
and W. Scott Stornetta were written an article [3] on develop the IoT apps with the power of BCs.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


152
Tanweer Alam Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 151-157

III. THE ROLE OF BC IN IoT publicly accessed, privacy is low to high and the
transferable identities while in a centralized database,
The IoT enables the connected physical things to the properties are centralized trust model, low in
exchange their information in the heterogeneous security, low publicly accessed, privacy is high and
network [18]. The IoT could be divided into the non-transferable identities. From the above properties,
following sections. the blockchain is more advanced than the centralized
1. Physical Things: The IoT provide the unique id for storage.
each connected thing in the network. The physical
things are able to exchange data with other IoT nodes.
2. Gateways: The gateways are the devices work
among physical things and the cloud to ensure that the
connection is established and security provided to the
network.
3. Networking: it is used to control the flow of data and
establish the shortest route among the IoT nodes.
4. Cloud: It is used to store and compute the data.
The BC is a chain of verified and cryptographic blocks (a)
of transactions held by the device connected in a
network. The blocks data are stored in the digital
ledger that is publicly shared and distributed. The BC
provides secure communication in IoT network. The
blockchain can be a private, public or consortium with
different properties. The following table represents
the differentiation among all kind of blockchains.

Table 1 : Kinds of Blockchains and their properties (b)


immovableness
Accord growth
BC/ Properties

Decentralized

Determining
Efficiency

Reading

Private good No yes Can Can Only


BC be be one
publi industr
cly y (c)
Public worse Yes no No publi All
Figure 3 : (a) Centralized (b) Decentralized (c)
BC cly miners
Consorti good Some yes Can Can IoT Distributed
um BC times be be devices
publi The following platforms are used to develop IoT
cly
applications using blockchain technology.
The database in blockchains has the properties such as
decentralized trust model, high security, highly

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


153
Tanweer Alam Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 151-157

a. IOTA: The IOTA is the new platform for the devices can communicate in the network and every
blockchain and IoT called Next generation block of the transaction will first verify by the miners
blockchains. This platform facilitates the high data then they can enter in the BC.
integrity, high performance of transactions and high 2. Reduce the Cost: This approach will reduce the cost
validity of blocks with using fewer resources. It because it communicates directly without the third
resolves the limitations of blockchains [19]. party. It eliminates all the third-party nodes between
b. IOTIFY: It provides web-based internet of things the sender and the receiver. It provides direct
solution to minimize the limitations of blockchains communication.
technology in the form of custom apps [20].
c. iExec: It is an open source blockchain based tool. It
facilitates your apps the decentralized cloud Trust

advantages [21].
d. Xage: It is the secure blockchain platform for IoT to Risk Management Cost Reduction

increase automation and secure information [22].


e. SONM: It is a decentralized blockchain based fog
Opportunities
computing platform to provide secure cloud services. in BC-IoT

The IoT and blockchains are increasing the business Financial Services Time Reduction

opportunities and opening the new markets where


everyone or everything can communicate in a real-
time with authenticity, privacy and security in a Social Services
Security and
Privacy

decentralized approach. The integration of these novel


technologies will change the current world where the
devices will communicate without the humans in
Figure 4 : Opportunities in BC-IoT
various stages. The objective of the framework is to get
the secured data on the right location, on the right
3: Reduce Time: This approach is reduced the time a
format, at real-time. The BC could be used to track
lot. It reduces the time taken in transactions from days
billions of IoT connected things, coordinate these
to second.
things, enabling the processing of the transactions,
4: Security and Privacy: It provides security and
resolving or eliminating the failures and making the
privacy to the devices and information.
flexible ecosystem for running the physical things on
5. Social Services: This approach provides public and
it. Hashing techniques are used in blocks of data by BC
social services to the connected devices. All connected
to create information privacy for the users.
devices can communicate and exchange information
between them.
IV. OPPORTUNITIES
6. financial Services: This approach transfer funds in a
secure way without the third party. It provides fast,
The BC-IoT integration approach has a lot of
secure and private financial service. It reduced transfer
remarkable opportunities. It opens the new doors for
cost and time.
both together. Some of the opportunities are described
7. Risk management: This approach is played the
as follows.
important roles to analyze and reduce the risk of
1. Building the Trust between parties: The BC-IoT
failing the resources and transactions.
approach will build trust among the various connected
devices because of its security features. Only verified

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


154
Tanweer Alam Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 151-157

V. CHALLENGES 6. Interoperability: The BC can be public or private. So,


the interoperability between public and private
The IoT and BC could face a lot of challenges such as blockchains is also a challenge in the BC-IoT approach.
scale, store, skills, discover etc. The following are the 7. Rules and Regulation: The IoT-BC will act globally,
challenges faced by the integration approach. so it faces many rules and regulations for
1. Scalability: The BC can become hang because of its implementing this approach globally.
heavy load of the transaction. The Bitcoin storage is
becoming more than 197 GB storage in 2019 [24]. VI. CONCLUSION
Imagine if IoT integrates with BC then the load will be
heavier than the current situation. The BC and IoT is a novel approach explored in this
2. Storage: The digital ledger will be stored on every article. Many opportunities and challenges are
IoT node. By the time, it will increase in its storage size described. Also, available platforms are listed in this
that will be a challenging task and become a heavy article. This approach can be the future of the internet
load on each and every connected device. because it can overhaul the current internet system
3. Lack of Skills: The BC is a new technology. It is and change it with the new one where every smart
known by very few people in the world. So, it is also a device will connect to other devices using the peer-to-
challenge to train the people about the technology. peer network in a real-time. It can reduce the current
cost and time and provide the right information to the
Scalability right device in a real-time. So, it can be very useful in
the future.
Rules and
Storage
Regulation

VII. ACKNOWLEDGMENT

Challenges
in BC-IoT This research is supported by Deanship of Scientific
Interoperability Lack of Skills
Research, Islamic University of Madinah, Kingdom of
Saudi Arabia. The grant number is 10/40.

Discovery and
Privacy
Integration VIII. REFERENCES

Figure 5 : Challenges in BC-IoT [1]. Reyna, Ana, et al. "On blockchain and its
integration with IoT. Challenges and
4. Discovery and Integration: Actually, BC is not opportunities." Future Generation Computer
designed for IoT. It is a very challenging task for the Systems (2018). DOI:
connected devices to discover another device in BC https://doi.org/10.1016/j.future.2018.05.046
and IoT. So, IoT nodes can discover each other but [2]. Zheng, Zibin, et al. "Blockchain challenges and
they can be unable to discover and integrate the BC opportunities: A survey." International Journal
with another device. of Web and Grid Services 14.4 (2018): 352-375.
5. Privacy: The ledger is distributed publicly to every DOI:
connected node. They can see the ledger transactions. https://doi.org/10.1504/IJWGS.2018.095647
So, privacy is also a challenging task in the integrated [3]. Haber, Stuart, and W. Scott Stornetta. "How to
approach. time-stamp a digital document." Conference on

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


155
Tanweer Alam Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 151-157

the Theory and Application of Cryptography. [12]. Alam, Tanweer, Parveen Kumar, and Prabhakar
Springer, Berlin, Heidelberg, 1990. Singh. "SEARCHING MOBILE NODES USING
[4]. Nakamoto, Satoshi. "Bitcoin: A peer-to-peer MODIFIED COLUMN MOBILITY MODEL.",
electronic cash system." (2008). International Journal of Computer Science and
[5]. Alphand, Olivier, et al. "IoTChain: A blockchain Mobile Computing, (2014).
security architecture for the Internet of Things." [13]. Alam, Tanweer, and B. K. Sharma. "A New
Wireless Communications and Networking Optimistic Mobility Model for Mobile Ad Hoc
Conference (WCNC), 2018 IEEE. IEEE, 2018. Networks." International Journal of Computer
[6]. Alam T, Benaida M. The Role of Cloud-MANET Applications 8.3 (2010): 1-4. DOI:
Framework in the Internet of Things (IoT). https://doi.org/10.5120/1196-1687
International Journal of Online Engineering [14]. Singh, Parbhakar, Parveen Kumar, and Tanweer
(iJOE). 2018;14(12):97-111. DOI: Alam. "Generating Different Mobility Scenarios
https://doi.org/10.3991/ijoe.v14i12.8338 in Ad Hoc Networks.", International Journal of
[7]. Alam T, Benaida M. CICS: Cloud–Internet Electronics Communication and Computer
Communication Security Framework for the Technology, 4(2), 2014
Internet of Smart Devices. International Journal [15]. Sharma, Abhilash, Tanweer Alam, and Dimpi
of Interactive Mobile Technologies (iJIM). 2018 Srivastava. "Ad Hoc Network Architecture
Nov 1;12(6):74-84. DOI: Based on Mobile Ipv6 Development." Advances
https://doi.org/10.3991/ijim.v12i6.6776 in Computer Vision and Information
[8]. Alam, Tanweer. "Middleware Implementation Technology (2008): 224.
in Cloud-MANET Mobility Model for Internet [16]. Alam, Tanweer. "Fuzzy control based mobility
of Smart Devices", International Journal of framework for evaluating mobility models in
Computer Science and Network Security, 17(5), MANET of smart devices." ARPN Journal of
2017. Pp. 86-94 Engineering and Applied Sciences 12, no. 15
[9]. Tanweer Alam, "A Reliable Communication (2017): 4526-4538.
Framework and Its Use in Internet of Things [17]. Conoscenti, Marco, Antonio Vetro, and Juan
(IoT)", International Journal of Scientific Carlos De Martin. "Blockchain for the Internet
Research in Computer Science, Engineering and of Things: A systematic literature review."
Information Technology (IJSRCSEIT), Volume Computer Systems and Applications (AICCSA),
3, Issue 5, pp.450-456, May-June.2018 URL: 2016 IEEE/ACS 13th International Conference
http://ijsrcseit.com/CSEIT1835111 of. IEEE, 2016.
[10]. Alam, Tanweer. (2018) "A reliable framework [18]. Gubbi, Jayavardhana, et al. "Internet of Things
for communication in internet of smart devices (IoT): A vision, architectural elements, and
using IEEE 802.15.4." ARPN Journal of future directions." Future generation computer
Engineering and Applied Sciences 13(10), 3378- systems 29.7 (2013): 1645-1660. DOI:
3387 https://doi.org/10.1016/j.future.2013.01.010
[11]. Alam, Tanweer, Arun Pratap Srivastava, [19]. https://www.iota.org
Sandeep Gupta, and Raj Gaurang Tiwari. [20]. https://iotify.org
"Scanning the Node Using Modified Column [21]. https://iex.ec/overview
Mobility Model." Computer Vision and [22]. https://xage.com
Information Technology: Advances and [23]. https://www.i-scoop.eu/blockchain-distributed-
Applications 455 (2010). ledger-technology/blockchain-iot

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


156
Tanweer Alam Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 151-157

[24]. https://www.statista.com/statistics/647523/worl
dwide-bitcoin-blockchain-size

Cite this article as :


Tanweer Alam, “Blockchain and its Role in the Internet of Things (IoT)”, International Journal of Scientific Research in
Computer Science, Engineering and Information Technology, pp. 151-157, 2019. DOI:
https://doi.org/10.32628/CSEIT195137

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


157
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195138
A Review on Various Payment and Billing Techniques
Alka D. Aswar1, Diksha M. Chhipane1, Dipali C. Sontakke1, Pallavi P. Onkar1, Snehal S. Helonde1, Prof. S. V. Sonekar2
1BE Scholar, Department of Information Technology, J.D. College of Engineering & Management, Nagpur, Maharashtra,
India
2Associate Professor, Department of Information Technology, J.D. College of Engineering & Management, Nagpur,
Maharashtra, India
ABSTRACT

Organizations have been compelled to change the manner in which they lead business to react to the ongoing
changes occurring both at the worldwide, provincial and at the national dimension. The motivation behind this
examination was to explore through an overview the electronic billing and payment framework. The premise
of the issue in regards to electronic billing and payment framework that the exploration examine planned to
investigate was talked about dependent on two research destinations; to set up the advantages acknowledged
with the dispatch electronic billing and payment, and to survey difficulties in the execution of electronic billing
and payment.
Keywords : Electronic Billing, Payment System, E-Billing, Online Payment

I. INTRODUCTION change are well on the way to get by in future. As it


states, "On the off chance that you don't care for
E-billing and e-payment frameworks are a piece of e- change, you will like superfluity even less" Some of
business. Electronic Business (e-business) might be the progressions saw in various establishments
characterized comprehensively as any business program, rebuilding, cost-cutting, expansion and the
procedure that depends on a robotized data adjustment of Information and Communication
framework. Electronic business techniques empower Technologies (ICTs).
organizations to connect their inner and outside
information preparing frameworks all the more In Information and Communication Technology, the
efficient and adaptable to work all the more progressions are quick. The fast changes in payment
intimately with providers and accomplices and to frameworks are making a wide exhibit of new
better fulfill the requirements and desires for their business openings and difficulties. One of the open
clients. Today, this is for the most part accomplished doors is electronic billing (e-billing) and electronic
utilizing web-based innovations and portable payment (e-payment). The business needs of KPLC
applications. clients are changing as purchasers grasp new
advances. The Company has left on a huge extension
Organizations have been compelled to change the process through the Rural Electrification Program
manner in which they direct business so as to react to subsidized incompletely by improvement
the ongoing changes occurring both at the worldwide, accomplices, for example, IMF and World Bank. The
territorial and at the national dimension. This is extra number of shoppers has made requirements in
unavoidable as organizations that rush to receive to

CSEIT195138 | Received : 25 Jan 2019 | Accepted : 03 Feb 2019 | January-February -2019 [ 5 (1) : 158-162 ]
158
Alka D. Aswar et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 158-162

the KPLC organize prompting power outages and a portrayal of the information which is filtered and
drop in the nature of administration. deciphered for data. Scanner tag is littler, lighter and
more affordable. The standardized identification must
With the progression in innovation, there are be in the immediate viewable pathway to be perused
different new advances or frameworks that outcome by scanner tag scanner. Scanner tag have no
in the smart billing and stock administration of the perused/compose capacities thus they can't contain
obtained things to spare the time required for the data for example, item expiry date and so on. The
billing and furthermore stock or stock administration scanner tags can be effectively repeated and produced
[1] [2]. The bill is created naturally when the user those outcomes in less security. In the event that the
distinguishes item tag and the data put away with the scanner tag is tore, harmed there's no real way to
tag. The data can be the item name and the expense examine the item.
of the item. The stock administration should be
possible by creating and interfacing the product with
these innovations. At the point when the user
peruses data from the label, it shows the expense and
amount of the item which is being acquired by the
client. After checkout, the bought things are
deducted from the fundamental database server
which deals with the stock and amount. These new
advances are as a rule exceptionally valuable in
Fig (1). Barcode
today’s life by diminishing the heaps of endeavours
and human work. There are numerous such
The standardized tag based information gathering
advancements like Barcode, QR Code, RFID, OCR.
framework can be regularly included any of the
The innovations utilized in current billing
segments like Barcode scanner; scanner tag based
framework. In this investigation, we talk about the
portable PC, standardized tag printers, scanner tag
different existing system for smart or e-billing.
name, standardized tag information accumulation
programming and so on. The PC looks into the cost
II. Review of Literature
in an ace database (the cost of an item isn't in the
standardized tag!), subtracts it from the store stock,
A. Scanner Tag
and ascertains the change. The product additionally
Scanner tag is otherwise called the Auto ID
makes reports with respect to stock dimensions,
(Automatic Identification). It was designed in the
demonstrates what items are the most and least
mid-1970s. The scanner tag was made to help huge
famous, makes statistic provides details regarding
retail locations and markets process their
singular items and clients, and tracks substantially
merchandise. The standardized identification
more. The way to the entire framework is exact
innovation is very effective for the administration
revealing of the item bought. Clerks are innately
and data refreshing in the PCs [1]. Standardized tag is
questionable and moderate. Barcoding is not one or
one of the recognizable business models. It is utilized
the other. Standardized identification benefits in
for examining the things at grocery store for
information exactness, consistency and effectiveness.
checkouts [2]. The standardized tag peruser is utilized
Standardized identification is utilized in Retail
to distinguish the tag on things and recognizes the
Operations, Receiving and dispatching Operations,
item and related data. Scanner tag result in a visual

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


159
Alka D. Aswar et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 158-162

resource the executives, producing tasks, Office and At the point when QR code is checked it connects to
Customer Service Applications and Warehousing and plain content, picture, website pages and interactive
so forth. Standardized identifications are estimated by media, maps areas or anything that can be gotten to
the width of the thin bar and are recorded in mils, or over the web. QR code generator is required to make
1/1000 inch. A 15 Mil standardized identification, for a QR code in which the required data is entered. A
example, has a restricted bar that is 15/1000 inches viable QR code ought to be created utilizing 40-80
wide. Further, "calm zones," or clear spaces to one characters and least of 2cm wide. The more
side and right of standardized identification images, characters result in the debased picture. QR code
are incorporated to safeguard the scanner tag can be enables us to choose the activities we need the clients
perused. to take. The QR code is ISO standard, advertise
demonstrated and totally quantifiable. QR codes
B. QR Code enable the normal individual to decipher the QR
code by examining and perused by a camera-prepared
A QR code is Quick Response code. This code is a 2D Smartphone when you’ve downloaded a scanner
standardized identification bringing about higher application [3].
speed at disentangling substance. The Japanese
company Denso-Wave made the QR code for C. OCR (Optical Character Recognition)
following reason in 1994. QR Codes utilize a 2
dimensional network so as to store information. The OCR is an innovation that empowers you to change
more lines and sections the more information it holds. over various kinds of archives, for example, checked
It underpins the mistake rectification. The code can paper reports, PDF records or pictures caught by an
be examined utilizing cell phone camera or QR advanced camera into editable and accessible
peruser and QR code peruser programming. There information. Albeit Optical Character Recognition
are 4 diff dimension of mistake redress low mid-range (OCR) filtering innovation has expanded quickly
quality high. The most elevated amount of mistake throughout the years, there are constraints with
rectification can re-establish up to 30% of a respect to the source materials and character
degenerate document. designing. Content from a source with a text
dimension of under 12 will result in more mistakes.
OCR additionally requires a spell checking editing
and reformatting for the ideal format. There are
numerous components that will influence the OCR,
for example, shapes, little messages, hazy duplicates,
scientific recipes, shaded paper, and written by hand
massage.

D. RFID

Presently days, radio recurrence ID innovation has


picked up the position from lack of clarity to standard
applications that outcomes in the speed of taking care
Fig(2). QR Code
of fabricated products and materials. RFID can
distinguish the labels utilizing RFID peruser from

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


160
Alka D. Aswar et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 158-162

separation without the viewable pathway. RFID they need to hang tight in the line for longer time.
bolster the immense measure of one of a kind ids that The thought which is proposed utilizing RFID
helps in making the one of a kind labels for innovation will conquer the issue and it gets the
everything. It permits contactless identification of the undertaking less demanding. The joined impacts of
items utilizing the radio recurrence. RFID label simple and adaptable usage, secure transmission of
arrangement is less compelled as it doesn't require record data, and diminished debate offer the
viewable pathway. It has most extreme and more accompanying advantages for all. It will spare time,
grounded read go. RFID bolsters the read/compose vitality and labor of Customer, Owner and provider.
memory capacity and it can hold substantial measure There are numerous advances which are right now
of information under a special identifier. The being utilized for billing frameworks in general stores.
unfriendly conditions, for example, dust, substance, The choice of the innovation relies on the execution,
physical harm don't influence on RFID without a proficiency and QoS of the innovation in regards to
doubt. This innovation has three sections: a filtering specific undertaking and condition.
reception apparatus, a handset with a decoder to
translate the information and a transponder (RFID IV.REFERENCES
tag) pre-set with data. RFID labels can be perused at a
quicker rate as they roughly can peruse 40 RFID [1] T. Sriram, K. Vishwanatha Rao, S. Biswas and
labels at same time. RFID contain abnormal amounts B. Ahmed "Applications of barcode technology
of security; information can be encoded, secret key in automated storage and retrieval systems",
ensured or set to incorporate an "execute" highlight Industrial Electronics, Control, and
to evacuate information for all time [4]. Instrumentation, 1996., Proceedings of the
1996 IEEE IECON 22nd International
The Tag or Transponder gets recognized by the Conference on, vol. 1, pp.641 -646
Reader or Interrogator when it goes inside the [2] Alistair Milne “The Rise and Success of the
receiving wire extend and the distinguished novel id Barcode: some lessons for Financial Services”
is sent to the Computer or server. feb 2013
[3] Andrew Kerry-Bedell IT‟s in Conservation
III.Comparative Analysis Consultants. Mobile: 07899 741939 email:
a.kb@live.co.uk
http://www.kbstconsulting.co.uk/QR/images/Q
R.pdf
[4] Mandeep Kaur, Manjeet Sandhu, Neeraj Mohan
and Parvinder S. Sandhu “RFID Technology
Principles, Advantages, Limitations & Its
Applications” International Journal of
Computer and Electrical Engineering, Vol.3,
No.1, February, 2011
[5] Abwao, V. (2012). Information Technology
The payment of bill by remaining in long line is a
Applications in Business Management within
monotonous factor when individuals need to buy
Kenyan Companies; A Survey of Selected
items from shops. In spite of the fact that individuals
Insurance Firms in Nairobi, Unpublished MBA
can pay immediately utilizing electronic cash office,
Thesis, University of Nairobi.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


161
Alka D. Aswar et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 158-162

[6] Andrew, J, W. (2010). Electronic bill payment


and presentment: Jump-starting the next great
Web enabled consumer empowerment engine,
Robertson Stephens e-Processing Research,
April 11.
[7] Atebe, E. (2011). An Effect of Business Process
Reengineering on Business Process Cylce. The
Case of KLPC, Unpublished MBA Thesis,
University of Nairobi.

Cite this article as :

Alka D. Aswar, Diksha M. Chhipane, Dipali C.


Sontakke, Pallavi P. Onkar, Snehal S. Helonde, Prof.
S. V. Sonekar, "A Review on Various Payment and
Billing Techniques", International Journal of
Scientific Research in Computer Science, Engineering
and Information Technology (IJSRCSEIT), ISSN :
2456-3307, Volume 5 Issue 1, pp. 158-162, January-
February 2019. Available at doi :
https://doi.org/10.32628/CSEIT195138
Journal URL : http://ijsrcseit.com/CSEIT195138

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


162
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195128

Finite/Infinite Queueing Models Performance Analysis in


Optical Switching Network Nodes
Waleed M. Gaballah*
* Department of Electronics and Communication Engineering, Mansoura High Institute of Engineering and
Technology, Mansoura, Egypt

ABSTRACT

In optical switching networks, optical information (packets/bursts) that are forwarded from an optical switch to
another through the network could entering a queue of a certain length in each node and may be waiting inside
this node before it will be transmitted to the next node. Various queueing models have been widely applied as
the main tool in optical switching networks for modeling and performance evaluation analysis of the switching
nodes. In this paper, studying the performance analysis of an optical switching node is discussed at Finite and
Infinite queueing models to have the optimum queueing model for optical core node switch design. The
waiting delay time, the average expected number of optical packets and the loss probability of packets in the
optical switch are estimated at variable traffic loads and different wavelength channels.
Keywords : Queueing Models, Optical Switching Network, Finite Queue, Infinite Queue, Blocking Probability

I. INTRODUCTION time. In the switch if the wavelength channel is busy


serving other packets, the arrivals are queued in the
An optical information traffic access to the network switch queueing buffer. Therefore, the model can
with a given capacity is modeled in optical nodes as a determine the distribution of the number of packets
queue with a certain distribution of traffic arrival in the system and their waiting time.
times and a certain distribution of traffic service time.
Queueing models in the optical switching networks
Queueing in the network nodes are modeled in a have various application aspects. Queueing models
variety of network techniques and principles [1, 2]. with optical delay lines in optical packet switching
The Queueing model is a theoretical aspect of such networks are studied in [4]. The edge OBS node
the network node [3]. Typically, a queueing model queueing is modelled in [5]. An analytical model
represents the network node physical configuration optical delay line buffers in OBS networks using
by determining the number of optical packets in the queueing theory are developed in [6].
switch and how fast that switch serves the optical
packet traffic. Also, the queueing models give a In this paper, study the performance analysis of the
statistical nature of the optical switch node, by optical switching network node is performed at
specifying the variability in the arrival process to the different queueing models. Validating that the
switch and in the switch service process. In an optical various queueing model parameters and the blocking
switching, optical packets arrive at a system (switch) probability are affected by the wavelength channels
as random intervals and are served during a random number and traffic loads.

CSEIT195128 | Received : 25 Jan 2019 | Accepted : 03 Feb 2019 | January-February -2019 [ 5 (1) : 163-170 ]
163
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

For this, our work first presents some specific capacity. In finite queue, the optical arrivals that
queueing models overview that can be used in optical attempt to enter the full-occupied system are denied
switching networks, at section 2. The infinite and entry or blocked.
finite queueing models’ numerical analysis is
obtained in section 3. Finally, conclude in section 4. Table 1 represents the main differences between
Finite and Infinite queueing models.
II. Queueing Models in Optical Switches
TABLE I
FINITE VS. INFINITE QUEUEING MODELS
At the optical switches, the arriving optical
information traffic request some specific amount of Infinite Queueing
Finite Queueing Model
resources such as; circuit, bandwidth, wavelength Model
channel, etc. to be served. The most common M/M/w/m M/M/w
queueing models assume that the optical information The arrival rate The arrival rate is not
traffic inter-arrival and service times follow the depends on the number affected by the number
exponential distribution or equivalently follow a of served and waiting of packets being served
Poisson distribution process with Markovian or packets in the system. and waiting.
memoryless properties [7]. A commonly used Limited buffer capacity. Unlimited buffer
shorthand notation, called Kendall’s notation [8], for capacity.
such queue models describes the arrival process, Faster and has lower Packets waiting long
service distribution, the number of servers and the average number of times in the buffer.
buffer size (waiting line). The complete notation waiting packets in the
expressed as (a/b/c/d) where, Arrival process/service system.
distribution/ number of servers/waiting line. There is a packet loss No packet loss
probability. probability.
In optical switching networks, the commonly used
characters for the first two positions in the shorthand III. Infinite/Finite Queueing Models Performance
notations are M (Markovian – Poisson for the arrival Analysis
or Exponential for the service time). The third
positions used for the number of the output optical In this section, a numerical performance analysis of
wavelength channels w. The fourth position indicates infinite and finite queueing models in the optical
the switch queueing size m and it's usually not used switching node are represented. This analysis study
in infinity waiting room buffers. aimed to determine the suitable queuing model for
enhancement of an optical switch performance,
There are single server queueing models such as which queueing model has low loss probability,
M/M/1 and M/M/1/w, and multiple server's systems which one is faster and has lesser number of optical
such as M/M/w, M/M/w/m, and M/M/w/w systems packets in the optical switch system serviced and
[9]. There is an infinite queue system such as M/M/w, waiting for service. In our analysis taking in
where the optical traffic arrivals are hold waiting for consideration the effect of the number of optical
service and not affected by the number of packets switch wavelength channels w whether at Infinite
already on the queue because there is unlimited model M/M/w or Finite one M/M/w/m. The queueing
buffer size. In addition, there is a finite queue system models are used at suitable average arrival rate 
such as M/M/w/m, which has a limited buffer

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


164
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

packet/sec and average service rate µ packets/sec wavelengths (servers), w = 1, 2, 8, 16, 32. In Fig. 1,
values. the service rate per channel is constant µ = 1
The performance parameters that measured are: packet/time unit. While at figures 2 to 4, the entire
- The average number of optical packets resident in switching capacity wµ is kept constant by µ = 1/w
the system Ls packets. packets/time unit, which gives  = λ becomes
- The average number of optical packets waiting in independent on w.
the queue Lq packets.
- The average time of optical packets spend in the From Fig. 1 it is clear that for a lower offered load,
system (the average switch queueing delay time) the mean time in the system is very low and equal to
Ws sec. 1/µ = 1 time unit. As the load increases, in the M/M/1
- The average time of optical packets waiting in the queue, the mean total time in the optical switch
queue (the average waiting time to serviced) Wq increases greatly and increased slightly as the number
sec. of wavelengths w increased. At the M/M/32 queue,
- The blocking probability of optical packets PB (at the mean total time in the system only has a slightly
finite queueing models). increment. Therefore, as the number of wavelengths
increased the queue has a superior performance than
In this analysis, the waiting time in the system Ws
M/M/1 queue.
established at two different cases [10]. First, if it is
considered that the service rate per channel is
constant µ = 1 packet/sec, which gives  = λ/w. Later, 4
M/M/1
Average waiting time in the switch Ws time unit

the total service rate in the switch is kept constant M/M/2


wµ = 1, which gives  = λ becomes independent on w. M/M/8
3 M/M/16
At Infinite model we will study the influence of the Ws1(  )
M/M/32
Ws2(  )
wavelength channel numbers at the different two
Ws3(  )
analysis cases. At Finite model the effect of the Ws4(  )
2

switch queueing size and the number of wavelength Ws5(  )

channels are illustrated on the optical switch 1


performance taking in consideration the second
analysis case.
0
0 0.2 0.4 0.6 0.8 1
A. Infinite Queueing Models 
Offered Load 
Figure 1. Average waiting time in the system vs. the
offered load of M/M/w with different number of
With infinite queuing systems M/M/w, where w is
wavelength channels and μ = 1 packet/time unit per
the number of servers or optical wavelength channels,
channel
the queue buffer size is infinite. The system filling Ls
and the waiting times Wq and Ws, in the queue and If the switching capacity is constant, figures 2 to 4 are
the optical switch respectively, establish the main illustrated. Obviously, at low loads, less than 80%
optical switch characteristics. load or  < 0.8, the Waiting time in the optical switch
(mean flow time), Ws = Wq + 1/μ , shown is ruled by
In figures from 1 to 4, the optical switch performance the increased holding time, 1/μ = w, which results
is studied at an infinite queueing models. Ws, Wq and from decreasing the service rate for increased
Ls parameters are represented at different number of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


165
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

wavelength numbers w in order to keep the

Average number of packets in the switch Ls (packets)


50
switching capacity constant. w=1
w=2
40 w=8
At high loads the mean time in the optical switch Ls1(  ) w = 16
Ls2(  ) w = 32
increases significantly because the waiting time in 30
Ls3(  )
the queue component Wq becomes dominant. This
Ls4(  )
increase in queueing waiting time also causes the Ls5(  )
20

increase in the system filling (Number of packets in


10
the system) Ls shown in Fig. 3. However, this only
results from the longer serving interval required per 0
0 0.2 0.4 0.6 0.8 1
server and not from an increase in the mean waiting 
Offered Load 
queueing time Wq. Figure 4 shows the mean waiting
time in the switch queue Wq. It is decreased Figure 3. Average number of packets in the system vs.
dramatically with increasing wavelength numbers w, the offered load of M/M/w with different number of
and the service rate per channel be slower. So, in this wavelength channels and normalized to equal system
case, it is clear that a single wavelength queueing load  =  by setting μ = 1/w packet/time unit
model is preferred over a multi-wavelength model
Average waiting time in the queue Wq time unit

because it has less average number of packets resident 0.2

in the system and lower mean time in the system.


0.15
Wq1(  )

Wq2(  )
Average waiting time in the switch Ws time unit

M/M/1 Wq3(  )
0.1
Wq4(  )
M/M/2
40 Wq5(  )
Ws1( ) M/M/8 w =1
Ws2( )
M/M/16 0.05 w =2
30 M/M/32 w =8
Ws3( ) w = 16
w = 32
Ws4( ) 0
20 0 0.2 0.4 0.6 0.8 1
Ws5( ) 
Offered Load 
10

Figure 4. Average waiting time in the queue vs. the


0
0 0.2 0.4 0.6 0.8 1 offered load of M/M/w with different number of
Offered Load  wavelength channels and normalized to equal system
load  =  by setting μ = 1/w packet/time unit
Figure 2. Average waiting time in the system vs. the
offered load of M/M/w with different number of In a realistic view, if we consider that the switching
wavelength channels and normalized to equal system capacity is constant, the best service is provided for
load  =  by setting μ = 1/w packet/time unit per the lower number of wavelengths possible. While, if
channel the service rate per channel is constant, increasing a
system wavelength channels is preferred.

Consequently, if the costs of using more wavelengths


switch less, constant service rate per channel is more
preferred choice. However, constant switching
capacity is better at lower wavelengths number. An

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


166
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

optical switch with an adequately chosen number of Figure 5. Average number of packets in the system
wavelengths is the better choice. vs. the offered load of M/M/w/m with different
switch buffer size m and w = 8 wavelengths
B. Finite Queueing Models

Average waiting time in the switch Ws time unit


In finite queueing model systems M/M/w/m, an 80
m=8
arriving optical packet may be admitted to the free m = 16
wavelength channel immediately. It may be placed in 60
m = 32
Ws1(  )
m = 64
the queue until a wavelength channel is available, or
Ws2(  )
it may be blocked due to all wavelengths are busy Ws3(  )40

and all buffer places are occupied. Ws4(  )

20

Figures from 5 to 8 illustrate how the finite queueing


M/M/w/m performance depends on the system size m. 0
1 1.2 1.4 1.6 1.8
These figures show the mean time in the optical 
Offered Load 
switch Ws, the mean waiting time in the switch
Figure 6. Average waiting time in the switch vs. the
queue Wq, the mean optical packet number in the
offered load of M/M/w/m with different switch
switch Ls and the blocking probability PB all over the
buffer size m and w = 8 wavelengths
increasing system load  of M/M/w/m queueing
systems. The analysis is using eight wavelengths, w =
Average waiting time in the queue Wq time unit

8, and the switch buffer size m = (8, 16, 32, 64). 40


Impact of the system size m on the optical switch
declared at high traffic loads  > 1. As the system Wq1(  ) 20

buffer size increases, with a constant wavelength Wq2(  )

Wq3(  )
number, the more load can be buffered and that may 0
Wq4(  )
increase the mean optical packets number in the m=8
m = 16
system (switch filling) Ls, Fig. 5. Therefore, it is − 20
m = 32
increasing the waiting time in the queue and in the m = 64

system Wq and Ws, figures 6 and 7 respectively. 1 1.2 1.4 1.6 1.8

Offered Load 
Average number of packets in the switch Ls (packets)

100 Figure 7. Average waiting time in the queue vs. the


offered load of M/M/w/m with different switch
buffer size m and w = 8 wavelengths
Ls1( )
10
Ls2( )
However, this cause a heavily decreases in the
Ls3( )

Ls4( ) switching blocking probability PB as in Fig. 8.


1 m=8
m =16
m = 32
m = 64
0.1
1 1.2 1.4 1.6 1.8
Offered Load 

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


167
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

Figure 9. Average waiting time in the switch vs.


− 13
110 Average number of wavelengths for M/M/w/m with
Blocking Probability of the switch PB

− 27
110 different switch buffer size m at  = 
− 41
pk1(  ) 110
− 55
pk2(  ) 110
− 69
pk3(  ) 110 100

Average number of packets in the switch Ls


− 83
pk4(  ) 110
− 97 m=8
110
m = 16
− 11 1
110 m = 32 Ls1(  )
− 12 5 10
110 m = 64
Ls2(  )

(packets)
− 13 9
110
1 1.2 1.4 1.6 1.8 Ls3(  )
Offered Load 
Ls4(  )
1 m=8
Figure 8. Packets blocking probability PB in the m =16
m = 32
system vs. the offered load of M/M/w/m with m = 64
different switch buffer size m and w = 8 wavelengths 0.1
20 40 60
Average number of wavelengths w
For the special case m = w at queue model M/M/w/w,
no loads can wait for service, such that no waiting Figure 10. Average number of packets in the switch
time in the queue, Wq = 0. In this case, all loads that vs. Average number of wavelengths for M/M/w/m
cannot be served immediately become blocked, and with different switch buffer size m at  = 
consequently, these systems are commonly called loss
systems.
Average waiting time in the queue Wq time

40

To study the influence of the number of wavelength


channels on the queueing model performance, figures Wq1(  ) 20

Wq2(  )
from 9 to 12 show the performance of M/M/w/m
unit

Wq3(  )
switching system over the number of provided 0
Wq4(  )
wavelengths w. This study at high load  = 2, at m=8
different system size m values (m = 8, 16, 32, 64), and − 20
m = 16
m = 32
constant system capacity wμ = 1. m = 64
20 40 60

Average waiting time in the switch Ws time unit

100
Average number of wavelengths w

80 Figure 11. Average waiting time in the queue vs.


Ws1(  ) Average number of wavelengths for M/M/w/m with
Ws2(  ) 60
different switch buffer size m at  = 
Ws3(  )

Ws4(  ) 40
m=8
m = 16
20
m = 32
m = 64
0
0 20 40 60

Average number of wavelengths w

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


168
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

−4
110
Blocking Probability of the switch PB

− 36 0.1
110

Blocking Probability of the switch PB


−3
− 68 110
110
−5
110
pk1(  ) 110− 10 0
− 13 2 −7
110
pk2(  ) 110 pk1(  )
− 16 4 −9
pk2(  ) 110
pk3(  ) 110
− 19 6 pk3(  ) − 11
110
pk4(  ) 110
− 22 8 m=8 − 13
110 110
− 26 0
m = 16 − 15
110 110 M/M/8/8
m = 32
− 29 2 − 17 M/M/16/16
110 m = 64 110
M/M/32/32
− 19
0 110
0 20 40 60 80 0 0.5 1 1.5 2

Average numberof wavelengths w Offered Load 

Figure 12. Packets blocking probability PB in the Figure 13. Packet blocking probability PB in the
system vs. average number of wavelengths w of system vs. the offered load of M/M/w/w at different
M/M/w/m with different switch buffer size m at  =  servers and  = 

Again, for m = w no waiting space exists, and thus, It is clear that as the load increases, the blocking
the waiting time in the queue Wq equal zero in case probability also increases at low traffic and saturated
of M/M/8/8. In addition that the queue M/M/8/8 at high traffic. However, as the number of
gives the better performance, less average number of wavelength channels increases, the blocking
packets and low waiting time in the system. However, probability decreases because the incoming traffic has
the strong reduction of the blocking probability PB is a more chance to be serviced.
gained from the implicit queue size increase.
IV.CONCLUSION
Then, on finite queueing optical switches, the
blocking probability decreases at increasing the Optical switching network node performance
number of wavelengths (servers). analysis was done with Finite and Infinite queueing
models to indicate the optimum model at optical
Finally, Fig. 13 investigates the switch blocking switch design. Queue delay time, expected number of
probability PB for the M/M/w/w model at different packets in the system and blocking probability are
number of wavelength channels w = 8, 16, 32. The the main parameters in the analysis to demonstrate
switch performance illustrated at different traffic the optical switch performance. At Infinite queueing
loads  = 0 to 2. model, two different analysis cases were used to
represent the wavelength channel number impact on
the optical switch node performance. First analysis
was considered a constant service rate per channel
which is preferred if the costs of using more
wavelengths switch less. However, fixed switching
capacity analysis is better at lower wavelength
number. For Finite queueing model, impact of the
system size on the optical switch declared at high

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


169
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

traffic loads. As the system buffer size increases, with [5] Lisong Xu, Harry G. Perros and G. N. R., “A
a constant wavelength number, the more load can be Queueing Network Model of an Edge Optical
buffered and that may increase the mean optical Burst Switching Node,” in INFOCOM 2003.
packets number in the system. Therefore, it is Twenty-Second Annual Joint Conference of the
increasing the waiting time in the queue and in the IEEE Computer and Communications. IEEE
system. Also, the switching blocking probability PB is Societies, 2003, pp. 2019–2029.
heavily decreases. [6] Kumar, OS Sunish, C. S. Jaison, Indu Reena
Varughese and V. K. J., “The Analysis of M/M/r
As the number of wavelength channels increased Queuing Model for Fiber Delay Lines to
with fixed queue buffer size, the blocking probability Resolve BLP in OBS Networks,” In Emerging
increased due to the switch buffer capacity lowered Research Areas and 2013 International
until it reaches zero at M/M/w/w queueing system. Conference on Microelectronics,
At M/M/w/w queue system, it is faster and has less Communications and Renewable Energy
waiting packets in the switch, while it has high (AICERA/ICMiCR), 2013 Annual International
blocking probability due to no queue buffer. At this Conference, 2013, pp. 1–3.
model, as the number of wavelength channels [7] Bolch Gunter, Stefan Greiner, Hermann de
increases, the blocking probability decreases due to Meer, and Kishor S. Trivedi, Queueing
that the incoming traffic has a more chance to networks and Markov chains: modeling and
serviced. Therefore, using M/M/w/w queueing performance evaluation with computer science
system can be modelled well at optical switching applications, John Wiley & Sons, 2006.
nodes under a certain predefined number of [8] Harchol-Balter, Performance modeling and
wavelengths to lower the blocking probability design of computer systems: queueing theory in
problem. action, Cambridge University Press, 2013.
[9] Zukerman M., “Introduction to queueing
V. REFERENCES theory and stochastic teletraffic models,”
arXiv:1307.2968, 2018.
[1] Van Do, Tien, et al., eds., Queueing Theory and [10] Bhat, U. Narayan. An introduction to queueing
Network Applications, Springer International theory: modeling and analysis in applications,
Publishing, 2016. Birkhäuser, 2015.
[2] E. Yue, Wuyi, Quan-Lin Li, Shunfu Jin, and
Zhanyou Ma, “Queueing Theory and Network Cite this article as :
Applications,” in 12th International Waleed M. Gaballah, "Finite/Infinite Queueing
Conference, QTNA 2017, 2017. Models Performance Analysis in Optical Switching
[3] Thomopoulos, Nick T., Fundamentals of Network Nodes", International Journal of Scientific
Queuing Systems: Statistical Methods for Research in Computer Science, Engineering and
Analyzing Queuing Models, Springer Science & Information Technology (IJSRCSEIT), ISSN : 2456-
Business Media, 2012. 3307, Volume 5 Issue 1, pp. 163-170, January-
[4] Zhu, Xiaoming, and Joseph M. Kahn, February 2019. Available at doi :
"Queueing models of optical delay lines in https://doi.org/10.32628/CSEIT195128
synchronous and asynchronous optical packet- Journal URL : http://ijsrcseit.com/CSEIT195128
switching networks," Optical Engineering, vol.
42, no. 6, pp. 1741-1748, 2003.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


170
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195118
Improving Central and Remote Access Units in Cloud Based Wi-
Fi Network for Better Client and Service
Kavitha-R1, Sathiyabama-G2
1Department of Computer Science and Engineering, Velalar College of Engineering &Technology, Erode,
Tamilnadu, India
2 ME, Department of Computer Science and Engineering, Velalar College of Engineering &Technology, Erode,
Tamilnadu, India
ABSTRACT

People show attention to cloud computing since it is efficient and scalable. But maintaining the stability of
processing so many jobs in the cloud computing environment with load balancing is a very difficult problem
and receives much attention by researchers. Load balancing in cloud computing platform has performance
impact also. Good load balancing makes cloud computing more efficient and increases user satisfaction. This
paper introduces a better load balance model for public cloud based on cloud partitioning concept with a switch
mechanism to choose various strategies in different situations. The algorithm applies game theory to load
balancing strategy and improves the efficiency in the public cloud environment. MCC is a concept which aims
to mitigate these limitations by extending the capabilities of smart devices by employing cloud services, as
required. In MCC both the data storage and processing occur external to the mobile device, while in cloud
computing it is usually only the data storage which is external. In this following a some of these architectures as
categorized in augmented execution, elastic partitioned/modularized applications, application mobility, ad-hoc
mobile cloud and add a fifth category; cyber foraging. This system addresses whether MCC techniques can be
used to extend the capabilities of resource-constrained mobile-devices to provide the illusion of infinite, elastic
resources on demand. The existing system limitations of mobile-devices, and identified five key limited
resources as being CPU, memory, battery, data usage and time. In this research explored existing solutions for
these limitations and identified offloading computation and storage from the device as a possible solution.
Keywords : Mobility, Load Balancing, MCC, Cyber Foraging

I. INTRODUCTION The most sanctioning technology for cloud


computing is virtualization. Virtualization computer
Cloud computing is internet-based computing during code permits a physical computer to be electronically
which massive teams of remote servers area unit separated into one or additional "virtual" devices,
networked to permit sharing of data-processing tasks, every of which may be simply used and managed to
centralized information storage, and on-line access to perform computing tasks. Cloud computing adopts
pc services or resources. Clouds are classified as ideas from Service directed design (SOA) that may
public, non-public or hybrid. Cloud computing may facilitate the user break these issues into services that
be a form of computing that depends on sharing may be integrated to supply an answer. Cloud
computing resources instead of having native servers computing provides all of its resources as services,
or personal devices to handle applications. and makes use of the well-established standards and

CSEIT195118 | Received : 25 Jan 2019 | Accepted : 05 Feb 2019 | January-February -2019 [ 5 (1) : 171-179 ]
171
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

best practices gained within the domain of SOA to 3. Microsoft Azure


permit world and simple access to cloud services in 4. There are many small and medium scale cloud
an exceedingly standardized method. operators that include IBM, Oracle etc.

The internet of Things is setting out to rework daily


tasks area unit completed. the net of Things (IoT)
consists of everyday objects – physical devices,
vehicles, buildings etc. with embedded natural
philosophy, software, sensors, and network property,
permitting them to gather, send and receive
knowledge. The IoT generates a massive quantity of
huge knowledge and this successively puts a large
strain on web Infrastructure. As a result, this forces
corporations to search out solutions to minimize the
pressure and solve their drawback of transferring
massive amounts of knowledge. Cloud computing has
entered the thought of knowledge technology,
Cloud computing may be a quite grid computing; it's
providing quantifiability in delivery of enterprise
evolved by addressing the QoS (quality of service)
applications and software system as a Service (SaaS).
and responsibleness issues. Cloud computing provides
Corporation area unit currently migrating their data
the tools and technologies to make data/compute
operations to the cloud. Several cloud suppliers will
intensive parallel applications with rather more
yield your knowledge to be either transferred via
reasonable costs compared to ancient parallel
your ancient web association or via a passionate
computing techniques.
direct link. The advantage of an on the spot link into
the cloud can make sure that your knowledge is
Characteristics
uncontended which the traffic isn't crossing the net
Cloud computing exhibits the following key
and also the Quality of Service are often controlled.
characteristics
• Agility
II. RELATED WORKS
• Application programming interface
• Cost reductions
1. Niroshinie Fernando, Seng W. Loke, Wenny
• Device and location independence
Rahayu
• Maintenance
• Multitenancy
Despite increasing usage of mobile computing,
• Performance
exploiting its full potential is troublesome thanks to
• Productivity
its inherent issues like resource scarceness, frequent
• Reliability and Scalability
disconnections, and quality. Mobile cloud computing
• Security
will address these issues by corporal punishment
mobile applications on resource suppliers external to
Cloud computing is dominated by
the mobile device. during this paper [1], to produce
1. Amazon Web Services (AWS),
an intensive survey of mobile cloud computing
2. Google Compute Engine, Google App Engine
analysis, whereas lightness the precise considerations

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


172
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

in mobile cloud computing. this taxonomy supported facilitate building a lot of powerful mobile
the key problems during this space, and discusses the applications.
various approaches taken to tackle these problems. To
conclude the paper with a assessment and challenges 3. HanQi , Abdullah Gani
that haven't nevertheless been absolutely met, and Mobile Cloud Computing (MCC) which mixes mobile
highlight directions for future work. computing and cloud computing, has become one
The increasing usage of mobile computing is clear by amongst the trade buzz words and a serious
the study by Juniper analysis, that states that the discussion thread within the IT world since 2009. As
buyer and enterprise marketplace for cloud-based MCC continues to be at the first stage of development,
mobile applications is anticipated to rise to $9.5 it's necessary to understand a radical understanding
billion. In recent years, applications targeted at of the technology so as to entails the direction of
mobile devices have started changing into plentiful future analysis. With the latter aim, this paper
with applications in numerous classes like presents a review on the background and principle of
amusement, health, games, business, social MCC, characteristics, recent analysis work, and
networking, travel and news. The recognition of future analysis trends. a quick account on the
those ar evident by browsing through mobile app background of MCC: from mobile computing to
transfer centers like Apple’s iTunes or Nokia’s Ovi cloud computing is given and so followed with a
suite. discussion on characteristics and up to date analysis
work. It then analyses the options and infrastructure
2. Dejan Kovachev, Yiwei Cao and Ralf Klamma of mobile cloud computing.
Cloud computing is an rising construct combining
several fields of computing. the inspiration of cloud 4.Preeti Garg, Dr. Vineet Sharma
computing is that the delivery of services, computer Describes a cloud computing extremely climbable
code and process capability over the net, reducing computing resources are equipped as associate outer
price, increasing storage, automating systems, service through net on pay-as-usability basis. Portio
decoupling of service delivery from underlying analysis estimates that mobile subscribers can reach
technology, and providing flexibility and quality of half dozen.5 billion by the tip of 2012, 6.9 billion by
knowledge. However, the particular realization of the tip of 201 three. thanks to increasing use of
those edges is much from being achieved for mobile mobile devices the necessity of cloud computing in
applications and open several new analysis queries. In mobile devices arise, that gave birth to Mobile Cloud
order to higher perceive the way to facilitate the Computing (MCC). Mobile Cloud Computing refers
building of mobile cloud-based applications, they to associate infrastructure wherever processing and
need surveyed existing add mobile computing storage will happen faraway from mobile device.
through the prism of cloud computing principles. The Mobile devices don't have to have massive storage
given definition of mobile cloud computing and capability and powerful electronic equipment speed.
supply an outline of the results from this review, Thanks to storing information on cloud there's
specially, models of mobile cloud applications. To associate issue of information security. as a result of
focus on analysis challenges within the space of the chance related to information storage several IT
mobile cloud computing. To conclude with professionals aren't showing their interest towards
recommendations for a way this higher Mobile Cloud Computing. This paper[10] explores: (i)
understanding of mobile cloud computing will The thought of Mobile Cloud Computing and
problems associated in it (ii) Security of information

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


173
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

hold on in cloud with varied mechanisms (iii) resource utilization while maintaining affordable
projected a attainable answer to supply quality of service. However, such a goal cannot be
confidentiality, access management similarly as achieved while not economical resource management.
integrity of information. The most physical resources, like processor cores and
I/O devices, are shared among virtual machines
5.Byung-Gon Chun, Sunghwan Ihm, Petros Maniatis exploitation time slicing and are often scheduled
Describe a mobile applications are becoming flexibly supported priority, allocating an acceptable
increasingly ubiquitous and provide ever richer quantity of main memory to virtual machines is
functionality on mobile devices. Meanwhile, such tougher. Totally different applications have different
devices often enjoy strong connectivity with more memory needs. Even one application shows varied
powerful machines ranging from laptops and operating set sizes throughout its execution. A best
desktops to commercial clouds. This paper presents memory management strategy below a virtualized
the design and implementation of Clone Cloud, a environment therefore has to dynamically alter
system that automatically transforms mobile memory allocation for every virtual machine that
applications to benefit from the cloud. The system is more needs a prediction model that forecasts its host
a flexible application partitioned and execution physical memory wants on the fly. This paper
runtime that enables unmodified mobile applications introduces Memory Balancer (MEB) that dynamically
running in an application-level virtual machine to monitors the memory usage of every virtual machine,
seamlessly off-load part of their execution from accurately predicts its memory wants, and
mobile devices into device clones operating in a sporadically reallocates host memory. MEB uses 2
computational cloud. Clone Cloud uses a effective memory predictors that, severally, estimate
combination of static analysis and dynamic profiling the number of memory accessible for reclaiming
to partition applications automatically at a fine while not a notable performance drop, and extra
granularity while optimizing execution time and memory needed for reducing the virtual machine
energy use for a target computation and paging penalty. The advantage of this study is that
communication environment. During runtime, the swap usage from the OS performance statistics
application partitioning is achieved by migrating a and posts the swap usage statistics to the central
thread from the mobile device at a chosen point to information storage. The limitation of this study is
the clone in the cloud, executing there for the residue reducing overhead while up prediction accuracy.
of the partition, and re-integrating the migrated
thread back to mobile. Their evaluation shows that 7.Weizhe Zhang et al
Clone Cloud can adapt application partitioning to Cloud computing emerges as a brand new computing
different environments, and can help some paradigm involved by each domain and trade.
applications achieve as much as a 20x execution Resource management of multiple virtual machines is
speed- up and a 20-fold reduce of energy spent on that the core of Infrastructure as a Service.
mobile. Specializing in the mainframe resources, the aim of
this paper is to extend the QoS of net service by
6.Weiming Zhao properly planning the mainframe resource across the
Virtualization primarily enables multiple operating virtual machines. The authors of this study formulate
systems and applications to run on one physical pc by the mainframe planning of multiple virtual machines
multiplexing hardware resources. A key motivation into an whole number programming drawback. Then,
for applying virtualization is to enhance hardware a worldwide regulation algorithmic rule supported

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


174
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

utility optimization theory is planned. It principally The administrator selects a minimum and a most
consists of 4 parts: the monitor, the regulator, the information measure limit. The primary pre-copy
native effectiveness of management and therefore the spherical transfers pages at the minimum information
international effectiveness of management. The measure. Every resultant spherical counts the
regulator is within the Xen virtual machines’ monitor. quantity of pages dirtied within the previous
The monitor, native effectiveness of management and spherical, and divides this by the length of the
therefore the net server are settled in DomainU’s previous spherical to calculate the soiling rate. The
interior. The international effectiveness of information measure limit for consecutive spherical is
management is created of the worldwide utility then determined by adding a relentless increment to
generating module and global optimizing and finding the previous round’s soiling rate have by trial and
module, realized in the Domain0. Monitor: the most error determined that 50Mbit/sec may be a
role of monitor is to discover an internet server QoS appropriate worth. It terminates pre-copying once
metrics, in the text it means that the typical latency the calculated rate is bigger than the administrator’s
of net. In their experimental environment Apache chosen most, or once but 256KB remains to be
module was developed for the Apache server to transferred. By integration live OS migration into the
record period of time server latency. Additionally, we Xen virtual machine monitor we tend to alter speedy
have a tendency to monitor the changes of the system movement of interactive workloads at intervals
resources. It’s principally to try to do some tests for clusters and knowledge centers. The dynamic
proc CPUinfo, and procure the condition of the network-bandwidth adaptation permits migration to
system mainframe etc. proceed with stripped-down impact on running
services, whereas reducing total time period to below
8. Christopher Clark et al discernible thresholds. Feature of this study is
Migrating software package instances across distinct moving the contents of a VM’s memory from one
physical hosts may be a useful gizmo for directors of physical host to a different is approached in any
knowledge centers and clusters: It permits a clean range of how. And also the limitation may be a live
separation between hardware and computer code, service it's vital that this transfer happens in a very
and facilitates fault management, load equalization, manner that balances the necessities of minimizing.
and low-level system maintenance. By finishing up
the bulk of migration whereas OSes still run, atuhors 9. Albert M.K. Cheng et al
reach spectacular performance with stripped-down Describe a Virtualization technology recently
service downtimes; the study demonstrate the becomes a hot analysis topic once more in each
migration of entire OS instances on a artifact cluster, business and teachers. Some physical hardware like
recording service downtimes as low as 60ms. And processors and I/O devices area unit shared among
also the author shows that that our performance is virtual machines victimization time slicing, however
adequate to create live migration a sensible tool even the memory resource management is comparatively
for servers running interactive hundreds. During this sophisticated. Light-weight memory management
paper, the look choices for migrating OSes running design for multiple virtual machines is projected, and
services with aliveness constraints, specializing in it includes a mix of self adjustment and global-
knowledge center and cluster environments is taken adjustment policies, each of that collaborate with one
into account. another to boost the memory potency. IC module: it
consists of 2 parts: Virtual Machine Collector (VMC)
and Physical Machine Collector (PMC). VMC is

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


175
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

found in DomainU, accountable for aggregation every PROPOSED WORK


domain’s memory usage and disc space usage statistics.
PMC is put in at Domain0 to gather the scale of total MCC is a concept which aims to mitigate these
and idle memory within the physical machine, etc. limitations by extending the capabilities of smart
Regulator module: it's accountable to control devices by employing cloud services, as required. In
resources. The authors used Balloon-Driver as a MCC both the data storage and processing occur
restrictive mechanism of memory to trigger the external to the mobile device, while in cloud
memory rescaling and XenStore to exchange computing it is usually only the data storage which is
knowledge between varied domains. external

PM module: it's the core of the system. If the These Paper describes the choices to synchronize the
resource is decent, memory sizes area unit adjusted offline folder contents to the online. Word
by the VMs themselves that is named self-adjustment. documents, surpass worksheets and alternative files is
Otherwise, the global-adjustment is activated for the designated and uploaded to the online application.
complete memory management consistently. Additionally, they'll be reborn into bytes and saved
Following area unit the benefits of this study: Self- within the info provided within the net server
adjustment is running as a daemon method once the location. In anyplace, the administrator will read the
OS of a particular domain starts. contents within the server choose a file and transfer
to the system where the administrator is functioning.
10. Tudor-Ioan Salomie, Gustavo dancer, Timothy Likewise, the files are uploaded from the online page
Roscoe additionally. The windows application is wont to
Describes a program like databases and language transfer the content from the online website to the
runtimes usually manage memory themselves to use native machine. So the applying synchronizes the
application information out of stock to the OS. online content and windows content. The applying
Historically deployed on dedicated machines, they're acts like content storage also as document version
designed to be statically designed with memory manager.
comfortable for peak load. In virtualization
eventualities (cloud computing, server consolidation), Likewise, all the documents square measure keep in
however, static peak provisioning of RAM to info additionally with the name, the system from that
applications dramatically reduces the potency and it's uploaded, the information and time of uploading
cost-saving edges of virtualization. also because the original file copy is additionally
unbroken within the server. The files uploaded from
III. METHODOLOGY the system in foreign places apart from the native
workplace if presents within the information
EXISTING WORK processing system, then the administrator are
intimated specified some files gift within the
The existing system introduces a load balance model information processing system isn't downloaded to
for the public cloud based on the cloud partitioning the native system. This helps the administrator to
concept with a switch mechanism to choose different synchronize the document expeditiously. The system
strategies for different situations, then after creating is implemented with the following methods:
all partitioning load balancing will starts to the all
node, these are main drawback.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


176
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

o Augmented Execution The applications area unit partitioned off into


o Elastic/Partitioned Modularized Applications elements such these may be remotely dead severally
o Ad-hoc Mobile Cloud on a cloud. With static partitioning, applications area
o Application Mobility unit partitioned off into mounted elements
o Cyber Foraging throughout compile time or runtime. In distinction,
Dynamic partitioning is that the partitioning of
Resource-intensive tasks can be relegated to the applications at runtime, sporadically (on interval
cloud or other resources such that the mobile device basis) or nonchalantly. Dynamic configuration
constraints are mitigated. In MCC both the data supported factors like CPU standing, memory, power,
storage and processing occur external to the mobile bandwidth and user preferences determines
device, while in cloud computing it is usually only wherever an application is run; whether or not on
the data storage which is external. the device or within the cloud server.

3.1 AUGUMENTED EXECUTION 3.3 AD-HOC MOBILE CLOUD

Processes on mobile devices area unit offloaded to the In Ad-hoc Mobile Cloud, the given task by a mobile
cloud so as to beat good device limitations, like node, the method execution is happened within the
process power, battery life and memory. During this shopper if the server is unavailable. Ad-hoc mobile
module thought-about the restricted Memory-Client/ cloud may be a set of mobile devices that replace the
Excessive Memory-Server. the first practicality cloud infrastructure by providing their computation
outsourcing (intensive computation is offloaded), resources to different devices. Mobile devices have
background augmentation (processes that don't need storage, procedure power, battery power and sensing
user interaction area unit offloaded), inject capabilities, that provides a chance for exploiting the
augmentation (changes within the execution path collective power of those devices. This is often
may be offloaded), hardware augmentation particularly helpful in things wherever there's very
(augmentation of underlying platform) Association little or no affiliation to the network, restricted
augmentation through multiplicity (running multiple power (short vary communications consume less
parallel copies of an application to create optimum energy).
decisions).
3.4 APPLICATION MOBILITY
The computationally overpriced tasks area unit full to
an online service containing image of the device, Application quality relies on the idea of method
whereas straightforward tasks area unit unbroken on migration, whereby processes may be paused,
the device. Once the net service in cloud task transferred to a distinct machine and seamlessly
completes the execution, results area unit integrated resumed. The distinction between this and increased
back to the mobile node. Execution is that in application quality, migration
will occur between completely different underlying
3.2 ELASTIC/PARTIONED MODULARIZED mobile architectures. The applying quality, one
APPLICATIONS method is dead in client/ different method is
processed within the server that is Cloud choice
Some of the processes are dealt in shopper and based mostly. it's engineered on a distributed filing
remaining process area unit are dealt from the server. system, that support the transmission of state from

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


177
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

one web site to a different, and also the modification to figure effectively in each non-public and public
in state from suspend to resume or contrariwise. The cloud.
purchaser’s area unit that internets suspend / resume
have association asynchronous dependence on the Furthermore, to gauge a best help in document
network, whereas purchasers would historically be management, the application become helpful if the
synchronously captivated with the network. below enhancements square measure created in
future. If the appliance is meant as internet service, it
3.5 CYBER FORAGING may be integrated in several internet sites. Non-
synchronized documents may be alerted to
The process or the given task of a node is partly dead synchronize. the appliance is developed specified on
in an un-trusted server. It's a way whereby proximate top of aforementioned enhancements may be
offered resources (called surrogates) area unit used integrated with current method.
for remote execution of applications. These resources
area unit dynamically discovered and used V. REFERENCES
notwithstanding the servers don't seem to be
expressly trustworthy by the user. Surrogates area [1]. N. Fernando, S. W. Loke, and W. Rahayu,
unit supported 2 main premises; un-trusted and "Mobile cloud computing: A survey," Future
unmanaged. This reduces the overall price of Generation Computer Systems, vol. 29, no. 1,
possession and maintenance and so encourages their pp. 84-106, 2015.
broad adoption. The offload tasks to a distant server [2]. D. Kovachev, Y. Cao, and R. Klamma, "Mobile
area unit supported many factors together with cloud computing: a comparison of application
computation quality, network strength, and device models," arXiv preprint arXiv:1107.4940, 2012.
capability and information neighborhood. [3]. [H. Qi and A. Gani, "Research on mobile cloud
computing: Review, trend and perspectives," in
IV. CONCLUSION Digital Information and Communication
Technology and it’s Applications (DICTAP),
Through this analysis, the information management 2012 Second International Conference on .
method becomes straightforward. All the regular IEEE, 2012, pp. 195-202
activities square measure assigned to them through [4]. Preeti Garg, Dr. Vineet Sharma, "Secure Data
browser interface. The administrator will read the Storage In Mobile Cloud Computing ," in
contents within the server choose a file and transfer Computer Communications Workshops
to the system where the administrator is functioning. (INFOCOM WKSHPS), 2011 IEEE Conference
Likewise, the files may be uploaded from the online on . IEEE, 2014, pp. 1060-1065.
page additionally terribly straightforward manner. [5]. Byung-Gon Chun , Sunghwan Ihm, Petros
The new system eliminates the difficulties within the Maniatis, "Securing elastic applications on
existing system. It’s developed in a very easy manner. mobile devices for cloud computing," in
The system is incredibly quick and any dealings may Proceedings of the 2009 ACM workshop on
be viewed or retaken at any level. Error messages Cloud computing security . ACM, 2016, pp.
square measure given at every level of input of 127-134
individual stages. Several of the tested practices and [6]. Weiming Zhao, "Dynamic Memory Balancing
technologies for managing trust relationships in for Virtual Machines ", 2016 14th International
ancient enterprise IT environments may be extended Conference on Modelling and Simulation

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


178
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

[7]. Weizhe Zhang C. Wang, K.Q. Yan, W.P. Liao


and S.S. Wang, Multiple Virtual Machines
Resource Scheduling , Proceedings of the 3rd
IEEE International Conference on Computer
Science and Information Technology, pp. 108-
113, 2015.
[8]. Christopher Clark , H. Jamal, A. Nasir, K.
Ruhana, K. Mahamud and A.M. Din, Live
Migration of Virtual Machines, Proceedings of
the Second International Conference on
Computational Intelligence, Modelling and
Simulation, pp. 160-165, 2014.
[9]. Albert M.K. Cheng, R. Subrata, and A.Y.
Zomaya, "LVMM: A lightweight virtual
machine memory management architecture for
virtual computing environment," Proc. 25th
IEEE Int’l Performance Computing and Comm.
Conf. (IPCCC ’06), 2016.
[10]. Tudor-Ioan Salomie, Gustavo Alonso, Timothy
Roscoe " Game-Theoretic Approach F Or Load
Balancing In Computational Grids ",IEEE
transactions on parallel and distributed systems,
vol. 19, no. 2, February y 2018.

Cite this article as :

R. Kavitha, G.Sathiyabama, "Improving Central and


Remote Access Units in Cloud Based Wi-Fi Network
for Better Client and Service ", International Journal of
Scientific Research in Computer Science, Engineering
and Information Technology (IJSRCSEIT), ISSN :
2456-3307, Volume 5 Issue 1, pp. 171-179, January-
February 2019. Available at doi :
https://doi.org/10.32628/CSEIT195118
Journal URL : http://ijsrcseit.com/CSEIT195118

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


179
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195122
Effective Prediction for Rock Burst Dataset Using Classification
Algorithms with Particle Swarm
Sadesh-S1, Banupriya-GK2
1Department of Computer Science and Engineering, Velalar College of Engineering &Technology, Erode,
Tamilnadu, India
2 ME, Department of Computer Science and Engineering, Velalar College of Engineering &Technology, Erode,
Tamilnadu, India

ABSTRACT

Rock burst and Slope Stability is one of the common failures in hard rock mining and civil construction. This
study focuses on the prediction of rock burst and Slope Stability classification with case instances using cloud
models and attribution weight. First, cloud models are introduced briefly related to the rock burst and Slope
Stability classification problem. Then, the attribution weight method is presented to quantify the contribution
of each rock burst and Slope Stability indicator for classification. In addition, analysis and prediction of slope
stability is of great importance in geotechnical engineering. With the development of economics, the number
of slopes is increasing and landslides caused by slope instability have become one of the three major geological
disasters in the world along with earthquakes and volcanoes. To reduce or prevent landslide damage, slope
stability analysis and stabilization are required. However, accurately predicting slope stability is challenging
because of the complexity of slope structures and the difficulty to determine the precise input data associated
with key geotechnical parameters the proposed methodology PSO feature extraction preserves important
distance relationships, such as : The Random forest, Naive Bayes of each object of the original dataset. This
leads to preservation of any mining operation that depends on the ordering of distances between objects, such
as Random forest, Naive Bays -search, SVM, J.48 and MLP classification, as well as many visualization
techniques. In particular, it establishes a restricted isometric property, i.e., tight bounds on the
contraction/expansion of the original distances.
Keywords: Raspberry Pi, USB Camera, Motor Driver, Motor, LCD Display, Road Side Speed Sign Controlling.

I. INTRODUCTION correlation or patterns among dozens of fields in the


large relational database. [14]
Data mining is analyzing data from different
perspective and summarizing it into useful ISSUSES IN DATA MINING
information. The information can be used to increase
revenue, cut cost, or both. Data mining allows the Mining different kinds of knowledge in databases
user to analyze data from many different dimensions Data mining is used to cover a wide spectrum of data
or angles, categorize it, and summarize the analysis and knowledge discovery task. Different user
relationships identified. It is the process of finding can use different kinds of knowledge.

CSEIT195122 | Received : 25 Jan 2019 | Accepted : 05 Feb 2019 | January-February -2019 [ 5 (1) : 180-184 ]
180
Sadesh S, Banupriya GK Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 180-184

Interactive mining at multiple levels of abstraction XIAO FAN and KAN WANG [2] In the field of
It is difficult to know what can be discovered within nuclear engineering, deterministic and stochastic
database. The Data mining process should be methods are used to solve radiation transport
interactive. Database containing a huge amount of problems. To solve the transport equation for the
data, appropriate sampling techniques can be used to average particle behavior and also contain
facilitate interactive data exploration. [9] uncertainties associated with the discretization of the
independent variables such as space, energy and
Pattern evaluation angle of the transport equation and can admit
It can uncover thousands of patterns. Many of the solutions that exhibit non - physics features. The
patterns discovered may be uninteresting to the given Monte Carlo method obtains s results by simulating
user they present common knowledge or lacking individual particles and recording some aspects of
novelty. their average behavior. This method enables detailed,
explicit geometrical, energy and angular
Handling noisy or incomplete data representations and hence is considered the most
The data stored in a database may reflect noise such accurate method presently available for solving
as exceptional cases or incomplete data objects. The complex radiation transport problems. One of the
accuracy can be poor. Data cleaning and data analysis difficulties associated with Monte Carlo method is
method is used to handle the noisy data. the amount of computer time required to obtain
sufficient precision in the simulations.
II. LITERATURE SURVEY
QIANG YANG AND PING- AN DU [3] To present a
CHUN'AN TANG AND JOHN A. HUDSON [1] he novel approach for determining the weights of
subject of rock failure has been studied in a decision makers (DMs) based on rough group
coordinated way since the 1960s.It can be studied by decision in multiple attribute group decision-making
examination of natural rock formations that have (MAGDM) problems. At first, we construct a rough
been stressed and strained over geological time, by group decision matrix from all DMs’ decision
laboratory experiments on the rock samples in situ matrixes on the basis of rough set theory. The second
experiments takes place, and by observing the results process we derive a positive ideal solution (PIS)
of rock excavation and loading during engineering founded on the average matrix of rough group
construction. In this book, rock failure mechanisms decision, and negative ideal solutions (NISs) founded
explanations and illustrations are given. Over the on the lower and upper limit matrixes of rough group
years, there have been three main developmental decision. After that we obtain the weight of each
phases supporting rock engineering design: analysis group member and priority order of alternatives by
based on elasticity theory; the use of rock mass using relative closeness method, which depends on
classification systems; and computer modeling. The the distances from each individual group member’
elasticity theory approach is more useful because it decision to the PIS and NISs. Comparisons with
enables the stresses around circular and elliptical existing methods and an on-line business manager
holes to be determined, although the approach is with the selection example, the proposed method
most useful for deep excavations where the rock show that it can provide more insights into the
behavior is essentially elastic. subjectivity and vagueness of DMs’ evaluations and
selections.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


181
Sadesh S, Banupriya GK Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 180-184

MATTHIJS J. WARRENS [4] The kappa coefficient, ZONG- XIAN ZHANG [7] are two missions for rock
denoted by κ, is widely used as a descriptive statistic mechanics to accomplish in mining engineering: (1)
for summarizing the cross-classification of two to destroy rock efficiently; (2) to make rock
variables with the same unordered categories. structures safe. The missions are completed and
Originally proposed as a measure of agreement mining operations are well managed, best mining
between two raters classifying subjects into mutually results should be achieved. For accomplishments of
exclusive categories, Cohen’s κ has been applied to the two missions, rock mechanics faces following
square cross-classifications encountered in challenges: (1) how to make drilling, crushing and
psychometrics, educational measurement, grinding more efficiently, in particular for grinding
epidemiology, diagnostic imaging, map comparison, whose energy efficiency is less than 1%; (2)now how
and content analysis. The popularity of Co- hen’s κ to make full use of explosive energy and destroy rock
has led to the development of many extensions, effectively; (3) how to manage, reduce and finally
including multi-rater kappas, kappas for groups of predict seismic events and rock bursts; (4) how to the
raters, and weighted kappas. The value of κ is 1 when develop various mining methods; (5) how to reduce
perfect agreement between the two observers occurs, borehole damage in deep mines or in the mines with
0 when agreement is equal to that expected under high in - situ stresses; (6) how to increase ore
independence, and negative when agreement is less recovery and decrease dilution; (7) how to improve
than expected by chance. mining safety; (8) how to make rock support de signs
more scientifically. All these challenges will be
YU ZHOU AND TINGLING WANG [5] Rock burst is analyzed in this paper. In addition, few topics such as
one of main engineering geological problems greatly rock mass classification, environment protection and
threatening the safety of construction. Prediction of the effects of loading rates, temperatures, and cyclic
rock burst is always to important for the safety of loading on mining engineering will be discussed.
workers and equipments in tunnels. Here in this
paper, PNN-based rock burst prediction model is III. METHODOLOGY
proposed to determine whether rock burst will
happen in the underground rock projects and how NORMALIZATION PROCESS
much the intensity of rock burst.
It is the process of classify data into an associated
ZAOBAO LIU and JIANFU SHAO [6] Rock burst is table it also eliminates redundancy and increases the
one of the common failures in hard rock mining and reliability which improves output of the query. For
civil construction. It focuses on the prediction of rock normalizing a database, we divide the dataset into
burst classification with the case instances using tables and establish relationships between the
cloud models and attribution weight. At first, the tables.The dataset normalization can essentially be
cloud models are introduced briefly related to the defined as the practice of optimizing table structures.
rock burst classification problem. Now, the Optimization is accomplished with the result of a
attribution weight method is presented to quantify thorough investigation of the various pieces of data
the contribution of each rock burst indicator for that will be stored within the database, in particular
classification. This approach is implemented to concentrating upon how this data is interrelated
predict the classes of rock burst intensity for the 164
rock burst instances collected.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


182
Sadesh S, Banupriya GK Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 180-184

Min Max Normalization


SYSTEM FLOW DIAGRAM
Min max normalization is a normalization strategy
which linearly transforms x to y= (x-min)/(max-min),
where min and max are the minimum and maximum
values in X, where X is the set of observed values of x.
It can be easily visible when x=min, then y=0.

y=x-(min(x)) / (max(x)-min(x))

FEATURE SELECTION

HCR-PSO feature extraction model for Rock Burst


dataset and applied an improve probability in many
Geo-graphical application such as training artificial
neural networks, linear constrained function
optimization, wireless network optimization, data
classification, and in many other areas where GA is
applied. Computation in HCR-PSO is based on a
swarm of processing elements called particles in
which each particle represent a candidate solution.
The system is initialized with a Rock Burst dataset
swarm of random solutions and searches for optima
by updating Rock Burst dataset generations. The
search process utilizes the combination of
deterministic and also tbe probabilistic rules that
depend on information sharing among their
population members to enhance their search
IV.CONCLUSION
processes. Rock Burst dataset prediction system
sharing mechanism in HCR-PSO is considerably
This project analyzed the Rock Burst dataset using
different.
algorithms such as J48, MLP, SVM, Random Forest,
and Bayesian Classification. These algorithm gives
In GAs, chromosomes share information with each
various result based on HCR-PSO feature extraction
other, so the whole Rock Burst dataset moves like
model. It has been seen that MLP and J48
one group towards a selected area. In HCR-PSO, the
Classification gives better results compare to other
global best swam particle found among the swarm is
classification algorithms. There are many criterions
the only Rock Burst dataset shared among particles. It
for evaluating the selected feature subset; here this
is a one - way Rock Burst dataset prediction sharing
project used features to evaluate the performance of
mechanism. The Rock Burst dataset prediction
different classification algorithm. In future, we have
computation time in HCR-PSO is much less than in
attempted to classify different feature selection
GAs because all swam particles in HCR-PSO tend to
algorithms into four groups: complete search,
meet to the best solution fast.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


183
Sadesh S, Banupriya GK Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 180-184

heuristic search, meta-heuristic methods and [8]. Tan P.-N., Steinbach M. and Kumar V.,
methods that use artificial neural network. The "Introduction to Data Mining", Addison
future methodology is used to analyze the Rock Burst Wesley, 2006
dataset region into separable compartments i.e. Rock [9]. Weber.M, Welling.M, and Perona.P.
Burst dataset etc. However, the method requires "Unsupervised Learning of Models for
further improvement mostly regarding feature Recognition
selection of the Rock Burst dataset into multiple
components: renal cortex, renal column, renal Cite this article as :
medulla and renal pelvis. Apart from that, this is
planned to expand the database on which the system Sadesh S, Banupriya GK, "Effective Prediction for
will be tested. And also, the proposed method in this Rock Burst Dataset Using Classification Algorithms
project can be employed for detecting the Rock Burst with Particle Swarm", International Journal of
dataset Geo-graphical in future with the Rock Burst Scientific Research in Computer Science, Engineering
dataset and classification of the Geo-graphical. and Information Technology (IJSRCSEIT), ISSN :
2456-3307, Volume 5 Issue 1, pp. 180-184, January-
V. REFERENCES February 2019. Available at doi :
https://doi.org/10.32628/CSEIT195122
[1]. Akay M.F. "Support vector machines combined Journal URL : http://ijsrcseit.com/CSEIT195122
with feature selection for breast cancer
diagnosis" (2009) Expert Systems with
Applications, 36 (2 PART 2) , pp. 3240-3247.
[2]. Cristianini N and Shawe-Taylor J., "An
Introduction to Support Vector Machines and
Other Kernel-based Learning Methods",
Cambridge University Press, 2000.
[3]. Fergus.R, Fei-Fei.L, Perona.P, and Zisserman.A.
"Learning Object Categories from Google’s
Image Search".
[4]. Han, Jiawei; Kamber, Micheline (2001). Data
mining: concepts and techniques. Morgan
Kaufmann. p. 5. ISBN 978-1-55860-489-6.
[5]. Han, Kamber, Pei, Jaiwei, Micheline, Jian (June
9, 2011). Data Mining: Concepts and
Techniques (3rd ed.). Morgan Kaufmann. ISBN
978-0-12-381479-1.
[6]. Pal S.K. and Mitra P, "Pattern Recognition
Algorithms for Data Mining", CRC Press, 2004.
[7]. R language and environment Hornik, Kurt
(November 26, 2015). "R FAQ". The
Comprehensive R Archive Network. 2.1 What
is R. Retrieved 2015-12-06.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


184
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195140
Enhancement in the Lifetime of Wireless Sensor Network
Bhumika Yadav1, Ram Neresh Sharma2
1Computer Science Engineering Department, Maharana Pratap College of Technology, Madhya Pradesh, India
Assistant Professor, Computer Science Engineering Department, Maharana Pratap College of Technology,
2

Madhya Pradesh, India


ABSTRACT

Technology changes every day and everyone wants to add more comfort in our life but some technologies
increase the safety. Wireless Sensors are one of the technologies which increase the safety in our life in some
manner. Wireless sensors are used in remote areas where human cannot survive to collect results of atmosphere,
temperature and pressure. All the technical devices are work on electricity/battery and in remote area it is
impossible to change batteries or provide continues power supply to sensors that is why we want to increase
the lifetime of wireless sensor network in this paper. In this paper we are explaining our approach which is
developed by us to increase the lifetime of wireless sensor network as well as we are comparing its results to
some other technologies which are already present now a days.
Keywords : EES Scheme, LBM, REAR, Hybrid, Network Topology, Routing, Random Event Model

I. INTRODUCTION To validate our proposed schemes the simulation are


running on 3 different network with different size
In this paper I am evaluating the performance of my and also with randomly distributed sensor nodes:
proposed approach and compare the sink lifetime
achieved with existing EES schemes to place sink. 1. 160mX160m square sensing area with 100 sensors
Firstly I will describe my experimental setup to 2. 220mX220m square sensing area with 200 sensors
calculate the performance and later we discuss the 3. 360mX360m square sensing area with 500 sensors
results.
I also consider a static or uniform grid topology in
addition to a random topologies, so that all the
Experimental Setup
sensors are evenly distributed in the network area.

In this experiment I am going to measure how the


network lifetime improves which is under a set of
random events. The Coverage area of each event,
start time, location and duration is randomly
generated. Simulation is programmed with C
language and running on Xcode in Mac OS C Lion. I
have also used a number of different-different size
networks and for each network size I am randomly
generate several sensor distributions.

CSEIT195140 | Received : 25 Jan 2019 | Accepted : 05 Feb 2019 | January-February -2019 [ 5 (1) : 185-190 ]
185
Bhumika Yadav, Ram Neresh Sharma Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 185-190

Below mentioned grid topologies are used in our


simulation

1. 160mX160m square sensing area with 100 sensors


2. 220mX220m square sensing area with 225 sensors
3. 360mX360m square sensing area with 400 sensors

Random Event Setup

For every set of simulation I consider 500 random Random event launch sequence
events. The attributes of events such as start and end
time etc are generated randomly and stored in a file Other Parameters
as an input to simulator. In performance evaluation
comparison in EES with 50 events in the simulation,
my experiment has higher occurrence rate because I
am considering the 500 random events are put into
launce sequence during one simulation. The coverage
are for each event varies from 6m to 30m in the
network and the duration of the event varies from 1
unit time to maximum 3000 unit time in the system.
The actual duration of an event will depend on Algorithm
specific application and the time may vary in
millisecond, or in several minutes or in several hours. LBM can be represented as
• S = the set of all sensors in the network
In each time interval many events can start, many • pfinal = (Xs, Ys) the final position of the sink node
can end or many can continue. To define each time • ptentative = (X’s, Y’s) the tentative position of the
unit of the simulation the simulator uses a timeline, sink node (for finding routes)
event file access, read Startn and Endn of every event • pi = (xi, yi) the position of sensor node 𝑠𝑠𝑖𝑖∈𝑆𝑆.
and place them into corresponding sequence of • bi = Total number of bits transmitted by sensor node
launch for every time interval. When the simulation 𝑠𝑠𝑖𝑖∈𝑆𝑆.
start simulation process and the simulation reach the • wi = Total number of bits generated by sensor node
time interval then it will generate pop event in the 𝑠𝑠𝑖𝑖∈𝑆𝑆 .
sequence and perform start or end activities. • dij = distance from node si to node sj
• cij = cost of edge e(i& j)
• dmax = maximum transmission range of a sensor
node
• rij = 1 if and only if node si transmits its data
directly to node sj
Algorithm 1:-
Step 1: Calculate tentative sink position ptentative =
(X’s, Y’s

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


186
Bhumika Yadav, Ram Neresh Sharma Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 185-190

Based on EES [51], use equations (3) and (4) to • Metric 1: First sensor node depletes energy
compute values for X’s and Y’s • Metric 2: 10% of total sensors depletes energy
Step 2: Calculate paths • Metric 3: Network becomes disconnected, i.e.
Use Dijkstra’s algorithm [52] to compute the path at least one sensor node cannot find a valid
from each sensor 𝑠𝑠𝑖𝑖∈𝑆𝑆 to sink node at (X’s, Y’s) path to route data back to the sink
Step 3: Determine load on each sensor node

Load on node-
Step 4: Calculate final sink position

For the LBM algorithm

Fig 2 – Network Lifetime in network with 100


Finally the new location of the sink is determined in sensors
step 4

Result Analysis and Performance Fig 3 – Network Lifetime in network with 200
sensors
I am comparing 4 different strategies in our
comparisons
1. Location based sink placement (EES)
2. Load based sink placement (LBM)
3. Residual Energy aware routing (REAR)
4. Hybrid approach that combines both LBM
and REAR (Hybird).

For each network size, we generated 5 different Fig 4 – Network Lifetime in network with 500
random topologies, and tested each topology with 5 sensors
sets of random events sequences. So, the results Comparison of lifetimes for grid topologies
reported for the random networks are averages from
the 25 runs. As mentioned earlier, different
researchers have used different metrics to measure
the lifetime of a wireless sensor network. In our
simulations, we have taken the following commonly
used metrics to measure network lifetime:

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


187
Bhumika Yadav, Ram Neresh Sharma Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 185-190

Fig 5 – Network Lifetime in grid with 100 sensors Fig 8 – Network lifetime under different maximum
transmission range in network with 200 sensors

Fig 6 – Network Lifetime in grid with 225 sensors


Fig 9 – Metric 1 under different maximum
transmission range in network with 100 sensors

Fig 6 – Network Lifetime in grid with 400 sensors

Fig 10 – Metric 1 under different maximum


Effect of transmission range
transmission range in network with 200 sensors

Fig 7 – Network lifetime under different maximum


transmission range in network with 100 sensors
Fig 11 – Metric 1 with no maximum transmission
range.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


188
Bhumika Yadav, Ram Neresh Sharma Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 185-190

II. CONCLUSION [5]. Luo.F. Relay node placement and trajectory


computation of mobile data collectors in
In this paper I have created an algorithm in which I wireless sensor networks. University of
have introduce two new methods in network lifetime Windsor, Windsor, Master Thesis 2011
in random event and time-based methods. In these [6]. Akyildiz, I.F., Su, W.,Sankarasubramaniam,
kind of cases rand event happens regularly and end Y.,Cayirci, E. Wireless sensor networks: a
for a specific time slot. In the proposed scheme it is a survey. Computer Networks, (2002), 393-422.
mixture of LBM and REAR Technology which can [7]. W. Heinzelman, A. Chandrakasan and H.
balance the lifetime of wireless sensor network in Balakrishnan.Energy efficient communication
much more efficient manner. In LBM method, it protocol for wireless micro-sensor networks.
takes load of transmission and adjust the sink node HICSS, (2000) 3005- 3014
position when an event occurs and the sink node is [8]. J. Pan, Y. T. Hou, L. Cai, Y. Shi, and S. X. Shen.
very heavily loaded. REAR technology is very helpful Topology Control for Wireless Sensor
in the calculation to find routing path in the network. Networks. International Conference on Mobile
Computing and Networking, (2003) 286-299.
We have used a simulator or we can call it custom [9]. ZhiAngEu; Seah, W.K.G. Impact of
simulator to calculate the performance of the Transmission Power And Routing Algorithms
proposed algorithms and also compare our results to in Designing Robust Wireless Sensor Networks.
EES technology. In the above-mentioned result, all Personal, Indoor and Mobile Radio
the graph shows that our method performs better Communications, (2007) 1-5 56
then EES technology continuously when only sink [10]. Y. Bi, L. Sun, N. Li. 2009. BoSS: a moving
node move from one place to another. We have also strategy for mobile sinks in wireless sensor
found that a particular matric is calculating the networks. International Journal of Sensor
lifetime has very significant effect on performance. Networks, 5(2009), 173-184
[11]. Nazir, B. and Hasbullah, H. Mobile Sink based
III. REFERENCES Routing Protocol (MSRP) for Prolonging
Network Lifetime in Clustered Wireless Sensor
[1]. Romer, K and Mattern, F. The Design Space of Network. Computer Applications and Industrial
Wireless Sensor Networks. IEEE Wireless Electronics, (2010), 624 - 629.
Communications, 11(6), (2004) 54-61 [12]. G. Gupta, M. Younis. Fault-tolerant clustering
[2]. D. Estrin, R. Govindan, J. Heidemann, S. of wireless sensor networks. Proceedings of
Kumar, Next century challenges: scalable IEEEWCNC, (2003), 1579 - 1584
coordination in sensor networks, ACM [13]. G. Gupta, M. Younis. 2003. Performance
MobiCom’99, (Washington, USA, 1999), 263- evaluation of load-balanced clustering of
270. wireless sensor networks. 10th International
[3]. Bari. A. Relay node in Wireless Sensor Conference on Telecommunications, (2003),
Network: a survey. Computer Science, 1577-1583
University of Windsor, Windsor, (2005) [14]. Bari, A and Jaekel, A, Techiques for Exploiting
[4]. Chong.C.Y. andKumar.S.P. Sensor Networks: Mobility in Wireless Sensor Networks.
Evolution, Opportunities and Challenges. Handbook of Research on Mobile Business:
Proceeding of the IEEE, 91(8), (2003) 1247- Technical, Methodological and Social
1256 Perspective, (2008), 445-455

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


189
Bhumika Yadav, Ram Neresh Sharma Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 185-190

[15]. Tian, K., Zhang, B., Huang, K. and Ma, J. Data


Gathering Protocols for Wireless Sensor
Networks with Mobile Sinks. IEEE Global
Telecommunications Conference, (2010), 1 - 6.
[16]. Wang, B., Xie, D., Chen, C., Ma, J. and Cheng,
S. Deploying Multiple Mobile Sinks in Event-
Driven WSNs. Communications, (2008), 2293 -
2297.

Cite this article as :

Bhumika Yadav, Ram Neresh Sharma, "Enhancement


in the Lifetime of Wireless Sensor Network",
International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 185-190, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195140
Journal URL : http://ijsrcseit.com/CSEIT195140

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


190
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195155

A Review Paper on IoT Based Smart Irrigation System


Anjali Dokhande1, Chetna Bomble1, Rakshanda Patil1, Puja Khandekar1, Nayan Dhone1,
Prof. Chandrashekhar Gode2
1BE Students, Department Computer Science and Engineering, Priyadarshini J.L. College of Engineering,
Nagpur, Maharashtra, India
2 Assistant Professor, Department Computer Science and Engineering, Priyadarshini J.L. College of Engineering,
Nagpur, Maharashtra, India

ABSTRACT

Automation of farm activities can transform agricultural domain from being manual and static to intelligent and
dynamic leading to higher production with lesser human supervision. This paper proposes an automated
irrigation system which monitors and maintains the desired soil moisture content via automatic watering.
Microcontroller ATMEGA328P on arduino uno platform is used to implement the control unit. The setup uses
soil moisture sensors which measure the exact moisture level in soil. This value enables the system to use
appropriate quantity of water which avoids over/under irrigation. IOT is used to keep the farmers updated
about the status of sprinklers. Information from the sensors is regularly updated on a App using WiFi modem
through which a farmer can check whether the water sprinklers are ON/OFF at any given time. Also, the
sensor readings are transmitted to a Thing speak channel to generate graphs for analysis.
Keywords : Irrigation System, Soil Moisture Sensor, Temperature Sensor, Humidity Sensor, IOT, Android

I. INTRODUCTION promising solution to farmers where the presence of


farmer in the field is not compulsory. Now a day's
In our country, agriculture is a major source of food internet is widely used. Using internet farmer know
production to the growing demand of the human about the agriculture field irrigation status. Thus,
population. In agriculture, irrigation is an essential mobile applications will be helpful in fulfilling this
process that influences crop production by supplying purpose. It helps farmers to know the status of farm
water to the needed land. Farmers have to visit their field moisture and temperature through a mobile app
land to check how much amount of water is required and it will make the farmer at some remote places to
for their field. This irrigation method takes a lot of decide whether he needs to water the field or not. The
time and effort particularly when a farmer needs to soil moisture sensor is used for sensing moisture level
irrigate multiple agriculture fields distributed in to find out whether the soil is dry or wet. The
different geographical areas. Traditionally farmers will moisture sensor and temperature sensor are interfaced
present in their fields to do irrigation process. But with Arduino microcontroller. Based on that sensor
nowadays farmers need to manage their agricultural values, the controller will activate the DC motor or
activity along with other occupations. Automation in pump by sending the commands from the mobile
irrigation system makes farmer work much easier. phone that will operate the pump through the relay.
Sensor-based automated irrigation system provides a Water is a very precious resource and must be

CSEIT195155 | Received : 25 Jan 2019 | Accepted : 05 Feb 2019 | January-February -2019 [ 5 (1) : 191-196 ]
191
Anjali Dokhande et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 191-196

properly utilized. Agriculture is one of those areas


which consume a lot of water. The aim of the paper is
to develop an intelligent irrigation system controller
which measures the moisture of the soil and
temperature of the atmosphere and helps to take
decision accordingly to turns on or off the water
supply system. It is important to know the amount of
water that has been used in the irrigation process
during agriculture. This paper describes a simple
system by using Arduino microcontroller and Android
software to automate the existing manual irrigation
system.

II. RELATED WORK Fig 1. water holding property of soil

A. Moisture Measurement This stage is called as the field limit. Void spaces are
currently loaded with water and air parcels (Fig. 1(b)).
Soil is comprised of blend of segments including Each harvest has a Critical Soil Moisture Deficit level,
mineral and natural particles, with water and air enabling soil to dry out past this level, water pulling
making up the spaces in the middle. Soil can be from edit can't occur, diminishing the yield. Advance
essentially characterized into following 4 classes: more water expulsion from soil lead little holding of
water by soil particles much firmly on account of
surface strain impact for the yield to remove this is
said to be as withering point as appeared in figure 1(c).
The accessible water limit is the measure of water a
dirt can make accessible to plants, by and large
It is encouraged to break down soil to reason its
characterized as the contrast between measure of
classification. Every classification has distinctive
water put away in a dirt at field limit and the measure
properties henceforth their water holding limit
of water put away in the dirt at the lasting shrinking
changes from one sort to the next. As water invades
point. Plants get the vast majority of water from the
soil, it begins to fill the hole between the void spaces
upper part of the root zone.
in the middle of soil particles (Fig. 1(a)), when every
one of the spaces are totally immersed with the water,
The term viable root zone alludes to about the upper
the state is known as immersion point. This state goes
portion of the root zone profundity, where around 70%
on for brief time. With time overabundance water
of harvests water is taken up.
permeates descending through water profile because
of gravitational power. At same time narrow activity
give inverse power to gravity and give adjusted
condition so descending development of water is
ruined.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


192
Anjali Dokhande et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 191-196

changes relying upon the air temperature. More


sizzling air, for instance, can hold more moisture.

III. LITERATURE REVIEW

For the development of automated irrigation structure,


soil moisture content is increasingly basic parameter
when stood out from others as it has a significant part
in plant improvement segment and availability of
water for irrigation is genuine stress for the farmers
uncommonly the ones who are dependent on rain. In
this manner, water organization has a high need while
arranging an automated irrigation system as saw in
most of the composition. The experts have used
Fig 2. Effective root zone
distinctive systems to evaluate moisture substance
precisely. The electrical conductivity estimation is the
Numbers of techniques are produced for soil moisture
most fundamental, sagacious and control capable
estimation from straightforward feel strategy to most
strategy for all. In any case, it isn't correct and its
progress electronic ones.
results vacillate after some time. In spite of its burdens,
it is comprehensively supported by various
B. Temperature Measurement
investigators. [5] Have used this technique to execute
Temperature monitoring is key in numerous modern
an automated irrigation control structure using spill
situations. It likewise assumes imperative part in plant
irrigation framework. The data was accumulated and
development subsequently monitoring temperature is
arranged by the ARM7 board. To give UI GSM
basic for good farming practice. Numerous standard
technique was used and at the customer end android
methods exist which relies on estimation of physical
based application was created to demonstrate data and
properties of the working material that differs with
with that information, the customer can pick what
temperature. Thermocouple, thermistor, RTD,
move to make. According to the call given by the
pyrometer, Langmuir tests, infrared, and so forth are a
customer, the solenoid valve was attempted to control
portion of the cases.
the irrigation. An abnormal state of flexibility was
obtained with this execution. The remaining burden
C. Humidity Measurement
for the farmer was decreased and moreover increase in
There are three approaches to speak to Humidity. It is
the benefit of the estate was viewed.
the measure of water vapour (water that has
abandoned a fluid to an undetectable gas) noticeable
In [6] acoustic methodology has been used to measure
all around. Outright humidity is the genuine measure
water substance of the earth in transit that
of water vapour in a predetermined volume of air.
development time of sound wave is assorted in dry
Relative humidity is the proportion of moisture
and wet soil. In light of the discernment, the chart
noticeable all around when contrasted with the most
was set up for development time versus moisture
extreme measure of moisture the air can hold, which
content from which moisture can be found. Another

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


193
Anjali Dokhande et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 191-196

methodology for moisture estimation is assessing moisture sensor (VH400) in the perspective of
assortment in dielectric steady [7]. Dielectric electromagnetic estimation and second is soil
unfaltering of water is (~80) high when appeared temperature sensor (DS1822). The recorded data was
differently in relation to dry soil (2-3). In [8] secured locally in the memory chip and was moreover
assortment of the dielectric was assessed as an transmitted to the web. The system is controlled
assortment in the capacitor using capacitor and capable and thoroughly wears down sun controlled
resistor interface. The response is generally immediate essentialness. The promising delayed consequence of
which is required in case of exactness agriculture. S. S. around 90% water saving was viewed. P. Bhosale and
Mathurkar and D. S. Chaudhari [9] showed a model in V. Dixit made atmosphere monitoring structure. They
perspective of moisture, temperature and humidity used a broad assortment of sensors for monitoring
sensor. All of the sensors were balanced for the specifically ecological temperature and humidity
straight response. The standard purpose of the (SHT1x), Soil temperature (LM35), Radiation and
structure is to develop an exact system which can be sunshine, soil moisture (gypsum square in perspective
utilized in real estate and offer preferences to the of the resistive framework), wind speed and course
farmer. (anemometer) and rainfall. In control board, PIC
microcontroller was used. The total data was secured
An epic method to manage design automated in memory card amassing and was also sent to the
irrigation system is the use of Plant water weight remote customer by methods for the GSM module.
examination. It is overcome selection of optical and IR Powerful usage of water was practiced through this
pictures of plant overhang. This particular enrollment structure [12].
speaks to a couple of challenges as no anticipated
essential segment or right organizing can be found IV. PROPOSED METHODOLOGY
from the data pictures. X. Wang et al [10] developed
an Automatic Cross-Correlation course of action Nowadays agricultural field is facing lot of problems
count which uses the information of smart picture due to lack of water resources. In order to help the
structure yet clears out the effect of picture shading farmers to overcome the difficulties, smart irrigation
and power in the relationship method and thusly system has been used. In this system, various sensors
achieves a pleasing enlistment result. They similarly such as pH, soil moisture and DHT11 are connected to
realized capable computation approach which can the input pins of Arduino microcontroller. The sensed
fundamentally diminish the estimation eccentrics of values from the sensors are displayed in LCD. If the
the ACC count while keeping up needed exactness. sensed value goes beyond the threshold values set in
Game plan control was also improved by accepting N- the program, the pump will be automatically switched
maxima procedure in the control point count. ON/OFF by the relay circuit and it is connected to the
Preliminary comes about showed that the proposed driver circuit which helps to switch the voltage. The
system defeated each and every other procedure for farmer will be intimated about the current field
region-based methods. J. Gutierrez et al [11] executed condition through Wi-Fi module and also updated in
amazingly capable automated system with remote the web page. By using this system, the farmer can
sensor arrange. The remote interface was given access the details about the condition of the field
through the GPRS module (MTSMC-G2-SP). The anywhere at any time.
structure has two information parameters at first is

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


194
Anjali Dokhande et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 191-196

V. COMPONENTS USED
• Wi-Fi Module
• Soil Moisture Sensor The ESP8266 Wi-Fi module is a self-contained SOC
Soil Moisture sensor is used to measure the moisture (System on Chip) with integrated TCP/IP
content present in the soil. When the soil moisture (Transmission Control Protocol/Internet Protocol)
value read by the sensor is above the threshold value, protocol stack that can give any microcontroller
low level (0V) will be the digital output and if it is access to any Wi-Fi network. Each ESP8266 module
below the threshold level, high level (5V) will be the comes preprogrammed meaning, it can be simply
digital output. The digital pin is used to directly read hooked up to Arduino device to get Wi-Fi ability. This
current soil moisture value to see if it is above module has a powerful enough on-boarding process
threshold or not. The threshold voltage can be and high storage capacity that allows it to be
regulated with help of potentiometer. integrated with the sensors and other application
specific devices.

VI. APPLICATION

• On different region of land: It can be used on


shallow soil-hilltops, hillsides and throughout
mountain regions.
• For Greenhouse: Smart Greenhouse parameter-
watering of crops, on fertilizers, temperature and
moisture level.
• Very Efficient for Paddy, Rice Fields-This plant
Fig 3. Moisture Sensor requires much water that’s why apply system on
their farm.
• DHT11 SENSOR • Irrigation in Fields: This is used to give watering
DHT11 sensor is used for measuring temperature and to the crop fields.
humidity. It uses a capacitive humidity sensor and a • Food production and safety etc.
thermistor to measure the surrounding air. This sensor
is cost effective, provides low power consumption and VII.ADVANTAGES
up-to 20 meter signal transmission is possible.
• Increase the productivity: Productivity on
farmland is going too increased.
• Reduce water consumption.
• No manpower required.
• Reduce soil erosion and nutrient leaching.
• Cost effective method.
• High quality crop production.
• System not damage by weathers and birds.
Fig 4. DHT11 Sensor
• Efficient use of water.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


195
Anjali Dokhande et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 191-196

VIII. CONCLUSIONS Instrumentation and Measurement, Vol. 53, No.


4, pp. 891-898, May 2004.
In light of all above talk it can be reason that [7]. Darold Wobschall, A frequency shift dielectric
programmed irrigation framework utilizing remote soil moisture sensor, IEEE Transactions on
innovation can give effective framework equipped for Geoscience Electronics, Vol. 16, No. 2, pp. 112-
moderating assets and human exertion. The 118, Apr. 1978.
framework likewise encourages ongoing remote [8]. S. Saxena and G. M. Tayal, Capacitive moisture
monitoring of the current ecological state of field. meter, IEEE Transaction on Industrial
Present day innovation can be consolidated to let Electronics and Control Instrumentation, Vol.
down the cost. These electronic frameworks are 28, No. 1, pp. 37-39, Feb. 1981.
control proficient thus devours less power and depend [9]. S. Mathurkar and D. Chaudhari, Smart Sensors
on auxiliary sources such sunlight based vitality for Based Monitoring System for Agriculture using
finish independence. Field Programmable Gate Array”, International
Journal of Innovative Technology and Exploring
IX. REFERENCES Engineering (IJITEE), Vol. 3, Issue 4, May 2013.
[10]. X. Wang, W. Yang, A. Wheaton, N. Cooley, and
[1]. Plant Growth Factors: Temperature, Colorado B. Moran, Efficient registration of optical and IR
State University, And Available (as on 14-09- images for automatic plant water stress
2015) at: assessment, Comput. Electron. Agricult., vol. 74,
http://www.ext.colostate.edu/mg/Gardenno- no. 2, pp. 230–237, Nov. 2010.
tes/143.html#heat. [11]. J. Gutiérrez, J. F. Villa-Medina, A. Nieto-
[2]. Plant Growth Factors: Water, Colorado State Garibay, and M. Á. Porta-Gándara, Automated
University, And Available (as on 14-09-2015) at: irrigation system using a wireless sensor
http://www.ext.colostate.edu/- network and GPRS module, IEEE transactions
mg/Gardennotes/144.html. on instrumentation and measurement, Vol. 16,
[3]. Harriot Bigas (Ed.), The Global Water Crisis: Issue 1, pp. 166-176, 2013.
addressing an urgent security issue‖, Paper for [12]. P. Bhosale and V. Dixit, Water saving-irrigation
InterAction Council, Hamilton, Canada: UNO- automatic agricultural controller, International
INWEH, 2011-12. Journal of Scientific and Technology Research,
[4]. A. I. Johnson, Methods of measuring soil Vol. 1, Issue 11, Dec. 2012.
moisture in the field‖, U. S. Geological Survey,
1992 (Third Reprint). Cite this article as :
[5]. V. Divya, A. Akhouri, C. Kumar, R. Rishabh, R. Anjali Dokhande, Chetna Bomble, Rakshanda Patil,
Bagla, A Real time implementation of a GSM Puja Khandekar, Nayan Dhone, Prof. Chandrashekhar
based automated irrigation control system using Gode , "A Review Paper on IoT Based Smart Irrigation
drip irrigation methodology, International System", International Journal of Scientific Research
Journal of Scientific and Engineering Research, in Computer Science, Engineering and Information
Vol. 4, Issue 5, pp. 146-151, May 2013. Technology (IJSRCSEIT), ISSN : 2456-3307, Volume 5
[6]. F. Adamo, An acoustic method for soil moisture Issue 1, pp. 191-196, January-February 2019.
measurement, IEEE transactions on Journal URL : http://ijsrcseit.com/CSEIT195155

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


196
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/ CSEIT195153

Marg-Sudhar : Pothole Reporting System Using Android Smart


Phone
Riya Agrawal, Payal Patel, Yash Telang, Ankita Bhandarkar, Irfan Mushtaq, Dr. S. V. Sonekar
Department of Computer Engineering, JD College of Engineering and Management, Nagpur, India

ABSTRACT

Marg-Sudhar: Pothole Reporting System is a mobile application for detecting and reporting road condition.
Dangerous road surface conditions are major distractions for safe and comfortable transportation and therefore
road damage needs to be reported by locality. Data can be gathered by clicking image to identify potholes and
other severe road surface anomalies using geotagging technique and upload on the server.
Keywords : Potholes, GPRS, Geotagging, Road Maintainance, Strategy i.e. Click, "Report" and "Fix".

I. INTRODUCTION caused due to potholes and government authorities


try hard to fix each and every pothole but due to
One major problem in developing countries is large amount of roads in India it has become rather
maintainence of road. Since India is a developing difficult to locate each and every pothole and get it
nation there is a constant demand for good quality fixed.
infrastructure, transportation and services. But since
India is a huge country with quite a sizable So to reduce time consumption in visiting the sites
population this problem still has not been addressed the describe system can be use to directly report to
in totality. the highway Authority about the potholes with the
help of android based application. By this a lot of
In order to improve the conditions of roads efforts time gets reduced of authorities and they get
began way back in the 1980s. It is during this time information about the where about of potholes and
that roads were built to link major highways, to the preventive measures can be taken on time and a
expand the width of existing roads and to construct lot of casualties can be reduced from occurring.
important bridges.

Most roads in India are narrow and congested with


poor surface quality and road maintenance need are
not satisfactorily meet.

Our topic basically concerns about the safety of road


users. Potholes, formed due to heavy rains and
movement of heavy vehicles, also become a major
reason for traumatic accidents and loss of human lives.
As it is seen that many disasters occurring on road are Fig. 1. Condition of roads with potholes

CSEIT195153 | Received : 25 Jan 2019 | Accepted : 05 Feb 2019 | January-February -2019 [ 5 (1) : 197-202 ]
197
Riya Agrawal et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 197-202

II. PROPOSED SYSTEM

Marg-Sudhar is a android based mobile application


for reporting potholes and various road anomalies.

It works on strategy that is “Click”, “Report” and


“Fix”. Using app Marg-Sudhar users can click picture
which use geotagging technique along with
description and this information gets uploaded on
server.

The pothole information obtained from pothole Fig. 2. Pothole detection system and its application
reporting system is sent to server and can be applied
to a pothole alert service and road management Application has been implemented on following
system. All uploaded information are shown as a list android features:
view on client side. Also, the obtained pothole
information is provided to Road Management System. • For designing part various activities and their
components are used.
III. IMPLEMENTATION • Communication between activities has been
implemented using Intent object. Intent is an
Marg-Sudhar has been implemented using following messaging object use to request an action from
technologies (1) Android (2) Java (3) SQL (4) another app component.
Geotagging (5) Web Technologies. • Components registration, is an entry point
through which the system or a user can enter into
(1) Android: app.
• Also, various background services rights has been
Android studio is the IDE for google’s Android OS used.
build on JetBrains’ InteliJ IDEA software and • This application needs various build app access
designed specifically for android development. The rights such as: Storage, Camera, Data access,
following features are provided in the current stable Internet access, GPRS access.
version: Gradel-Based build support, Android-specific • Data is Authenticated due to security purpose at
refactoring and quick fixes, template-based vizards to various point such as at time of login.
create common android designs and components, a
rich layout editor that allows user to drag-and-drop (2) Java:
UI components, build-in support for google cloud
platform, enabling integration with firebase cloud Java is an object-oriented, class-based, concurrent,
messaging and google app engine. secured and general-purpose computer-programming
language.

After front-end elements are finalized java classes and


packages are use to make them work.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


198
Riya Agrawal et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 197-202

(3) Database: (1) Client Side:

A database is an organised collection of data,


generally stored and accessed electronically from a
computer system where databases are more complex
they are often develop using formal design and
modelling techniques.

Android comes with in-built SQLite database. SQLite


supports all the relational database features.

(4) Geotagging:

Geotagging or GeoTagging, is the process of adding


geographical identification metadata to various media Fig. 3. Client Side Working
such as a geotagged photographh or video, websites, Client side involves user interaction. Using app user
SMS messages, QR Codes. This data usually consists can login/ register. If user gets successfully login or
of latitude and longitude coordinates, though they register they can click image of pothole. The snapped
can also include altitude, bearing, distance, accuracy image is automatically geotagged with latitude &
data, and place names, and perhaps a time stamp. longitude using GPS and GPRS. This image is sent on
There are two main options for geotagging photos: server along with some description added by user.
Additionally, user can also see lists of complaints
capturing GPS information at the time the photo is
added by other users and their status.
taken or "attaching" the photograph to a map after
the picture is taken. In order to capture GPS data at (2) Server Side:
the time the photograph is captured, the user must
have a camera with built in GPS or a standalone GPS
along with a digital camera.

IV. WORKING

Marg-Sudhar: Pothole reporting system is


implemented into two modules: (1) Client side, (2)
Server Side.
Fig. 4. Server Side Working

Image via app is uploaded to web server and


complaint is stored on server for further processing
and authenticity. Also admins get all the privileges
through this server. Confirmation SMS and
continuous status for these complaint are send to
uploader.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


199
Riya Agrawal et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 197-202

V. APPLICATION LAYOUT (3) Launch Complaint:

(1) Login/ Registration screen:

(2) Home Screen:

(4) List View:

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


200
Riya Agrawal et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 197-202

(5) Server: VIII. CONCLUSION

Marg-Sudhar will enable road maintainers to obtain


information regarding road condition easily with the
help of locality. Through this it will reduce time of
Municipal corporation in gathering information and
locality will get better roads in less time.

IX. ACKNOWLEDGEMENT

First, and foremost we would like to thank God for


the wonderful opportunities and challenges he has
given to us. We express our sincere gratitude to our
VI. ADVANTAGES & LIMITATIONS guide Dr. S.V. Sonekar for providing his valuable
guidance, patience and for encouraging us to do our
(A) Advantages: best. We wish to thank Prof. P. Lohe, Head,
• It becomes easy for road maintainers as well as for Department of Computer Science and Engineering
locality to monitor the road condition. for his valuable contribution in fulfilling the
• Smooth roads will lead to less vehicle damage and requirement related to the project.
government investment.
• Rapid increase in the rate of Smartphone, users X. REFRENCES
can help to identify Potholes easily.
[1]. Pavan Gorade, Digambar Karde, Poonam
(B) Limitations: Khilare, Abhishek Sontakke, Prof.Ranjana
• To complain on real time basis an android mobile M.Kedar, “Real Time Pothole Tracking System
with GPRS connection is required. Using Android Smart Phone”, International
• It is completely based on user interaction and Journal of Advanced Research in Computer
requires good internet connectivity. Engineering & Technology (IJARCET) Volume
3 Issue 5, May 2014.
VII.FUTURE SCOPES [2]. Salman, Waheed Ahmad, Riaz Ali, Salman
Saleem, “a research on mobile applications for
• Can be implemented in cities under Municipal location tracking through web server and short
Corporation. messages services (sms) “,vfast Transitions on
• It can also be implemented in automated manner Software Engineering Volume 7, Number 02,
using GPS sensors on vehicle July-August, 2015.
• Further research is necessary for improvement in [3]. Lizeth Chandi, Catarina Silva, “Mobile
image processing. application development process: A practical
experience”.
[4]. Seung-Ki Ryu,1 Taehyeong Kim,1 and Young-
Ro Kim2, “Image-Based Pothole Detection
System for ITS Service and Road Management
System”, Hindawi Publishing Corporation
Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com
201
Riya Agrawal et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 197-202

Mathematical Problems in Engineering Volume


2015.

Cite this article as :

Riya Agrawal, Payal Patel, Yash Telang, Ankita


Bhandarkar, Irfan Mushtaq, Dr. S. V. Sonekar, "Marg-
Sudhar : Pothole Reporting System Using Android
Smart Phone", International Journal of Scientific
Research in Computer Science, Engineering and
Information Technology (IJSRCSEIT), ISSN : 2456-
3307, Volume 5 Issue 1, pp. 197-202, January-
February 2019.
Journal URL : http://ijsrcseit.com/CSEIT195153

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


202
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195152

Smart Meals
Shivali Singh1, Aishwarya Narnaware1, Kirtikumar Jain1, Kash Patel1, Prof. S Dhawale2
1Department of Computer Science Engineering, Rashtrasant Tukadoji Maharaj Nagpur University,
Nagpur, Maharashtra, India
2 Professor, Department of Computer Science Engineering, J D College of Engineering and Management,
Nagpur, Maharashtra, India
ABSTRACT

Online business, also known as e-commerce or electronic commerce, refers to transaction and the activities that
have been performed by the customers and seller over the business is facilitated by the web based technologies.
Online food delivery system is one of the most popular online business products. Customers are able to view
and select their favourite food from the list, add to the cart, choose the delivery types, make payments and the
order is complete.
Keywords : Automated Food Ordering System, Dynamic Database Management, Global Positioning System

I. INTRODUCTION application front end is developed using .NET


technology and the back end is developed by using
With the online food ordering system, food is MySQL.
delivered online to the customer. This is possible
using electronic gadgets. Due to rapid growth in the Thus time as well as cost of the developers will be
use of internet and the technologies associated with it, reduced. In our system the user will first register and
several opportunities are coming up on the web. So login themselves into the database. So that he/she can
many business and companies are undertaking into then order their food and make their transaction
their business because of internet. happen.
II. PROBLEM STATEMENT
In this project, an efficient online food ordering
system is proposed. By using the shortest path To create a database of the eating joints falling within
algorithm, we have established a system to reduce the the origin and destination points of the journey and
time consumption of the customers. Also we have placing the orders as per the available menu at
used Heuristic algorithm for the same. In today’s particulars eating joint with the time of service. Our
world, Android and IOS operating system are more idea is to develop an application that provides the
popular. travellers, the meal without any time consumption.
Basically the application will contain the information
Different types of programming language is used for about the hotels i.e. the eating joints that will provide
such operating system. So, in order to mitigate these the food to the travellers.
problems, web base solutions can be used. Here we
came with a food delivery application, where the

CSEIT195152 | Received : 25 Jan 2019 | Accepted : 05 Feb 2019 | January-February -2019 [ 5 (1) : 203-206 ]
203
Shivali Singh et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 203-206

III. LITERATURE REVIEW by the restaurant owners to adopt information and


communication technologies like PDA, wireless
In [1] a food ordering system is proposed with LANs, costly multi-touch screens, etc. to enhance the
Delivery Routing Optimization using GPS dining experience. This paper highlights limitations
Technology and Google Maps. In this paper they based on PDA base food ordering system using an
have explained about the use of GPS and Google Map android smartphone or tablet as a solution. The
for ordering purpose. The major problem in this customer tablet, kitchen display connects with each
paper is routing optimization. The system also uses other through Wi-Fi. Orders did by the customer
GPS in Android mobiles and take advantage of then reach the kitchen module instantly. This
Google Map for map solution. wireless application is user friendly which improves
accuracy and efficiency for hotels by saving their
In [2] a Digital Ordering System for Restaurant using time, reduces human errors and provides customer
Android was proposed. In this application they feedback.
combine the hotel management system by web IV. PROPOSED SYSTEM
technologies. Digital Hotel Management combines
lots of system of restaurant and hotels such as To overcome the limitations of the above system, an
Ordering System Kitchen Order Ticket, Billing Online Food Ordering System based on web based
System Customer Relationship Management System technologies is proposed. Here .NET technology is
together. This combine solution can add or expand mainly used. Our main objective in the project is in
the software system in any size of hotel chain designing completely automated menu in hotels with
environment. This system increases the speed of the help of Android phones using Wi-Fi module and
service and quality. This system also increases to provide a user-friendly environment. The menu
attraction of many customers. will be displayed automatically on the customer
mobile application using wireless connectivity and
In [3] this system explained Implementing we can directly order the menu with the help of press
Customizable Online Food Ordering System Using on the menu. The greatest advantage of this system is
Web Based Application. Previously the process its flexibility.
requires that the customer has to be seated in the
hotels before starting. An alternative method of this ASP.NET is an open-source server side web
for the customer is that ‘Food Pre-Ordering System application framework which is designed for web
using Web Based Application’ in which customer can development to produce dynamic web-sites, web-
be able to do the order before they approach to the applications and web services.
hotel, the saved order can be confirmed by touching
the smartphones. The list of theie selected order There are following steps in our process which first
items shall be shown on the kitchen screen, and includes that the customer has to log in to our system.
when confirmed, order slip shall be printed for If he is new to the system, he need to first register
further processing. This solution provides the easy himself and then log in to the system. The next page
and convenient way to select the order transaction that will appear will be the menu card where the
from customers. customer needs to choose the items from the list and
place his order. This order is collected and
In [4] an Implementation of Smart Restaurant with e- acknowledge by our system. After this he has to fill
menu card is proposed. Lots of work have been done all regarding details like his credit card details, etc.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


204
Shivali Singh et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 203-206

Now the verification of the customer details will be shortest path. In this project through .NET
done to see that whether it is valid or invalid. So for technology we have created login page, checking our
this we have to contact the credit card company. password and through MySQL we created database.
After this if we find all the details correct in all After login our site will show variety of nearby hotels
means then the order is placed successfully. But if through which a customer can order the meals.
given details mismatch then the order will be
cancelled. In the future research, we suggest adding more
variable for routing optimization process such as type
Now, the last step is to make the product ready and vehicle, the size of food packages, holiday session the
get the delivery. driver license type owned by delivery service and the
maximum capacity of the vehicle type. These
ADVANTAGES additional variables will add the complexity to our
application. This system can also enhance to produce
• Online Ordering and easy to use for all. an online system by adding the email notification
• Freedom for demanding food. feature for order confirmation with official receipt.
• Food available on decided places.
• Remove time consumption. VI. ACKNOWLEDGMENT
• Meal comes with good packing.
• Hygiene has first priority. Special Thanks to our Head of Department CSE/IT
• Hassle-free food preparation. Prof Prasanna Lohe , for his valuable suggestions and
encouragement.
FLOW DIAGRAM
VII.REFERENCES

[1]. Roy Deddy Hasiolan Tobing, “A Food Ordering


System with Delivery Routing Optimization
Using Global Positioning System Technology and
Google Maps”, Internetworking Indonesia
Journal, vol. 8, 2016.
[2]. Apurava Joshi, Prachi Oke, Niranjan Jadhav,
Ashutosh Bhargave, Prof. Mr. S.R. Lahane,
“Digital Ordering System for Restaurant Using
Android”. In International Journal of Scientific
and Research Publications, Volume 3, Issue
4April 2013.
[3]. Varsha Chavan, Priya Jadhav, Sneha Korade,
V. CONCLUSION Priyanka Teli,Prof. Mr. R.B.Anpat
“Implementing Digital Restaurant and Inter-
In this project, we have worked on the A* algorithm Restaurant Navigation Using Smart Phone.” in
for GPS tracking, All Pair Shortest Path Algorithm International Journal of Computer Science and
for finding the nearest route of the customer, Mobile Computing, Volume 4. Issue 2, February
Heuristic Search algorithm for again to find out the 2015.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


205
Shivali Singh et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 203-206

[4]. Mayur D. Jakhete, Piyush C. Mankar,


“Implementation of Smart Restaurant with e-
menu card.” in International Journal of Computer
Application, Volume 119, June 2015.
[5]. N. M. Z. Hashim “Smart Ordering System via
Bluetooth” in International Journal of Computer
Trends and Technology (IJCTT) – volume 4 Issue
7–Month 2013
[6]. S. R. Patil “e-converse An affordable touch screen
solution to intrigue Dining experience”

Cite this article as :

Shivali Singh, Aishwarya Narnaware, Kirtikumar Jain,


Kash Patel, Prof. S Dhawale, "Smart Meals",
International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 203-206, January-February 2019.
Journal URL : http://ijsrcseit.com/CSEIT195152

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


206
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195159

A Review on Different Face Recognition Techniques


Aishwarya Admane1, Afrin Sheikh1, Sneha Paunikar1, Shruti Jawade1, Shubhangi Wadbude1,
Prof. M. J. Sawarkar2
1 BE Students, Department of Computer Science & Engineering, Priyadarshini J. L. College of Engineering,
Nagpur, Maharashtra, India
2Assistant Professor, Department of Computer Science & Engineering, Priyadarshini J. L. College of
Engineering, Nagpur, Maharashtra, India

ABSTRACT

Face recognition has been a quickly developing, testing and intriguing region progressively applications. An
extensive number of face recognition calculations have been produced in a decade ago. In this paper an
endeavor is made to survey an extensive variety of techniques utilized for face recognition exhaustively. This
incorporate PCA, LDA, ICA, SVM, Gabor wavelet delicate registering instrument like ANN for recognition and
different cross breed blend of this systems. This audit examines every one of these techniques with parameters
that difficulties face recognition like illumination, pose variation, facial expressions.
Keywords: Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), Face Recognition,
Independent, Component Analysis (ICA), Artificial Neural Networks (ANN)

I. INTRODUCTION the 1970s [4] and after the fundamental work of


Kanade [5].
Face recognition is an essential piece of the ability of
human discernment framework and is a standard In 1995, an audit paper [6] gave an exhaustive review
errand for people, while building a comparable of face recognition innovation around then [7].
computational model of face recognition. The Around then, video-based face recognition was still in
computational model add to hypothetical bits of a beginning stage. Amid the previous decades, face
knowledge as well as to numerous down to earth recognition has gotten expanded consideration and
applications like mechanized group observation, get to has progressed actually. Numerous business
control, outline of human PC interface (HCI), content frameworks for still face recognition are presently
based picture database administration, criminal accessible. As of late, noteworthy research endeavors
distinguishing proof et cetera. The soonest chip away have been centered around video-based face
at face recognition can be followed back in any event displaying/following, recognition and framework
to the 1950s in brain science [1] and to the 1960s in incorporation. New databases have been made and
the designing writing [2]. A portion of the soonest assessments of recognition strategies utilizing these
examines incorporate work on facial demeanor databases have been done. Presently, the face
feelings by Darwin [3]. In any case, inquire about on recognition has turned out to be a standout amongst
programmed machine recognition of faces began in the most dynamic uses of example recognition, picture
examination and comprehension.

CSEIT195159 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 207-213 ]
207
Aishwarya Admane et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 207-213

II. FACE RECOGNITION ALGORITHMS LDA is more touchy than PCA and ICA on fractional
impediments, however PCA is less delicate to halfway
A. Principal Component Analysis (PCA) impediments contrasted with LDA and ICA. PCA is
PCA otherwise called Karhunen-Loeve technique is utilized as a measurement lessening method in [12]
one of the famous strategies for include determination and for demonstrating articulation disfigurements in
and measurement lessening. Recognition of human [13].
faces utilizing PCA was first done by Turk and
Pentland [8] and recreation of human faces was A recursive calculation for figuring the discriminant
finished by Kirby and Sirovich [9]. The recognition highlights of PCA-LDA method is presented in [14].
technique, known as eigenface strategy characterizes This technique focuses on testing issue of figuring
an element space which lessens the dimensionality of separating vectors from an incrementally arriving
the first information space. This diminished high dimensional information stream without
information space is utilized for recognition. In any processing the comparing covariance grid and without
case, poor separating power inside the class and huge knowing the information ahead of time. The proposed
calculation are the outstanding basic issues in PCA incremental PCA-LDA calculation is exceptionally
technique. This restriction is overwhelmed by Linear proficient in memory utilization and it is extremely
Discriminant Analysis (LDA). LDA is the most productive in the estimation of first premise vectors.
predominant calculations for highlight choice in This calculation gives an adequate face recognition
appearance based techniques [9]. achievement rate in correlation with extremely
acclaimed face recognition calculations, for example,
Be that as it may, numerous LDA based face PCA and LDA. Two appearance– based systems, for
recognition framework initially utilized PCA to lessen example, Modified PCA (MPCA) and Locality
measurements and after that LDA is utilized to boost Preserving Projections (LPP) are consolidated in [15]
the separating energy of highlight determination. The to give a high face recognition rate. PCA is utilized as
reason is that LDA has the little specimen measure a component extraction strategy in [16]. These
issue in which dataset chose ought to have bigger component vectors are thought about utilizing
examples per class for good segregating highlights Mahalanobis separations for basic leadership. Tensor
extraction. Accordingly executing LDA based Multilinear PCA approach is proposed in [17]
straightforwardly brought about poor extraction of which removes highlight specifically from the tensor
separating highlights. In the proposed strategy [10] portrayal as opposed to the vector portrayal. This
Gabor channel is utilized to channel frontal face technique demonstrates a superior execution in
pictures and PCA is utilized to decrease the correlation with the outstanding strategies in separate
measurement of separated component vectors and changing situations.
after that LDA is utilized for highlight extraction. The
exhibitions of appearance based factual techniques, for PCA can outflank over numerous different systems
example, PCA, LDA and ICA are tried and looked at when the span of database is little. In proposed
for the recognition of hued faces pictures in [11]. PCA calculation [18] the database was sub grouped utilizing
is superior to LDA and ICA under various a few highlights of enthusiasm for faces. Just a single
illumination variations however LDA is superior to of the acquired subgroups was given by PCA to
ICA. recognition. Notwithstanding the great aftereffects of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


208
Aishwarya Admane et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 207-213

PCA, this system has the inconvenience of being Another face recognition technique in light of PCA,
computationally costly and complex with the LDA and neural system were proposed in [21]. This
expansion in database estimate, since every one of the technique comprises of four stages:
pixels in the picture are important to get the portrayal
used to coordinate the information picture with all i. Preprocessing
others in the database. ii. Dimension lessening utilizing PCA
iii. include extraction utilizing LDA
Diverse dimensionality diminishment methods, for iv. grouping utilizing neural system.
example, PCA, Kernel PCA, LDA, Locality
safeguarding Projections and Neighborhood Blend of PCA and LDA were utilized for enhancing
Preserving inserting were chosen and connected so as the capacity of LDA when a couple of tests of pictures
to decrease the loss of grouping execution due to were accessible and neural classifier was utilized to
changes in facial appearance. The execution of diminish number misclassification caused by not-
recognition while utilizing PCA and in addition LDA directly distinguishable classes. The proposed strategy
for dimensionality diminishment is by all accounts was tried on Yale face database. Exploratory outcomes
parallel as far as precision. However, it was watched on this database showed the viability of the proposed
that LDA requires long time for preparing more technique for face recognition with less
number of numerous face pictures notwithstanding misclassification in examination with past strategies.
for little databases. In the event of Locality Preserving
Projections (LPP) and NPE techniques, the An alternate approach for face recognition was
recognition rate was less if expanding number of face proposed in [22] which limit calculation time while
pictures were utilized when contrasted with that of accomplishing higher discovery precision. PCA was
PCA and KPCA strategies. The proposed technique utilized to decrease the measurement separating a
[19] gave significant enhancements on account of component vector. GRNN utilized as a capacity
illumination variations, PCA and bit PCA are the best estimate system to identify whether the information
entertainers. picture contains a face or not and if existed then
reports about its introduction. The proposed
Altered PCA calculation for face recognition were framework had demonstrated that GRNN can perform
proposed in [20], this strategy depended on superior to back propagation calculation and give
diminishing the impact of eigenvectors related with some answer for better regularization.
the expansive eigen esteems by normalizing the
component vector component by its comparing B. Support Vector Machine (SVM)
standard deviation. The reenactment comes about Support Vector Machines (SVM) are a standout
demonstrate that the proposed strategy brings about a amongst the most valuable systems in grouping issues.
superior execution than customary PCA and LDA One clear case is face recognition. Be that as it may,
approaches and the computational cost continues as SVM can't be connected when the component vectors
before as that of PCA and significantly less than that characterizing tests have missing sections. A grouping
of LDA. calculation that has effectively been utilized as a part
of this structure is the all-known Support Vector
Machines (SVM) [23], which can be connected to the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


209
Aishwarya Admane et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 207-213

first appearance space or a subspace of it acquired in This approach depended on the exhibitions of a firmly
the wake of applying an element extraction technique segregating optical connection strategy alongside the
[24] [25] [26]. The benefit of SVM classifier over heartiness of the ICA show. Independent component
customary neural system is that SVMs can accomplish analysis (ICA) demonstrate had started enthusiasm for
better speculation execution. looking for a direct change to express an arrangement
of arbitrary factors as straight mixes of factually
C. Independent Component Analysis (ICA) independent source factors [36]. ICA gave a more
Independent component analysis (ICA) is a strategy intense information portrayal than PCA as its
for finding basic elements or components from objective was that of giving an independent instead of
multivariate (multidimensional) measurable uncorrelated picture deterioration and portrayal. A
information. There is have to execute face recognition quick incremental primary non Guassian headings
framework utilizing ICA for facial pictures having analysis calculation called IPCA_ICA was proposed in
face introductions and distinctive illumination [37].This calculation figures the key components of an
conditions, which will give better outcomes as arrangement of picture vectors incrementally without
contrasted and existing frameworks [27] [28] [29] . evaluating the covariance grid and in the meantime
What recognizes ICA from different strategies is that, change these important components to the
it searches for component that is both measurably independent bearings that amplify the non-
independent and non-gaussian [27]. The ICA is like Guassianity of the source. IPCA_ICA is extremely
visually impaired source detachment issue [30] that effective in the estimation of the main premise vectors.
comes down to finding a straight portrayal in which PCA_ICA makes higher normal progress rate than
the components are measurably independent. The Eigenface, the Fisherface and FastICA strategies.
correlation of face recognition utilizing PCA and ICA
on FERET database with various classifiers [31] [32] III. CONCLUSIONS
were talked about and discovered that the ICA would
be advised to recognition rate as contrasted and PCA This paper has endeavored to survey significant
with factually independent premise pictures and number of papers to cover the current advancement
furthermore with measurably independent in the field of face recognition. Introduce think about
coefficients. Face recognition utilizing ICA with uncovers that for upgraded face recognition new
extensive revolution edges with poses and variations calculation needs to advance utilizing half and half
in illumination conditions was proposed in [33]. A strategies for delicate figuring apparatuses, for
novel subspace technique called successive line example, ANN, SVM, SOM may yields better
segment independent component analysis for face execution. The rundown of references to give more
recognition is proposed in [34]. In ICA each face itemized comprehension of the methodologies
picture is changed into a vector before figuring the portrayed is enrolled. We apologize to specialists
independent components. RC_ICA decreases face whose critical commitments may have been neglected.
recognition blunder and dimensionality of recognition
subspace ends up noticeably littler. A novel system for
face recognition consolidated the independent
component analysis (ICA) demonstrate with the
optical connection procedure was proposed in [35].

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


210
Aishwarya Admane et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 207-213

IV. REFERENCES [11]. Önsen TOYGAR Adnan ACAN ,‖Face


recognition using PCA,LDA and ICA approaches
[1]. Bruner, I. S. and Tagiuri, R. The perception of on colored images‖, Journal Of Electrical and
people. In Handbook of Social Psychology, Vol. Electronics Engineering, vol- 13,2003
2, G. Lindzey, Ed., Addison-Wesley, Reading, [12]. Y. Cheng, C.L. Wang, Z.Y. Li, Y.K. Hou and
MA, 634–654.1954 C.X. Zhao,‖ Multiscale principal contour
[2]. Bledsoe, W. W. The model method in facial direction for varying lighting face
recognition. Tech. rep. PRI:15, Panoramic recognition‖,Proceedings of IEEE 2010
research Inc., Palo Alto, CA.1964 [13]. F. Al-Osaimi·M. Bennamoun · A. Mian,‖ An
[3]. Ekman, P. Ed., Charles Darwin’s The Expression Expression Deformation Approach to Non-rigid
of the Emotions in Man and Animals, Third 3D Face Recognition‖, Springer
Edition, with Introduction, Afterwords and Science+Business Media, LLC 2008
Commentaries by Paul Ekman. Harper- [14]. Issam Dagher,‖Incremental PCA-LDA
Collins/Oxford University Press, New York, algorithm‖, International Journal of Biometrics
NY/London, U.K.1998 and Bioinformatics (IJBB), Volume (4): Issue (2)
[4]. Kelly, M. D. Visual identification of people by [15]. J. Shermina,V. Vasudevan,‖An Efficient Face
computer. Tech. rep. AI-130, Stanford AI recognition System Based on Fusion of MPCA
Project, Stanford, CA. 1970 and LPP‖, American Journal of Scientific
[5]. Kanade, T. Computer recognition of human Research ISSN 1450-223X Issue 11(2010), pp.6-
faces. Birkhauser, Basel, Switzerland, and 19
Stuttgart, Germany 1973 [16]. Ishwar S. Jadhav, V. T. Gaikwad, Gajanan U.
[6]. Chellapa, R., Wilson, C. L., and Sirohey, S. Patil,‖Human Identification Using Face and
Human and machine recognition of faces: A Voice Recognition‖, International Journal of
survey. Proc. IEEE, 83, 705–740.1995 Computer Science and Information
[7]. Samal, A. and Iyengar, P. Automatic recognition Technologies, Vol. 2 (3), 2011
and analysis of human faces and facial [17]. Yun-Hee Han,Keun-Chang Kwak,‖ Face
expressions: A survey. Patt. Recog. 25, 65– Recognition and Representation by Tensor-
77.1992 based MPCA Approach‖, 2010 The 3rd
[8]. M. Turk and A. Pentland, "Eigenfaces for International Conference on Machine Vision
recognition," J. Cognitive Neuroscience,vol. 3, (ICMV 2010)
71-86., 1991. [18]. Neerja,Ekta Walia,‖Face Recognition Using
[9]. D. L. Swets and J. J. Weng, "Using discriminant Improved Fast PCA Algorithm‖,Proceedings of
eigenfeatures for image retrieval", IEEE Trans. IEEE 2008
PAMI., vol. 18, No. 8, 831-836, 1996. [19]. S.Sakthivel,Dr.R.Lakshmipathi,‖Enhancing Face
[10]. C.Magesh Kumar, R.Thiyagarajan , Recognition using Improved Dimensionality
S.P.Natarajan, S.Arulselvi,G.Sainarayanan,‖ Reduction and feature extraction
Gabor features and LDA based Face Recognition Algorithms_An Evaluation with ORL database‖,
with ANN classifier‖,Procedings Of ICETECT International Journal of Engineering Science
2011 and Technology Vol. 2(6), 2010

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


211
Aishwarya Admane et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 207-213

[20]. Lin Luo, M.N.S. Swamy, Eugene I. Plotkin, ―A Applications‖ Neural Networks Research Centre
Modified PCA algorithm for face recognition‖, Helsinki University of Technology P.O. Box
Proceedings of IEEE 2003 5400, FIN-02015 HUT, Finland, Neural
[21]. A. Hossein Sahoolizadeh, B. Zargham Heidari, Networks, 13(4-5):411-430, 2000
and C. Hamid Dehghani,‖ A New Face [31]. Bruce A. Draper, Kyungim Baek, Marian
Recognition Method using PCA, LDA and Stewart Bartlett, ―Recognizing faces with PCA
Neural Network‖, International Journal of and ICA‖, Computer Vision and Image
Electrical and Electronics Engineering 2:8 2008 Understanding 91 (2003) 115-137.
[22]. Feroz Ahmed Siddiky, Mohammed Shamsul [32]. Jian Yang, David Zhang, Jing-yu Yang, ―Is ICA
Alam,Tanveer Ahsan and Mohammed Saifur Significantly Better than PCA for Face
Rahim,‖An Efficient Approach to Rotation Recognition?‖ Proceedings of the Tenth IEEE
Invariant Face detection using PCA,Generalized International Conference on Computer Vision
Regression Neural network and Mahalanobis (ICCV’05) 1550- 5499/05.
Distance by reducing Search space‖,Proceedings [33]. Kailash J. Karande, Sanjay N.Talbar,‖ Face
Of IEEE 2007 Recognition under Variation of Pose and
[23]. Vapnik. Statistical Learning Theory. JohnWiley Illumination using Independent Component
and Sons, New York, 1998. Analysis‖, ICGST-GVIP, ISSN 1687-398X,
[24]. E. Osuna, R. Freund, and F. Girosit. Training Volume (8), Issue (IV), December 2008
support vector machines: an application to face [34]. Quanxue Gao , LeiZhang, DavidZhang,‖
detection. Proc. of CVPR, pages 130–136, 1997. Sequential row–column independent
[25]. B. Heisele, T. Serre, and T. Poggio. A component analysis for face recognition‖,
componentbased framework for face detection Elsevier 2008
and identification. IJCV, 74(2):167–181, 2007. [35]. A. Alfalou and C. Brosseau,‖ A New Robust and
[26]. Q. Tao, D. Chu, and J. Wang. Recursive support Discriminating Method for Face Recognition
vector machines for dimensionality reduction. Based on Correlation Technique and
IEEE Trans. NN, 19(1):189–193, 2008. Independent Component Analysis Model‖,
[27]. Marian Stewart Bartlett, Javier R. Movellan, Optics Letters 36 (2011) 645-647
Terrence J. Sejonowski, ―Face Recognition by [36]. P. Comon, ―Independent component analysis:
Independent Component Analysis‖, IEEE a new concept?‖, Signal Process. 3, 287-314
Transactions on Neural Networks, vol-13, No- (1994).
6,November 2002, PP 1450- 1464. [37]. Issam Dagher and Rabih Nachar,‖Face
[28]. Pong C.Yuen, J.H.Lai, ―Face representation Recognition Using IPCA-ICA algorithm‖, IEEE
using independent component analysis‖, Pattern Transactions On Pattern Analysis and Machine
Recognition 35 (2002) 1247-1257. Intelligence, VOL. 28, NO. 6, JUNE 2006
[29]. Tae-Kyun Kim, Hyunwoo Kim, Wonjum [38]. Arindam Kar, Debotosh Bhattacharjee, Dipak
Hwang, Josef Kittler, ―Independent component Kumar Basu, Mita Nasipuri, Mahantapas
analysis in a local facial residue space for face Kundu,‖ High Performance Human Face
recognition‖, Pattern Recognition 37 (2004) Recognition using Independent High Intensity
[30]. Aapo Hyvärinen and Erkki Oja ―Independent Gabor Wavelet Responses: A Statistical
Component Analysis: Algorithms and Approach‖, International Journal of Computer

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


212
Aishwarya Admane et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 207-213

Science & Emerging Technologies 178 Volume Cite this article as :


2, Issue 1, February 2011
[39]. Lades M, Vorbruggen J, Buhmann J, Lange J, Aishwarya Admane, Afrin Sheikh, Sneha Paunikar,
Cvd Malsburg, Wurtz R, Konen W Distortion Shruti Jawade, Shubhangi Wadbude, Prof. M. J.
invariant object recognition in the dynamic link Sawarkar , "A Review on Different Face Recognition
architecture. IEEE Trans Comput 42(3):300– Techniques", International Journal of Scientific
311,1993 Research in Computer Science, Engineering and
[40]. Wiskott L, Fellous JN, Kruger N, Cvd Malsburg Information Technology (IJSRCSEIT), ISSN : 2456-
Face recogniton by elastic bunch graph 3307, Volume 5 Issue 1, pp. 207-213, January-
matching. IEEE Trans PAMI 197:775–779,1997 February 2019.
Journal URL : http://ijsrcseit.com/CSEIT195159

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


213
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195160

A Survey on Various Bio-Metric Techniques and its Application


for Child Care
Shanon Paul1, Payal Ramteke1, Sugam Sangai1, Minakshee Shende1, Aafreen Qureshi1, Prof. Umesh Samarth2
1 Department of Information Technology, Rashtrasant Tukdoji Maharaj Nagpur University, Nagpur,
Maharashtra, India
2 Professor, Department of Information Technology, JD College of Engineering and Management, Nagpur,
Maharashtra, India

ABSTRACT

With various developing applications requiring biometric acknowledgment of children (e.g., following tyke
immunization plans, recognizing missing children and counteracting infant swaps in emergency clinics),
exploring the worldly solidness of biometric acknowledgment precision for children is critical. The diligence of
acknowledgment precision of three of the most ordinarily utilized biometric qualities (fingerprints, face and iris)
has been examined for grown-ups. Notwithstanding, the industriousness of biometric acknowledgment
precision has not been concentrated methodicallly for children in the age gathering of 0-4 years. Given that
youthful children are regularly uncooperative and don't fathom or adhere to directions, as we would see it,
among all biometric modalities, fingerprints are the most suitable for perceiving children. This is principally in
light of the fact that it is simpler to catch fingerprints of youthful children contrasted with other biometric
attributes, e.g., iris, where a tyke needs to gaze straightforwardly towards the camera to start iris catch. In this
report, we detail our drive to explore the perseverance of unique finger impression acknowledgment for
children.
Keywords : Bio-Metric, Fingerprint, Child Identification

I. INTRODUCTION pay nations commonly don't have any type of ID


reports which can be utilized for this reason, we
One of the significant objectives of most national, address the accompanying inquiry: can fingerprints
worldwide and non-administrative wellbeing be adequately used to perceive children from birth to
associations is to kill the event of immunization- 4 years? We have gathered 1,600 unique mark
preventable youth sicknesses (e.g., polio). Without a pictures (500 ppi) of 20 babies and little children
high inoculation inclusion in a nation or a geological caught over a 30-day time frame in East Lansing,
locale, these lethal illnesses take an overwhelming Michigan and 420 fingerprints of 70 newborn
toll on children. In this manner, it is essential for a children and babies at two diverse wellbeing centers
viable vaccination program to monitor children who in Benin, West Africa. We conceived the
have been inoculated and the individuals who have accompanying systems to enhance the unique finger
gotten the required supporter shots amid the initial 4 impression acknowledgment exactness when looking
years of life to enhance the immunization inclusion. at the gained fingerprints against an all-
Given that children, just as the grown-ups, in low encompassing exhibition database of 32,768 newborn

CSEIT195160 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 214-220 ]
214
Shanon Paul et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 214-220

child fingerprints gathered by VaxTrac in Benin: (I) immunizations obtained, $50 will never go into the
upsample the obtained unique finger impression arm of a tyke in need"3. Accordingly, the kid death
picture to encourage particulars extraction, (ii) rates keep on being high in the low-salary and
coordinate the question print against formats made creating nations.
from every enlistment impression and wire the
match scores, (iii) intertwine the match scores of II. RELATED WORK
the thumb and forefinger, and (iv) refresh the
display with fingerprints procured over various A powerful inoculation program needs to monitor
sessions. which babies and little children have been
vaccinated and how frequently they have gotten
The United Nations Children's Fund (UNICEF's) the required sponsor shots from birth to 4 years old
"2013 Progress Report on Committing to Child (Fig. 1 demonstrates the UNICEF prescribed
Survival: A Promise Renewed" [2] makes reference vaccination plan for newborn children; for CDC
to that while more children currently make due suggested timetable for children up to 6 years see
past their fifth birthday celebration than at any [1]). In creating nations, regularly, there are no
other time, the poorest countries still lose an national distinguishing proof projects which can be
extensive number of children to immunization- utilized to recognize children all through the
preventable illnesses. The 2011 Grand Challenges in vaccination schedule4. This brings up the
Global Health Explorations Round 7 issued by the accompanying issue: can fingerprints, or, so far as
Bill and Melinda Gates Foundation [4] states that that is concerned some other biometric
"every year roughly 25 million babies don't get the methodology, be utilized to distinguish children
vital immunizations1, and in any event, 2.4 million from birth to 4 years old.
children bite the dust from antibody preventable
sicknesses."

With the point of destroying antibody preventable


infections, standard and obligatory immunization
programs are a standard in high salary nations. For
example, as per the Centers for Disease Control and
Prevention (CDC), "in the United States,
inoculation programs have wiped out or
fundamentally decreased numerous antibody
preventable infections" [3]. Thus, the youngster
death rates have diminished extensively in high Fig. 1 : Universally prescribed vaccination plan for
salary nations. Then again, routine inoculation newborn children by UNICEF
programs have not been as powerful in decreasing
the event of immunization-preventable sicknesses A few endeavours have been made to examine the
in low-pay nations. VaxTrac2, a non-legislative suitability of utilizing distinctive biometric qualities
association working in West-African nations, for distinguishing babies and little children. In 1899,
expresses that the "antibody wastage rates are Sir Francis Galton [4] first concentrated the varieties
higher than half is probably the most difficult experienced in the inked unique finger impression
geologies", and "for each $100 in new impressions of a baby caught after some time (from

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


215
Shanon Paul et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 214-220

around 9 days to 4.5 years old). He inferred that it


was not attainable to distinguish newborn children
in the age scope of 0-2.5 years utilizing inked unique
mark impressions. All the more as of late, the Joint
Research Center of the European Commission
distributed a specialized report [5] dedicated to the
(a) (b)
topic of regardless of whether robotized unique mark
Fig. 2 : Use of fingerprints for following the
acknowledgment for children is achievable.
inoculation calendar of newborn children and little
children.
The examination inferred that (I) children can be
recognized utilizing fingerprints when the time
Weingaertner et al. [9] researched the utilization of
contrast between the two caught impressions is
palmprints and impressions for distinguishing infants
under 4.5 years, and (ii) picture quality is a
(0-2 days old). Manual coordinating precision was
conclusive factor in unique mark acknowledgment.
accounted for to be around 83% and in Benin, Africa.
(a) Mothers holding up in a wellbeing center to get
Gottschlich et al. [5] contemplated the impact of
their children inoculated, and (b) a human services
youthful development on the precision of unique
specialist fingerprinting a kid before overseeing
mark coordinating frameworks, and demonstrated
immunization. These pictures were caught by the
that (I) finger impression development can be
creators amid their visit to the immunization bases in
displayed utilizing an isotropic development model,
and on Cotonou, Benin in June 2014. Roughly 67%
and (ii) coordinating exactness of finger impression
for palm prints and impressions of infants, separately.
frameworks can be enhanced by upscaling the
In spite of the fact that various diverse biometric
unique mark pictures utilizing this model while
modalities for distinguishing children have been
coordinating unique finger impression pictures of
investigated, there is no unmistakable agreement on
young people gathered after some time.
(I) regardless of whether it is attainable to perceive
newborn children and little children utilizing
Corby et al. [6] considered the reasonability of
biometrics, and (ii) if biometric acknowledgment is
utilizing business sensors to catch iris pictures of 1.5-
surely plausible, which methodology is most
multi year old children. They announced a high
appropriate for this assignment. In view of various
inability to enlist (FTE) rate of around 57%, in spite
contemplations, for example, simplicity of catch
of the fact that the acknowledgment exactness for
the selected subjectswas extremely high (about 99%). (palmprints are hard to catch since babies and
newborn children keep their clench hands shut),
Tiwari et al. [7] and Bharadwaj et al. [8] caught face
parental concerns (e.g., infrared enlightenment for
pictures of infants (0-3 days old) and inferred that it
iris catch), tirelessness of biometric attribute (facial
was hard to catch great quality face pictures because
qualities change after some time), as we would like to
of (I) net head reflexes, and (ii) posture and
think, fingerprints seem, by all accounts, to be the
demeanor varieties.
most feasible biometric for baby and little child
acknowledgment (see Tab. 1). Undoubtedly, VaxTrac
has built up a versatile immunization library
framework which utilizes fingerprints to recognize
children in Benin1 (Fig. 2(b)). In the VaxTrac

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


216
Shanon Paul et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 214-220

framework, the left and right thumb prints of both The connection-based techniques utilized in spatial
the kid and his mom are gathered. In the event that or in the recurrence area associate two unique mark
the kid's fingerprints can't be coordinated effectively, pictures to process the similitude between them.
mother's fingerprints are utilized for building
up/confirming the youngster's personality. While A tale way to deal with unique mark arrangement
VaxTrac does not report the coordinating exactness and coordinating was proposed by Arun Ross et.al
of children's fingerprints, they notice that they [10]. The list of capabilities utilized in this strategy
constantly end up utilizing the mother's fingerprints was an edge include map. A lot of Gabor channels
for this reason in light of the fact that coordinating pre-tuned to a particular recurrence were utilized to
children's fingerprints flops frequently 5. Proceeded catch the neighborhood edge qualities at different
with endeavors are, thusly, expected to propel the introductions and concentrate the nearby edge
unique finger impression innovation, both detecting attributes. The convolution was performed in the
innovation just as coordinating calculation behind recurrence space. A standard deviation picture that
the versatile immunization library framework. In caught the variety in the edge quality at different
this paper, we present the underlying consequences introductions was developed utilizing the separated
of our progressing study on utilizing fingerprints to pictures. The edge highlight map was acquired
perceive newborn children and little children. utilizing standard deviation picture. A 2D connection
of the edge highlight maps of the question picture
III. LITERATURE SURVEY and format was resolved to produce a coordinating
score.
Countless for programmed unique finger impression
coordinating have been proposed in the writing. Two Karthik Nandakumar and Anil K. Jain [11] proposed a
given fingerprints are thought about by a connection based unique mark matcher that used the
coordinating calculation and return either a level of nearby relationship of districts around the details to
closeness (additionally called coordinating score) or assess the level of comparability between two unique
an acknowledgment/dismissal choice. This finger impression pictures. Particulars focus and the
workgroup the unique finger impression coordinating related edge focuses were separated from the layout
methodologies into connection based, particulars and the inquiry unique finger impression pictures.
based methodologies, and non-details include based, This technique utilized Procrustes examination to get
for example, edge shape, surface data and so forth. a decent gauge of comparing edge bends to adjust the
The motivation behind the study is to break down inquiry to the layout. The two pictures were
those calculations and talk about the execution improved utilizing a bank of Gabor channels of
assessment. various introductions. The standardized cross
correlation was processed to decide the nature of the
The current unique finger impression particulars coordinate. A database comprising of
acknowledgment frameworks utilize methods unique finger impression impressions of 160 clients
dependent on the neighborhood and worldwide were utilized to assess this technique.
element portrayals of the unique mark pictures, for
example, details, edge shape, surface data and so on. A Robust Fingerprint Matching Algorithm for
Verification dependent on connection was
introduced by Abdullah Cavusoglu et. al [12]. They
utilized a fluctuation based division technique to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


217
Shanon Paul et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 214-220

fragment the unique mark picture. Neighborhood the connection between's edges of neighboring
edge introduction was resolved to utilize Sobel particulars that shows the similarity of regions in the
administrators. The strategy used to discover the middle of the two comparing particulars sets were
Reference point depended on the differential entirety processed. The item rule was utilized to join the
of sine estimations of the headings of the pixels quantity of coordinated particulars sets,
situated on a specific sweep. The proposed calculation neighborhood connection score and edge relationship
determined 12 distinctive cross connections score that spoke to an all-out aggregate of
dependent on certain span esteems from the coordinating zones in two fingerprints. The proposed
reference point for both the info and format pictures coordinating calculation recognized real coordinated
and found an all-out total to decide if the pictures details and impostor coordinated particulars in this
relate to a similar unique finger impression. The manner enhancing the coordinating precision.
proposed calculation was assessed on an open space
database. Haiyun Xu et.al [15] exhibited a novel technique for
unearthly details portrayal for unique finger
Koichi Ito et.al [13] proposed a proficient unique impression check. The ghostly particulars spoke to
mark acknowledgment calculation utilizing the stage detail set as a settled length include vector and
parts in 2D Discrete Fourier Transforms of the depended on the move, scale, and revolution
pictures. They utilized Phase-Only Correlation (POC) properties of the two-dimensional (2-D) ceaseless
capacity and Band-Limited Phase-Only Correlation
(BLPOC) capacity to decide the stature and area of Fourier change. The two details portrayal techniques
the crest from stage attributes in Fourier space, which utilized for particulars coordinating were Location-
gave the comparability measure and the translational Based Spectral Minutiae and Orientation-Based
uprooting between the pictures separately. The Spectral Minutiae. To lessen the higher frequencies, a
revolution and the removal between the enlisted Gaussian low-pass channel was connected on the
unique mark and the information finger impression range. Two coordinating calculations were exhibited.
pictures were standardized utilizing BLPOC work. The relationship of two ghostly pictures (direct
The covered locale of the two pictures was removed coordinating) was picked as a comparability score.
lastly the coordinating score is assessed as the The second calculation is the Fourier– Mellin
entirety of the most elevated two pinnacles. The coordinating, in
execution of their calculation was assessed on a
database comprising of 330 unique finger impression which the extent of the Fourier change of the
pictures. particulars range was taken to compute the likeness
score. The proposed calculations were assessed on
Jiang Li et.al [14] depicted a unique mark three unique finger impression databases. A basic
coordinating calculation that consolidates minutia relationship based unique mark confirmation
based coordinating strategy with relationship-based framework was proposed by Asker M. Bazen et.al
coordinating technique. Minutia based coordinating [16]. The proposed framework specifically utilized
calculation was utilized to separate the rundown of the more extravagant dark scale data of the
coordinated minutia sets from the two unique mark fingerprints.
pictures. Relationship of the nearby neighborhood
districts around each coordinating details pair that Suitable trademark layouts were chosen in the
speaks to the neighborhood likeness and furthermore essential unique mark and their relating positions in

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


218
Shanon Paul et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 214-220

the info unique finger impression were resolved. The [2]. Committing to child survival: A promise
format was moved over the whole unique finger renewed.
impression picture and the area where the separation http://www.unicef.org/lac/Committing to Child
is insignificant was picked as the relating position of Survival APR 9 Sept 2013.pdf.
the layout in the info finger impression. The data of [3]. Vaccines and immunizations.
all layout sets was converged to get the ultimate http://www.cdc.gov/vaccines
choice. The proposed framework execution was /vacgen/whatifstop.htm.
assessed on pictures from four diverse unique mark [4]. F. Galton. Finger prints of young children.
databases comprising of 880 fingerprints altogether. British Association for the Advancement of
Science, 1899.
IV. CONCLUSIONS [5]. C. Gottschlich, T. Hotz, R. Lorenz, S.
Bernhardt, M. Hantschel, and A. Munk.
Current biometric information catch and Modeling the growth of fingerprints improves
acknowledgment arrangements cook essentially to matching for adolescents. IEEE Transactions on
grown-ups (more than 16 years old). National ID Information Forensics and Security, 6(3):1165–
programs, for example, India's Aadhaar program, 1169, Sept 2011.
order catching fingerprints and iris pictures of people [6]. P. M. Corby, T. Schleyer, H. Spallek, T. C. Hart,
who are 5 years old or more established. There is R. J. Weyant, A. L. Corby, and W. A. Bretz.
currently a developing requirement for creating Using biometrics for participant identification
capacities to perceive exceptionally youthful children in a research study: a case report. J. Am.
(from infant to 4 years of age) in light of biometrics. Medical Informatics Assoc., 13(2):233–235,
For biometric acknowledgment to be fruitful, 2006.
uniqueness and constancy properties of a biometric [7]. S. Tiwari, A. Singh, and S. K. Singh. Intelligent
characteristic should be fulfilled for the number of method for face recognition of infant.
inhabitants in intrigue. While these two central International Journal of Computer
fundamentals of biometrics have been researched for Applications, 52(4):36–50, 2012.
the three essential biometric characteristics [8]. S. Bharadwaj, H. Bhatt, R. Singh, M. Vatsa, and
(fingerprints, face, and iris) of the grown-up populace, S. Singh. Face recognition for newborns: A
there is no thorough examination led to explore the preliminary study. In 4th IEEE BTAS, pages 1–
perseverance of biometric acknowledgment 6, Sept 2010.
(longitudinal investigation) for children. In this [9]. D.Weingaertner, O. R. P. Bellon, L. Silva, and
examination, we will likely research the tirelessness M. N. Cat. Newborn’s biometric identification:
of unique mark acknowledgment for children in the Can it be done? In VISAPP (1), pages 200– 205,
age gathering of 0-4 years. 2008.
[10]. Arun Ross, James Reisman and Anil Jain,
V. REFERENCES (2002), “Fingerprint Matching Using Feature
Space Correlation”, Biometric Authentication.
[1]. 2014 recommended immunizations for children Springer Berlin Heidelberg, pp. 48-57.
from birth through 6 years old. [11]. Karthik Nandakumar and Anil K. Jain ,(2004),
http://www.cdc.gov/vaccines/ “Local Correlation- based Fingerprint
parents/downloads/parentver- sch-0-6yrs.pdf. Matching”, Conference proceeding of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


219
Shanon Paul et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 214-220

Computer Vision, Graphics and Image Cite this article as :


Processing, Kolkata, India, pp. 503-508.
[12]. Abdullah Cavusoglu and Salih Gorgunoglu, Shanon Paul, Payal Ramteke, Sugam Sangai,
(2007), "A Robust Correlation Based Minakshee Shende, Aafreen Qureshi, Prof. Umesh
Fingerprint Matching Algorithm for Samarth, "A Survey on Various Bio-Metric
Verification", Journal of Applied Sciences 7, pp. Techniques and its Application for Child Care",
3286-3291. International Journal of Scientific Research in
[13]. Koichi Ito, Ayumi Morita, Takafumi Aoki, Computer Science, Engineering and Information
Tatsuo Higuchi, Hiroshi Nakajima, and Koji Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
Kobayashi, (2005), “A Fingerprint Recognition 5 Issue 1, pp. 214-220, January-February 2019.
Algorithm using Phase-Based Image Matching Journal URL : http://ijsrcseit.com/CSEIT195160
for low quality fingerprints”, IEEE
International Conference on Image Processing,
Vol. 2, pp. 33-36.
[14]. Jiang Li, Sergey Tulyakov and Venu
Govindaraju, (2007), “Verifying Fingerprint
Match by Local Correlation Methods”, First
IEEE International Conference on Biometrics:
Theory, Applications, and Systems, pp.1-5.
[15]. Haiyun Xu, Raymond N. J. Veldhuis, Asker M.
Bazen, Tom A. M. Kevenaar, Ton A. H. M.
Akkermans and Berk Gokberk ,(2009),
“Fingerprint Verification Using Spectral
Minutiae Representations”,IEEE Transactions
On Information Forensics And Security, Vol. 4,
No. 3, pp. 397-409.
[16]. Asker M. Bazen, Gerben T.B. Verwaaijen, Sabih
H. Gerez, Leo P.J. Veelenturf and Berend Jan
van der Zwaag, (2000), "A correlation- based
fingerprint verification system", ProRISC 2000
Workshop on Circuits, Systems and Signal
Processing, pp. 205-213

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


220
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT183818
Liver Cancer Detection
L. S. Rohith Anand*, B. Shannmuka, R. Uday Chowdary, K. Satya Sai Krishna
CSE, Vasireddy Venkatadri Institute of Technology, Guntur, Andhra Pradesh, India

ABSTRACT

Machine learning techniques play an important role in building predictive models by learning from Electronic
Health Records (EHR). Predictive models building from Electronic Health Records still remains as a challenge
as the clinical healthcare data is complex in nature and analysing such data is a difficult task. This paper
proposes prediction models built using random forest ensemble by using three different classifiers viz. J48, C4.5
and Naïve Bayes classifiers. The proposed random forest ensemble was used for classifying four stages of liver
cancer. Using a feature selection method the reliable features are identified and this subset serves as input for
the ensemble of classifiers. Further a majority voting mechanism is used to predict the class labels of the liver
cancer data. Experiments were conducted by varying the number of decision trees generated using the J48, C4.5
and Naïve Bayes classifiers and compared with the classification made using decision stump and Adaboost
algorithms.
Keywords : Ensemble, Feature Selection, C4.5, J48 and Random Forest

I. INTRODUCTION format for documenting patient information than


narrative text which is unstructured in nature. EHR
In health care industry, patient’s medical data size helps to streamline the clinical workflow information.
grows day to day. The process of applying computer Ensemble learning is a well-known approach used in
based information system (CBIS), including new machine learning for prediction by combining
techniques, for discovering knowledge from data is various ensemble models [1]. Ensemble of classifiers
called data mining. The process of machine learning is aggregations of multiple classifiers are J48, C4.5 and
is similar to that of data mining. Machine learning Naive Bayes etc. [2]. Ensembles aim for better
algorithms may be distinguished by either supervised performance than any of the base classifiers. The
or unsupervised learning methods. Supervised proposed work aims to improve the accuracy of
learning methods are widely used for predictive healthcare data for prediction and classification, by
modelling. Predictive modelling is a branch of building a hybrid predictive classifier model using
clinical and business intelligence branch which is ensemble of classifiers [3][4].
used for health risk classification and also to predict
the future health status of the individuals. Electronic The remaining part of the paper described in the
health records (EHR) are used to store large scale following section.
information of patient conditions, treatments etc.
The EHR information may be structured or Section 2 describes related works on classification,
unstructured. Using controlled vocabulary, electronic feature selection, subset generation, pre- processing
health records are maintained in structured data and boosting algorithms such as adaptive boosting for

CSEIT183818 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 221-228 ]
221
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

electronic health records. Section 3 explains the constraint projection for mutli-label ensemble. They
overall architecture of proposed system. Section 4 have adopted boosting methods to construct a multi-
reports the experimental results. Section 5 concludes label ensemble to increase the generalization ability
the paper. [6].

II. RELATED WORK Jia Zhua et. al. (2015) employed multiple classifier
systems (MCS) to improve the accuracy of disease
This section discusses the existing methods for pre- detection for Type-2 Diabetes Mellitus. Multi
processing, feature extraction, boosting methods such classifier system performs worse when design is not
as adaptive boosting. Aydin et. al. (2009) investigated proper [7]. They have proposed a dynamic weighted
the various factors involved on ensemble voting scheme for multiple classifier decision
construction using a wide variety of learning combination. Yan Li et. al. (2015) stated data mining
algorithms, data sets and evaluation criteria [5]. They framework for distributed healthcare information
have provided the idea of subset selection to the level based on privacy preserving constraints [8]. Neesha
of discriminating whether the discrimination is Jothi et. al. (2015) surveyed the data mining
applicable or not at the level of classifier.Ping Li et. al. techniques and has classified the articles have
(2013) surveyed about supervisedmulti-label suggested data mining plays important role in
classification and proposed variable pairwise medical diagnosing for predicting diseases [9].

Table 1. Comparative analysis of different ensembles of classifiers

Author Methods Used Data Sets Used Number of Performance Metrics


Iterations

Nikunj C. K nearest neighbor, Learning Datasets from 110 Average, Geometric


Oza vector quantization, Multi-layer UCI repository mean
et.al.method perceptron’s, Radial basis
(2008)[10] functions, Support
vector machines
Yong Seog Naive Bayes, Support vector German credit 120 AUC, Accuracy,
Kim machines data and COII False positive rate,
et.al.method ,Artificial neural networks, 2000 competition Hit rate gain
(2009)[11] Pruned tree classifier data
Hesam Sagha Quadrant discriminant analysis 2 real datasets 45 Entropy, Mutual
et.al.method containing data Information
(2013)[12] from body
mounted inertial
sensors
Ping Li Bagging, Boosting, Random 12 datasets 20 Hamming loss,
et.al.method Forest, Random subspace, including test Ranking loss,
(2013) [6] Rotation forest categorization, One error,
image Coverage, Average

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


222
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

classification and Precision, F1-


bioinformatics metrics, Recall
Ritaban Binary tree, Linear 24 Holstein- 20 AUC, Accuracy
Dutta discriminant analysis classifier, Friesian cows
et.al.method Naïve Bayes Classifier, K- from Tasmanian
(2015) [13] nearest Institute of
neighbor, Adaptive Neuro Agriculture
Fuzzy Inference classifier Dairy Research
Facility
Yang Zhang SVM classifier, BPNN classifier Benchmark 11 Average regression
et.al.method datasets from
(2015)[14] UCI repository
Bing Gong Artificial neural network, Datasets from 30 F measure, G mean
et.al.method Support vector machines, UCI repository
(2016)[15] CART
Yan Li Adaboost algorithm 9948 real world 20 F-measure,
et.al.method EHRs of diabetes Sensitivity,
(2016)[8] patients Precision
Cátia M. Apriori decision, Aposteriori Benchmark 12 AUC,
Salgado decision datasets from Accuracy,
et.al.method UCI collection, Sensitivity,
(2016)[16] MIMIC II Specificity
datasets

Based on the literature survey carried out a proposed system and how limitations in existing
comparative analysis of the ensemble of classification system is resolved.
methods, the data sets used for experiments by
different researchers, the number of iterations for III. PROPOSED WORK
which the experiments were conducted and the
metrics used for measuring the classification accuracy The proposed architecture is shown in Figure 1. The
are tabulated in the table given below. Electronic health records contain features like patient
id, status, age, sex, hepato, ascites, edema, billi,
From the above table the conclusion drawn is an cholestrol, albumin etc. The data considered have to
ensemble of C4.5, J48 and Naïve Bayes classifier with be clinically transformed i.e. to make it suitable for
majority voting scheme was not studied and hence further processing. The clinical transformation step is
this work focusses on building a predictive model also identified as preprocessing step.
based on building a random forest using these three
classifiers. The proposed system has been compared The unprocessed has null values, irrelevant values
with the existing decision stump and Adaboost and noisy values. These data errors would lead to
algorithms. The next section discusses about the misclassification and hence need to be clinically
transformed. The missing data in the considered

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


223
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

dataset is imputed with values computed using mode The proposed system with its role and advantages is
function. discussed. The experimental result analysis of the
proposed work has been discussed in the next section.
After pre-processing of data, for classifying instances
under Random forest, three subsets from the datasets IV. RESULTS AND DISCUSSION
are generated. The subset will be generated
considering three features like platelet count, alkaline Experiment Results
phosphate and cholesterol values.
The proposed system is implemented using Java and
The random forests are built using three classification Weka tool. The liver cancer dataset having 500
algorithms namely C4.5, J48 and Naïve Bayes. There instances and breast cancer dataset are used for the
are many voting mechanisms followed for ensemble experiments. The ensemble of classifiers is used for
of classifiers, here we are using majority vote method classifying these datasets on which voting is
to perform voting with different classifiers. Here the performed.
output will be the final outcome of the majority of
classifiers Pre-processing

Figure 1 shows the architecture of proposed system. In this module, pre-processing of data is done. The
dataset which we have contains null values,
irrelevant values and noisy values. As missing values
in the dataset lead to misprediction of the final result,
the dataset is pre-processed by filling missed values
on basis of mode function. The dataset without pre-
processing which contains irrelevant values is shown
below in figure 2.

Figure 2. Dataset before preprocessing

Figure 1. Architecture of Proposed System The dataset is preprocessed to fill missing and
irrelevant values as shown in figure 3

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


224
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

Figure 5. Results achieved after implementing subset


Figure 3. Dataset after preprocessing
generation

Feature selection
Performance Evaluation

Feature selection method used for model


This section evaluates the performance of J48
construction by choosing a subset of relevant
Random forest classifier, C4.5 Random Forest
predictors. It also called as variable selection or
classifier and Naïve Bayes classifier using Accuracy,
attribute selection [17].
True positive, False positive, Precision, Recall and F-
measure. Figure 6 shows the performance of J48
Random Forest classifier for prediction different
stages of liver cancer. Figure 7 show the performance
of C4.5 Random forest classifier for prediction
different stages of liver cancer and Figure 8 shows the
performance of Naïve Bayes classifier for prediction
different stages of liver cancer.

Figure 4. Dataset after feature selection

Subset Generation

For classifying instances under Random forest, Figure 6. Comparison of J48 Random forest classifier
generate three subsets from the datasets. Subset will at different stages
be generated considering some features as first,
middle and last stages as shown in figure 5.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


225
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

Comparison of Existing and Proposed System

The comparative analysis of classification accuracy


for Liver and Breast cancer dataset shown in figure 9
and also existing and proposed system based on
different threshold values shown in Table 2 and
figure 10.

Figure 7. Comparison of C4.5 Random forest classifier


at different stages

Table 2. Accuracy of Existing and Proposed system


based on different threshold values

Syste Weight Weigh Weight Weight


m ed ted ed ed
Thresh Thresh Thresh Thresh
old old old old
value value value value Figure 9. Comparison of existing and proposed system
100 in 200 in 500 in 1000 in for two different datasets
% % % %
Existi
ng 46 48 46 44
Propo
sed 51 53 50 48

Figure 10. Comparison of existing and proposed


system based on weighted threshold

In this section, we had a brief discussion about


implementation and experimental results of Clinical
feature transformation, Feature selection, Subset
generation, J48 classifier, C4.5 classifier, Naive bayes
classifier and Majority voting.
Figure 8. Comparison of Naïve Bayes classifier at
different stages

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


226
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

V. CONCLUSION construction of classifier and discriminant


ensembles, Information Sciences 179 (2009)
The prediction model is built by series of steps such 1298–1318.
as clinical feature transformation, feature selection, [6] Ping Li , Hong Li , Min Wu , Multi-label
ensemble of classifiers and Majority voting which ensemble based on variable pairwise
aimed to improve rate of correct predictions. The constraint projection, Information Sciences
prediction accuracy is improved by an ensemble of 222 (2013) 269–281.
classifiers and when majority voting mechanism was [7] Jia Zhua,, Qing Xie, Kai Zheng,An improved
applied on them. The proposed system here achieve early detection method of type-2diabetes
an accuracy of 40% for C4.5 Random forest classifier, mellitus using multiple classifier system,
43% for J48 Random forest classifier, 38% for Naïve Information Sciences 292 (2015) 1–14.
bays classifier when tested for Liver cancer dataset [8] Yan Li, Changxin Bai, Chandan K. Reddy, A
which is more than existing system. The accuracy of distributed ensemble approach for mining
45% for C4.5 Random forest classifier, 52% for J48 healthcare data under privacy constraints,
Random forest classifier and 47% for Naïve Bayes Information Sciences (2015).
classifier was achieved when tested for Breast cancer [9] Neesha Jothi, Nur’Aini Abdul Rashid,
dataset which is more than existing system which has Wahidah Husain, Data Mining in
an accuracy of 46%. The number of trees generated Healthcare – A Review, Procedia Computer
was varied and the prediction accuracy of the Science 72 ( 2015 ) 306 – 313.
proposed work was studied. [10] Nikunj C. Oza , Kagan Tumer, Classifier
ensembles: Select real-world applications,
VI. REFERENCES Information Fusion 9 (2008) 4–20.
[11] YongSeog Kim, Boosting and measuring the
[1] Dietterich TG.(2000), Ensemble methods in performance of ensembles for a successful
machine learning. In: Proceedings of database marketing, Expert Systems with
Multiple Classifier System, vol. 1857, Applications 36 (2009) 2161–2176.
Springer (2000), pp. 1–15. [12] Hesam Sagha, Hamidreza Bayati, José del R.
[2] Zhi-Hua Zhou, Ensemble Methods: Millán, Ricardo Chavarriaga, On-line
Foundations and Algorithms, Machine anomaly detection and resilience in classifier
Learning & Pattern Recognition Series, ensembles, Pattern Recognition Letters
2012. (2013).
[3] Yongjun Piao, Minghao Piao, Keun Ho Ryu, [13] Ritaban Duttaa, Daniel Smitha, Richard
Multiclass cancer classification using a Rawnsley, Greg Bishop-Hurley, James Hills,
feature subset- based ensemble from Greg Timms, Dave Henry, Dynamic cattle
microRNA expression profiles, Computers in behavioural classification using supervised
Biology and Medicine 80 (2017) 39–44. ensemble classifiers, Computers and
[4] ManjeevanSeera, Chee Peng Lim, A hybrid Electronics in Agriculture 111 (2015) 18–28.
intelligent system for medical data [14] Yang Zhang, Li Zhang, M.A. Hossain,
classification, Expert Systems with Adaptive 3D facial action intensity
Applications 41 (2014) 2239– 2249. estimation.
[5] Aydın Ulas, Murat Semerci, Olcay Taner [15] Bing Gong, Joaquín Ordieres-Mere,
Yıldız, Ethem Alpaydın, Incremental Prediction of daily maximum ozone

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


227
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

threshold exceedances by preprocessing and


ensemble artificial intelligence techniques:
Case study of Hong Kong, Environmental
Modelling & Software 84 (2016) 290-303.
[16] Cátia M. Salgado, Susana M. Vieira, Luís F.
Mendonça, Stan Finkelstein, João M.C.
Sousa, Ensemble fuzzy models in
personalized medicine: Application to
vasopressors administration, Engineering
Applications of Artificial Intelligence
(2015).
[17] B. Seijo-Pardo, I. Porto-D’iaz, V. Bol’on-
Canedo, A. Alonso-Betanzos, Ensemble
Feature Selection: Homogeneous and
Heterogeneous Approaches, Knowledge-
Based Systems (2016).
[18] J.Prathyusha, G.Sandhya, V.Krishna Reddy,”
An Improvised Partition-Based Workflow
Scheduling Algorithm”, International
Innovative Research Journal of Engineering
and Technology, vol 02, no 04,pp.120-
123,2017.

Cite this article as :

L. S. Rohith Anand, B. Shannmuka, R. Uday


Chowdary, K. Satya Sai Krishna, "Liver Cancer
Detection", International Journal of Scientific
Research in Computer Science, Engineering and
Information Technology (IJSRCSEIT), ISSN : 2456-
3307, Volume 5 Issue 1, pp. , January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT183818
Journal URL : http://ijsrcseit.com/CSEIT183818

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


228
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195150

Emotion Detection using Facial expressions with Transfer


Learning
Sairama Geetanand, Shaik Nazeer Basha, Sajja Nageswara Rao, Rayudu Srinivas
CSE, Vasireddy Venkatadri Institute of Technology, Guntur, Andhra Pradesh, India

ABSTRACT

The use of machines to perform different tasks is constantly increasing in society. Providing machines with
perception can lead them to perform a great variety of tasks. Machine perception requires that machines understand
about their environment. Recognizing facial emotions will thus help in this regard. We use the TensorFlow library
and the Inception model and apply transfer-learning for our dataset to retrain the model. We then identify the facial
emotions: happiness, sadness, anger, surprise, disgust and neutral.
Keywords: Deep learning, TensorFlow, Inception, Transfer Learning, Convolutional neural networks.

I. INTRODUCTION recognition is particularly difficult for two reasons: (1)


There does not exist a large database of training images
Facial expressions play a very important role in human and (2) classifying emotions can be difficult depending
communication. The human face is the richest source of on whether the input image is static or in a transition
emotions. As society continues to make more use of frame into a facial expression. The latter issue is
human-machine interactions, it is important for particularly difficult for real-time detection where facial
machines to be able to interpret facial expressions in expressions vary dynamically.
order to improve their authenticity or to make them less
machine and more human. Our brains make vision In the last few years the field of machine learning has
seem easy. It doesn't take any effort for humans to tell made tremendous progress on addressing these difficult
apart a lion and a tiger, read an article, or recognizing a problems. In particular, the discovery of the model
human's face. But these are actually hard problems to called a deep convolutional neural network can achieve
solve with a computer: they only seem easy because our reasonable performance on hard visual recognition
brains are incredibly good at understanding images. tasks matching or exceeding human performance in
Emotion can be recognized through a variety of means some domains. It is now being used by popular tech-
such as voice intonation, body language, and more giants, Facebook being one of the leading users.
complex methods such electroencephalography (EEG). Researchers have demonstrated steady progress in
However, the easier, more practical method is to computer vision by validating their work against
examine facial expressions. There are seven types of ImageNet an academic benchmark for computer vision.
human emotions shown to be universally recognizable
across different cultures: anger, disgust, fear, happiness, II. METHODS AND MATERIAL
sadness, surprise, contempt. Interestingly, even for
complex expressions where a mixture of emotions A. Dataset
could be used as descriptors, cross-cultural agreement is
still observed. Therefore, a utility that detects emotion We scraped image urls from google images, using java
from facial expressions would be widely applicable. script, downloaded the images from the generated urls
Such advancement could bring applications in medicine, using python. The images are then filtered manually
marketing and entertainment. The task of emotion

CSEIT195150 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 229-234 ]
229
Sairama Geetanand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 229-234

and the ones that don‟t suit the emotion are removed. part of ImageNet Large Sale Visual Recognition
The current dataset of images we used contain 2192 challenge 2014. It can classify almost every day-to-day
images for the six emotions : Happy, Sad, Angry, objects. The architecture of the Inception is briefly
Surprise, Disgust, Neutral. described in . Inception consists of 22 layers. The
penultimate layer is called as “Bottlenecks”. The final
B. TensorFlow layer is called as softmax layer. This is the layer that
can be retrained to classify the required image group.
Tensor Flow (TF) [2] is an open source software library We retrained it to classify 6 emotions: neutral, happy,
for machine learning written in Python and C++. The sad, surprise, disgust and anger. There are 3 metrics.
main reason behind it is that TF was developed by Training accuracy describes the accuracy of the
Google Brain Team. Google has already been using TF inception model in classifying the images on which it
to improve some tasks on several products. These tasks was trained. Validation accuracy describes the accuracy
include speech recognition in Google Now, search of the model in classifying the images that are not part
features in Google Photos, and the smart reply feature of the training dataset. Cross entropy is the loss
in Inbox by Gmail. Some design decision in TF have function. It describes the error rate.
lead to this framework to be early adopted by a big
community. One of them is the ease of going from
prototype to production. There is no need to compile or
to modify the code to use it on a product. Then, the Fig.1: The 3 metrics: Train accuracy, Class entropy,
framework is not only thought as a research tool, but as Validation accuracy
a production one. Another main design aspect is that
there is no need to use different API when working on D. Implementation
CPU or GPU. Moreover, the computations can be
deployed over desktops, servers and mobile devices. A Once we are sure that inception is correctly installed
key component of the library is the data flow graph. and is working correctly we retrain the model for our
The sense of expressing mathematical computations dataset like so. Modern object recognition models have
with nodes and edges is a TF trademark. Nodes are millions of parameters and can take weeks to fully train.
usually the mathematical operations, while edges define Transfer learning is a technique that shortcuts a lot of
the input / output association between nodes. The this work by taking a fully-trained model for a set of
information travels around the graph as a tensor, a categories like ImageNet, and retrains from the existing
multidimensional array. Finally, the nodes are allocated weights for new classes. We retrained the final layer
on devices where they are executed asynchronously or from scratch, while leaving all the others untouched.
in parallel when all the resources are ready. How is it done?

C. Inception We evaluate whether features extracted from the


activation of a deep convolutional network trained in a
Inception is a pre-trained deep neural network, fully supervised fashion on a large, fixed set of object
developed to classify images. It took around 2 weeks to recognition tasks can be re-purposed to novel generic
train the inception model using 8 Tesla K40 GPUs and tasks. Our generic tasks may differ significantly from
it costed around $30,000. It considers around 25 million the originally trained tasks and there may be
parameters and does 5 billion add-multiply operations insufficient labelled or unlabeled data to conventionally
to classify a single image. Given today‟s computation train or adapt a deep architecture to the new tasks. We
power, it can be done in fraction of a second. Inception investigate and visualize the semantic clustering of
takes images as its input. It can process only JPEG deep convolutional features with respect to a variety of
format images. The recommended resolution is such tasks, including scene recognition, domain
299*299. If the image is of higher resolution, it will be adaptation, and fine-grained recognition challenges.
compressed automatically. It produces a private class of Before we start any training, we need a set of images to
array of the image as its output. It was developed as a teach the network about the new classes you want to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


231
Sairama Geetanand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 229-234

recognize. Here we use a dataset of emotions gathered


from a variety of sources. Once you have the images,
from the root of your TensorFlow source directory you
can begin the training process.

We load the pre-trained Inception v3 model, remove the


old top layer, and train a new one on the emotion
Fig.3: Bottleneck values generated for an image
photos we've downloaded. None of the emotions were
in the original ImageNet classes the full network was
Once the bottlenecks are complete, the actual training
trained on. The magic of transfer learning is that lower
of the top layer of the network begins. We'll see a series
layers that have been trained to distinguish between
of step outputs, each one showing training accuracy,
some objects can be reused for many recognition tasks
validation accuracy, and the cross entropy. The training
without any alteration. The script can take thirty
accuracy shows what percent of the images used in the
minutes or more to complete, depending on the speed of
current training batch were labelled with the correct
the machine. The first phase analyzes all the images on
class. The validation accuracy is the precision on a
disk and calculates the bottleneck values for each of
randomly-selected group of images from a different set.
them. 'Bottleneck' is an informal term we often use for
The key difference is that the training accuracy is based
the layer just before the final output layer that actually
on images that the network has been able to learn from
does the classification. This penultimate layer has been
so the network can overfit to the noise in the training
trained to output a set of values that's good enough for
data. A true measure of the performance of the network
the classifier to use to distinguish between all the
is to measure its performance on a data set not
classes it's been asked to recognize. That means it has
contained in the training data -- this is measured by the
to be a meaningful and compact summary of the images,
validation accuracy. If the train accuracy is high but the
since it has to contain enough information for the
validation accuracy remains low, that means the
classifier to make a good choice in a very small set of
network is overfitting and memorizing particular
values. The reason our final layer retraining can work
features in the training images that aren't helpful more
on new classes is that it turns out the kind of
generally. Cross entropy is a loss function which gives
information needed to distinguish between all the 1,000
a glimpse into how well the learning process is
classes in ImageNet is often also useful to distinguish
progressing. The training's objective is to make the loss
between new kinds of objects. Because every image is
as small as possible, so you can tell if the learning is
reused multiple times during training and calculating
working by keeping an eye on whether the loss keeps
each bottleneck takes a significant amount of time, it
trending downwards, ignoring the short-term noise.
speeds things up to cache these Bottleneck values on
disk so they don't have to be repeatedly recalculated
By default, it will run 4,000 training steps. Each step
and if you rerun the script they'll be reused so you don't
chooses ten images at random from the training set,
have to wait for this part again.
finds their bottlenecks from the cache, and feeds them
into the final layer to get predictions. Those predictions
are then compared against the actual labels to update
the final layer's weights through the back-propagation
process. As the process continues you should see the
reported accuracy improve, and after all the steps are
done, a final test accuracy evaluation is run on a set of
images kept separate from the training and validation
pictures. This test evaluation is the best estimate of how
the trained model will perform on the classification task.
Fig.2 : TensorFlow creating bottleneck values for each You should see an accuracy value of between 90% and
image 95%, though the exact value will vary from run to run

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


232
Sairama Geetanand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 229-234

since there is randomness in the training process. This


number is based on the percent of the images in the test
set that are given the correct label after the model is
fully trained.

Tensor Board summaries make it easier to understand,


debug, and optimize the retraining. We can visualize
the graph and statistics, such as how the weights or
accuracy varied during training. The Inception v3
network with a final layer retrained to our categories to
get the output graph and a text file containing the labels
–output labels.

III. RESULTS AND DISCUSSION


Fig.5: A line graph depicting the average percentage
values, obtained by testing various images of the six
After the training process, we provide the retrained
emotions.
model with the image we wish to classify. Note that
the retrained model can classify images only with the
The graph values are taken by testing the output label
classes we provided initially in the retraining
with various images, some taken from the dataset and
inception part, so if the image is for example, an
some outside of it. Each emotion is tested using 15
image of a tall building the result would be the best
images and the average of all these values is then
possible match among the classes. The system can
converted to percentage, which are given as inputs to
identify only the images it is trained for just like
the graph.
humans, seeing something we have never seen before
we shall not be able to identify it. It will classify the
IV.CONCLUSION
images in the weights of percentage, for example if an
image is happy it will identify say 0.9025 to say it
In this work, we detect and recognize emotions in
static human images using TensorFlow and Inception
model. The future enhancement can be an action that
Fig.4: Test image classification values. is done upon recognition of the emotions. If we get a
sad emotion, we can have the system play a song or
The above image shows the classification value of an tell a joke or send his/her best friend a message. This
image, which depicts the emotion „sad‟, and its value can be the next step of AI where the system can
generated is 0.931057 which approximates to 93%. understand, comprehend the user’s feelings and
emotions and react accordingly. This bridges the gap
between machines and humans.

We can also have an interactive keyboard where the


users can just use the app and the app will then
identify the emotion and convert that emotion to the
emoji of choice.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


233
Sairama Geetanand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 229-234

V. REFERENCES

[1]. Raghav Puri, Archit Gupta, Manas Sikri (2018).


Emotion Detection using Image Processing in
Python. 5th International Conference on
“Computing for Sustainable Global
Development (IndiaCom)” IEEE conference ID
: 42835.
[2]. https://codelabs.developers.google.com/codelab
s/tensorflow-for-poets
[3]. Hong-Wei Ng, Viet Dung Nguyen, Vassilios
Vonikakis, Stefan Winkler(2015). Deep
Learning for Emotion Recognition on Small
Datasets Using Transfer Learning.
[4]. Alexandr Rassadin, Alexey Gruzdev, Andrey
Savchenko(2017). Group-level Emotion
Recognition Using Transfer Learning from Face
Identification.
[5]. https://www.tensorflow.org/hub/tutorials/imag
e_retraining
[6]. https://www.tensorflow.org/install/source
[7]. https://docs.docker.com/toolbox

Cite this article as :

Sairama Geetanand, Shaik Nazeer Basha, Sajja


Nageswara Rao, Rayudu Srinivas , "Emotion
Detection using Facial expressions with Transfer
Learning ", International Journal of Scientific
Research in Computer Science, Engineering and
Information Technology (IJSRCSEIT), ISSN : 2456-
3307, Volume 5 Issue 1, pp. 229-234, January-
February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195150
Journal URL : http://ijsrcseit.com/CSEIT195150

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


234
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195167

A Review on Various Algorithms for Student Performance


Prediction
Mayuri S. Dongre1, Neha S.Vidya1, Vanita D. Telrandhe1, Shweta R. Chaudhari1, Anup A. Umrikar1,
Prof. Prachiti V. Adghulkar2
1 BE Students, Department of computer Science & Engineering, Suryodaya collage of Engineering &
Technology, Nagpur, Maharashtra, India
2Assistant Professor, Department of computer Science & Engineering, Suryodaya collage of Engineering &
Technology, Nagpur, Maharashtra, India
ABSTRACT

Data in educational institutions are growing progressively along these lines there is a need of progress this
tremendous data into helpful data and information utilizing data mining. Educational data mining is the zone of
science where diverse techniques are being produced for looking and investigating data and this will be
valuable for better comprehension of understudies and the settings they learned. Classification of data objects in
view of a predefined learning of the articles is a data mining and information administration procedure utilized
as a part of collection comparable data questions together. Decision Tree is a valuable and well known
classification method that inductively takes in a model from a given arrangement of data. One explanation
behind its prominence comes from the accessibility of existing calculations that can be utilized to assemble
decision trees. In this paper we will survey the different ordinarily utilized decision tree calculations which are
utilized for classification. We will likewise contemplating how these decision tree calculations are appropriate
and valuable for educational data mining and which one is ideal.
Keywords : Educational data mining (EDM), Classification, Dropout Prediction, Selection Failure

I. INTRODUCTION The data in any given educational association is


growing rapidly. There is a need to change this data
As of late the scientist concentrates on the new into helpful data and learning; henceforth we make
region of research, which is EDM (Educational Data utilization of data mining.
mining). EDM is the information disclosure in
database strategy or the data mining in instruction. Distinguish and find valuable data covered up in
The specialist concentrates on the improvement of substantial databases is a troublesome errand. An
technique to better comprehend understudies and the exceptionally making an assurance to answer for
settings in which they learn. There are great cases of accomplish this goal is the utilization of learning
how to apply EDM systems to make models that revelation in databases techniques or data mining in
foresee dropping out and understudy disappointment. training, called Educational Data Mining, EDM [15].
These works have indicated promising outcomes Educational data mining techniques drawn from an
regarding those sociological, monetary, or assortment of writings, including data mining and
educational qualities that might be more pertinent in machine learning, psychometrics and different ranges
the forecast of low scholarly execution.

CSEIT195167 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 235-239 ]
235
Mayuri S. Dongre et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 235-239

of measurements, data perception, and computational amid this stage the tree is recursively parcelled till
displaying. every one of the data things have a place with a
similar class name. Tree pruning is done is a base up.
As we examined over the new territory of center for It is utilized to enhance the forecast and classification
the specialists is EDM. In introduce the vast majority exactness of the calculation by limiting over-fitting.
of the procedures was produced for the educational Over-fitting in decision tree calculation brings about
data mining, which foresee the disappointment and misclassification mistake. Tree pruning is less
dropout understudies. Be that as it may, a large entrusting contrasted with the tree development
portion of the examination on the use of EDM to stage as the preparation data set is filtered just once.
determine the issues of understudy disappointment In this examination we will audit Decision tree
and drop-outs has been connected principally to the calculations executed, recognize the ordinarily
particular instance of advanced education and all the utilized calculations.
more particularly to on the web or separation
instruction. Less almost no data about particular II. RELATED WORK
research on basic and optional instruction has been Decision tree is a vital strategy for both acceptance
found, and what has been discovered uses just research and data mining, which is basically utilized
measurable strategies, not DM methods. Along these for display classification and expectation. ID3
lines, there is a need to proposed Educational data calculation is the most broadly utilized calculation in
mining system that is attainable for rudimentary and the decision tree up until this point. Through
optional instruction. outlining on the fundamental thoughts of decision
tree in data mining. ID3 calculation is a forerunner of
Classification can be depicted as a regulated learning C4.5 calculation and it was produced by a data
calculation in the machine learning process. It mining software engineering scientist Ross Quinlan
relegates class marks to data objects in light of earlier in 1983.It is utilized to build a decision tree by testing
information of class which the data records have a every hub's trait of tree in top-down way. ID
place. In classification a given arrangement of data calculation performs property choice component
records is partitioned into preparing and tests data utilizing Entropy and Information Gain idea. A
sets. The preparation data set is utilized as a part of decision tree is an essential method for data mining
building the classification display, while the test data and inductive realizing, which is generally used to
record is utilized as a part of approving the model. frame classifiers and expectation models [2].
The model is then used to order and foresee new
arrangement of data records that is not the same as C4.5 is a standout amongst the most exemplary
both the preparation and test data sets. Decision tree classification calculations on data mining. Decision
calculation is a data mining enlistment methods that Trees a Decision Tree is a helpful and prevalent
recursively segments a data set of records utilizing classification procedure that inductively takes in a
profundity first voracious approach or expansiveness model from a given arrangement of data. One
initially approach until the point when every one of explanation behind its fame originates from the
the data things have a place with a specific class. The accessibility of existing calculations that can be
tree structure is utilized as a part of arranging obscure utilized to manufacture decision trees [3].
data records. Decision tree classification system is
performed in two stages: tree building and tree CART is Classification and relapse trees. It was
pruning. Tree building is done in top-down way. It is presented by Breiman in 1984. It fabricates the two

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


236
Mayuri S. Dongre et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 235-239

classifications and regressions trees. The classification classification and for recognition of connection
tree development via CART. It depends on paired between factors [16].
part of the characteristics. It is additionally in view of
Hunt's model of decision tree development. It The present decision tree calculations likewise
additionally can be executed serially by Breiman et al. contrast in their capacity to deal with various sorts of
It utilizes gini record part measure for choosing the data, and by a wide margin the most exceptional
part quality. Pruning is finished by utilizing a calculation of the gathering is C4.5. C4.5 depends on
segment of the preparation data set in truck. This Quinlan's prior work with ID3 and is equipped for
approach utilizes both numeric and clear cut taking care of ceaseless data, data with missing
properties for building the decision tree and has in- esteems, and even has worked in ventures for
constructed highlights that arrangement with missing rearranging the resultant decision tree. Another
qualities [11]. purpose for the ubiquity of decision trees is that they
are regularly interpretable by human analyzers. The
The fundamental distinction between CART, ID3 and structure of a decision tree gives thinking to each
C4.5 is the manner by which the parceling of data is statement of class esteem, and it is trusted that this
performed. Truck utilizes the Gini list to choose the thinking is anything but difficult to understand.
part quality, though ID3 and C4.5 utilize an
estimation of data pick up and the data pick up The classifier is tried first to group inconspicuous
proportion. data and for this reason coming about decision tree is
utilized. C4.5 calculation takes after the standards of
SLIQ is the Supervised Learning in Ques approach. It ID3 calculation. Also C5.0 calculation takes after the
was presented by Mehta et al in 1996. It is a quick, tenets of calculation that is the vast decision tree can
adaptable decision tree calculation that can be see as an arrangement of guidelines which is
executed in serial and parallel example. It does not straightforward. C5.0 calculation gives the recognize
depend on Hunt's calculation for decision tree on clamor and missing data. Issue of over fitting and
classification. It segments a preparation data set mistake pruning is tackled by the C5 calculation. In
recursively. It utilizes expansiveness first ravenous classification procedure the C5 classifier can envision
procedure that is coordinated with pre-sorting which properties are important and which are not
method amid the tree building stage [17]. applicable in classification [5]

Dash is Scalable Parallelizable Induction of decision


Tree calculation. It was presented by Shafer et al. It is
a quick and adaptable decision tree classifier. Like
SLIQ it utilizes one time kind of the data things. It
has no limitation on the info data estimate [18].

CHAID is a kind of decision tree method. It depends


on balanced hugeness testing. This procedure was
produced in South Africa. It is a comparable form to
relapse examination. This adaptation of CHAID being
initially known as XAID. It is valuable for

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


237
Mayuri S. Dongre et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 235-239

[2]. IU Qin, “Data Mining Method Based on


Computer Forensics-based ID3 Algorithm”,
978-1-4244-5265-1/10/$26.00©2010IEEE
[3]. Rong Cao, Lizhen Xu, “Improved C4.5
Algorithm for the Analysis of Sales”, Sixth Web
Information Systems and Applications
Conference, 978-0-7695-3874-7/09 $25.00 DOI
10.1109/WISA36, © 2009 IEEE
[4]. Gaganjot Kaur , Amit Chhabra , “Improved J48
Classification Algorithm for the Prediction of
Diabetes” , International Journal of Computer
Applications (0975 – 8887) Volume 98 – No.22,
July 2014.
[5]. Tomasz Bujlow, Tahir Riaz, Jens Myrup
Pedersen, “A method for classification of
Table 1. Suervy Table network traffic based on C5.0 Machine
Learning Algorithm”, Workshop on Computing,
III. CONCLUSION Networking and Communications, 2012 IEEE.
[6]. Raisul Islam Rashu, Naheena Haq, Rashedur M
This paper is a review of the best in class regarding Rahman, “Data Mining Approaches to Predict
EDM and reviews the most applicable work around Final Grade by Overcoming Class Imbalance
there to date. It would be exceptionally hard to Problem”, 17th International Conference on
physically experience the tremendous arrangement of Computer and Information Technology (ICCIT)
scholarly records to distinguish the understudy 2014.
patterns and conduct and the example in which they [7]. Hina Gulati, “Predictive Analytics Using Data
learn. Rather, if client makes utilization of data Mining Technique”, 978-9-3805-4416-
mining strategies on the expansive measure of 8/15/$31.00c 2015 IEEE.
scholastic record, he/she can without much of a [8]. Rutvija Pandya , Jayati Pandya , “C5.0
stretch gathering the understudies, distinguish Algorithm to Improved Decision Tree with
shrouded designs about their learning styles, discover Feature Selection and Reduced Error Pruning ”,
unfortunate understudy conduct and perform International Journal of Computer Applications
understudy profiling. In this way, data mining can (0975 – 8887) Volume 117 – No. 16, May 2015.
unquestionably be a vital apparatus and part of [9]. Alana M. de, Morais and Joseana M. F. R.
mechanically progressed educational systems. Araújo, Evandro B. Costa, “Monitoring Student
IV. REFERENCES Performance Using Data Clustering and
[1]. Carlos Márquez-Vera, Cristóbal Romero Predictive Modelling”, 978-1-4799-3922-
Morales, and Sebastián Ventura Soto, 0/14/$31.00 ©2014 IEEE.
“Predicting School Failure and Dropout by [10]. Kin Fun Li, David Rusk and Fred Song,
Using Data Mining Techniques”, Ieee Journal “Predicting Student Academic Performance”,
Of Latin-American Learning Technologies, Vol. 2013 Seventh International Conference on
8, No. 1, February 2013 IEEE. Complex, Intelligent, and Software Intensive
Systems.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


238
Mayuri S. Dongre et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 235-239

[11]. Mishra, T., Kumar, D. Gupta, S.,” Mining


Students' Data for Prediction Performance”,
Advanced Computing & Communication
Technologies (ACCT), 2014 Fourth
International Conference on 8-9 Feb. 2014.
[12]. B.K.Bhardwaj and S.Paul, “Mining Educational
Data to Analyze Students Performance”,
International Journal Advanced Computer
Science and application Vol. 2 No. 6, 2011.
[13]. Yohannes Kurniawan, Erwin Halim, “Use Data
Warehouse and Data Mining to Predict Student
Academic performance in Schools: A Case
Study (Perspective Application and Benefits)”.
[14]. B.K.Bhardwaj and S.Paul, “Mining Educational
Data to Analyze Students Performance”,
International Journal Advanced Computer
Science and application Vol. 2 No. 6, 2011.
[15]. Lewis, R.J. (200). An Introduction to
Classification and Regression Tree (CART)
Analysis. 2000 Annual Meeting of the Society
for Academic Emergency Medicine, Francisco,
California.
[16]. Mehta, M., Agrawal, R., and Rissanen, J. (1996).
SLIQ: A fast scalable classifier for data mining.
In EDBT 96, Avignon, France
[17]. Shafer, J., Agrawal, R., and Mehta, M. (1996).
Sprint: A scalable parallel classifier for data
mining. Proceedings of the 22nd international
conference on very large data base. Mumbai
(Bombay), India.

Cite this article as :

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


239
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195166

E-Toll System Using RFID


Ankita Bambole , Shrishti Thakur1, Suvidha Ramteke1, Sneha Borkar1, Hitesh Kasturi1, Prof. M. M. Baig1
1

1Department of Computer Science and Engineering, JD College of Engineering and Management, Nagpur,
Maharashtra, India
2Assistant Professor, Department of Computer Science and Engineering, JD College of Engineering and

Management, Nagpur, Maharashtra, India

ABSTRACT

ETS (E-Toll System) is an electronic Toll Tax collection system by which we can collect the toll charges using
RFID card detection and scanning.These RFID Cards are related to the respective vehicle holders.This project
hence, reducing common problems like traffic congestion, Pollution caused due to vehicles, Time dilemma,
requirement of labours etc. Makes it effective to collect the required amount of charge from the driver. There
will be an RFID tag implemented on every vehicle. Whenever vehicle passes through the toll plaza RFID
reader with the help of sensors will detect tag and the required amount will be deducted. Driver don’t need to
stop to pay the money ,and this RFID card is used to track & bill the vehicle owner through a payment gateway.
Keywords : Electronic Toll Collection, RFID, XAMPP, Vehicle, Tag.

I. INTRODUCTION Sanral:-E-toll (in South Africa) consists of the


electronic toll collection (ETC) processes employed
The world is changing furiously, the role of by South Africa's roads agency Sanral on selected toll
automation in our day to day life is increasing at a roads or toll lanes, subject to the Sanral Act of 1998.
very fast rate, hence the consumption of fuel and use
of road is increasing proportionally. People charge of FASTag:-FASTag is an automatic electronic toll
using roads as nothing in this universe is free of Cost. collection system operated by National Highway
This leads to the motive of our project i.e. “The Authority of India[1] (NHAI). It is also based on
Automation”.We have chosen this topic because Radio Frequency Identification (RFID) technology
digitizing current toll plaza will overcome for making toll payments from savings account linked
disadvantages like time delay, traffic congestion etc. to it
The charge form of manual and semi-manual will not
meet demand of the charging management system E-ZPass:-The E-ZPass Group is comprised of toll
and may lead to many vehicle blocked at entrances entities stretching across 17 states that operate the
and exits causing huge economic losses when it extremely successful E-ZPass electronic toll
reached a certain threshold enhancing the collection program.
convenience and safety of the travelers.
The current status of RFID-based ETC system
worldwide exampled and described as follows:

CSEIT195166 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 240-243 ]
240
Ankita Bambole et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 240-243

II. THE COMPOSITION processing will be done. Transaction will begin


depending upon the balance in the account of the
Our Toll Collection system is madeup of following drivers card the amount will be deducted as the toll
modules:- tax if the vehicle has to encounter another toll plaza
through the journey she can request for deduction of
• Hardware part: This is a hardware-cum-sensor amount wrt the upcoming toll also.All these tolls are
system which manages vehicle detection and is connected using the centralized server using LAN.If
installed at the toll booth. The hardware is based there is no sufficient amount of money present the
on Arduino boards development platform. user can pay manually or other transactions means.
• Server: This is the backend server hosted in The software updates the data of the transaction after
XAMPP which is responsible for authenticating, the process successfully terminated.
identifying and billing a vehicle based on the
unique RFID tag. Thus it manages the complete Following is the algorithm associated with the
toll collection process. project:-
• Database: Consists of user and administrator data
and information regarding the total transaction Step 1: Start
and input output of the vehicles in and through Step 2: Sensors 0.5 km before the toll plaza detects
the toll plaza. transport.
• User website: It is a web site frontend we have Step 3: Scanner and Camera will start processing
created in PHP for users and toll people to according to step 2.
register new drivers,Tolls,Vehicles and updating Step 4: Vehicle slow down for some seconds at the
information regarding the vehicles. same time High quality scanner scans QR-Code
Step 5: if Vehicle= Truck or Other transport then
III. DESIGN AND IMPLEMENTATION return the amount of tax to be deducted.
Vehicle = special Transport then
When a person buys a new vehicle she’ll need to Return(null);
register herself and her vehicle at the toll office or Step 6: At the same time the high quality Camera
RTO office. The RTO officer will allot a unique ID to clicks picture of the vehicle.
the RFID card the user along with No. Plate which
will contain a unique ID related to the Card only. At If{
this time the officer will also create an account for Traffic rules are followed then
the user although user can herself make her account return (null);
using the website by signing up. After it has else
registered the vehicle a password and user id will be return(penalty);
generated and once it is generated the user can Step 7: if the Transaction is completed
recharge his/hers card any time. successfully then the Tax will be delivered on their
address.
Each time the drivers vehicle approaches the toll Step 8:Message will be sent on owners phone
plaza the IR sensor implemented on the Step 9:END
plaza(Hardware)will sense the vehicle approaching
and sends a ready signal to the RFID reader and
Scanner the scanner will scan the Card and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


241
Ankita Bambole et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 240-243

IV. HARDWARE SETUP (a)RS232 compatible output and (b)TTL compatible


output. A TTL compatible output pin is connected
At the toll booth, Arduino board hardware are directly to Arduino.
situated which manages the toll total collection
processes .We have used Arduino UNO which is a
microcontroller board ATmega328P consisting of 14
digital input/output pins ,6 analog input ,a quartz
crystal, USB cable connection , a power supply jack,
ICSP header and a reset button.
We have simply connected it to the computer using
USB cable.

Fig 3. RFID with Arduino Uno Rev3

V. DATABASE

Database is the Heart of our project. We have used


MY-SQL for database query generation and updating
the entries. We obtained a database for user, admin
Fig1. Arduino Uno Rev3,Code: A000066 and transaction using XAMPP control panel .There
are three Data information’s stored in the database as
To track the vehicles, Arduino UNO is interfaced follows:-
with the RFID reader and scanner for getting the 1.Admin’s Database:-Administrator database consists
information regarding the vehicle. of all the details of central database and all tolls under
Construction.I.e. RTO’s Database.
Both the RFID reader and the tags comes with the 2.Central Database:- Central database consist records
coil in them power supply is given and the data is of all toll plaza in work. This is managed by the RTO
fetched We collect the read data. administrator.The Customer registers themselves into
this account system. This account information about
the users is stored into The RTO database. When the
registered customer passes through the particular toll
plazas then, automatically data will be updated.
3.Integrated database:-Integrated database is
disconnected to the central database This database
will update automatically. This database includes all
Registered vehicles and the details of vehicle such as
Fig 2. Interfacing Arduino with RFID vehicle owner, vehicle number, license number,
account ID, account balance,current charges, etc.
.Power supply requirement of RFID Readers varies
for different products. Reader we have used is of 12
Volts. There are 2 outputs from RFID Reader.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


242
Ankita Bambole et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 240-243

VI.CONCLUSION Cite this article as :

In this report, the conclusion that came out is that, to Ankita Bambole, Shrishti Thakur, Suvidha Ramteke,
implement modern system of toll collection new Sneha Borkar, Hitesh Kasturi, Prof. M. M. Baig, "E-
technique RFID came into light. The RFID safety is Toll System Using RFID", International Journal of
key important o this project. Scientific Research in Computer Science, Engineering
This project is reliable and easy way to pay toll as and Information Technology (IJSRCSEIT), ISSN :
compared to manual one. 2456-3307, Volume 5 Issue 1, pp. 240-243, January-
February 2019.
Project is all about automated toll system which will Journal URL : http://ijsrcseit.com/CSEIT195166
overcome disadvantages of current toll system. Our
project helps driver to pass through toll without
stopping his vehicle to pay the toll tax. RFID tag on
each vehicle will be detected by RFID reader and
amount will be deducted. Open Road Tolling
System[2] and Automated Toll system[3] these are
two previously implemented projects similar to our
project. We have studied them and we are trying to
overcome disadvantages of these two. The project if
implemented by government then toll system will be
fully automated and one who is passing through toll
don’t need to wait for a long time this will also
reduce pollution, traffic and accidents.

VII. REFERENCES

[1]. Electronic Toll Collection System in India.


Operated by the National Highway Authority
of India
https://en.wikipedia.org/wiki/National_High
way_Authority_of_India.
[2]. OPEN ROAD TOLLING - IMPROVED
SAFETY, LESS CONGESTION RSS FIRST
PUBLISHED IN ITS INTERNATIONAL
JANUARY FEBRUARY 2010 AS ORT: THE
SILVER BULLET FOR SAFETY
[3]. AUTOMATIC TOLL COLLECTION
ITERNATIONAL JOURNAL OF
COMPUTER SCIENCE AND MOBILE
COMPUTING, VOL.5 ISSUE.8, AUGUST-
2016

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


243
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/ CSEIT195165

A Review on Indoor Navigation Systems and Technologies for


Guiding the Users
Meenakshi D. Dhobale1, Pooja S. Verma1, Pushpa R. Joshi1, Jyotsna K. Tekade1, Swati Raut2
1BE Students, Department of Computer Science and Engineering, JD College of Engineering and Management,
Nagpur, Maharashtra, India
2Assistant Professor, Department of Computer Science and Engineering, JD College of Engineering and
Management, Nagpur, Maharashtra, India

ABSTRACT

GPS is in actuality the main answer for outdoor positioning, no unmistakable arrangement has so far risen for
indoor positioning notwithstanding the business hugeness. Along these lines, the principal point is to grow
ongoing indoor following framework and utilizing cell phone sensors to give dependable and precise indoor
localization. Indoor navigation is a vital empowering innovation for application, for example, finding a
gathering room in a place of business, security exit amid a crisis or focused on the retail commercial in a
shopping center. Utilizing this thought, an application can be made and by methods for that application, we are
handling indoor navigation utilizing continuous navigation. The android application gives mindfulness,
straightforwardness to discover your direction, commercial and effortlessness. To give solid indoor navigation
to business structures like shopping centers, enterprises, universities, hospitals, inns. In this paper, we give an
order and arrangement of the current indoor positioning systems and distinguish some conceivable zones of
improvements.
Keywords : Indoor Navigation, Pedestrian Navigation, GPS, Object Tracking, Location Based Services

I. INTRODUCTION and Non-Line-of-Sight (NLoS) conditions, nearness


of moving individuals that adjust the indoor
Recently, the utilization of indoor localization proliferation channel, more noteworthy thickness of
methods has turned out to be progressively critical in obstructions that reason a high weakening and flag
an extensive number of uses and settings, for example, dissipating, request of a higher exactness and
human services, homecare, checking, following, and precision.
so forth. In outside localization settings, the most
notable and broadly spread innovation is the Global In the decision of the best innovation to plan an
Positioning System (GPS). It can ensure fantastic indoor localization framework, an extensive number
execution in outside situations yet not work of parameters ought to be considered (e.g., cost,
legitimately in indoor conditions because of poor precision, vigor, versatility, inclusion, and so on.).
inclusion of satellite flag. Additionally, acquiring Clearly, a solitary arrangement that works fine for
position data in indoor situations is especially testing any situation does not exist. At that point, it is vital to
a direct result of a few reasons: mistakes by multipath circumspect the execution parameters all things

CSEIT195165 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 244-250 ]
244
Meenakshi D. Dhobale et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 244-250

considered and coordinate them with the client II. LITERATURE REVIEW
prerequisites, which must be broke down and
portrayed decisively for every application. In Scientists and organizations have made considerable
addition, the estimations of the execution parameters commitments to the theme of Indoor Positioning
are not univocally definable since they thus rely systems and conditions. Indoor Positioning zone is
upon different factors and conditions. Consequently, brimming with valuable commitments. In this
it is important to locate the correct exchange off segment, we talk about the diverse indoor conditions
among execution parameters, client necessities, and and the best in the class of indoor positioning systems.
natural conditions so as to plan a modified
arrangement. A. Fixed Indoor Positioning

In an outside situation, the Global Positioning System Normal design of Fixed IPS systems is having settled
(GPS) works productively in positioning and focusing number of Base Stations (BS), which is the
on various sorts of substances. It has been utilized in fundamental PC that gets every one of the signs from
numerous outdoor applications for limiting the diverse hubs and ascertains the directions of the
individuals, autos, just as different items. Be that as it objective article utilizing the parameters sent from
may, GPS comes up short on a similar dimension of the sensor hubs. All BS are required to be introduced
effectiveness when utilized inside indoor situations. in settled areas inside the building. In addition, a
This issue is because of the presence of hindrances Wireless Sensor arranges including the sensor hubs
that can debilitate the flag of the GPS (e.g., building are required. In this structure, the sensor hubs are in
engineering, dividers) where the presence of various charge of proliferating the flag got from a portable
gear can cause a commotion in the GPS transmitted station to the base station just as doing the basic
flag. count of the goal or time of transmissions. This
conduct relies upon the structure of the framework.
Pedersen [1] proposed a miniaturized scale The third principle segment of the settled indoor
positioning procedure that ought to be actualized positioning engineering is the cell phones or labels
inside the indoor condition so as to position and track conveyed by individuals or joined to target objects.
objects. He expressed that this procedure would fill in These labels give a novel recognizable proof to each
as a trade for the GPS positioning framework. item or individual. Along these lines, blunders will be
Furthermore, Fhelelboom [2] found that a remote restricted. Next, we list every remote innovation
neighbourhood (WLAN) can be utilized inside any with the systems, which are produced dependent on
indoor condition to position objects. this innovation. Elements of IPS are talked about in
[3].
In this paper, we audit the distinctive positioning
conditions, the diverse systems connected for every a. Infrared Positioning Systems
condition and the calculations utilized inside every
framework. We indicate two situations for These systems utilize infrared flags so as to transmit
positioning individuals and articles inside an indoor signals from sensor hubs to the BS. A standout
domain. Every one of these situations has its very amongst the most well-known infrared positioning
own difficulties, which specialists attempted in the systems is the dynamic identifications created by
past to moderate by proposing a few arrangements AT&T Cambridge [4]. In this framework, clients
throughout the most recent ten years. convey an ID card furnished with an infrared LED.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


245
Meenakshi D. Dhobale et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 244-250

The infrared LED sends a special code at regular and its directions. Cricket utilizes the Time
intervals. Moreover, there are infrared sensors Difference of Arrivals (TDOA) credit so as to figure
introduced on a roof and if the IR identification is the directions of the objective item. That is by
inside six meters, the sensor can peruse the code. The ascertaining the contrast between the Radio
BS gets the information from the IR sensors Frequency flag Time Arrival and the ultrasonic flag
occasionally. At last, the BS can assemble a guide of Time Arrival. The Difference of the season of the
each identification area utilizing the data recovered landing of the two signs is determined to utilize
from the sensors. Dynamic Badges have for the most conditions since the Radio Frequency is a lot quicker
part four directions, WITH, LOOK, NOTIFY, and than sound. At that point, the TDOA is equivalent to
HISTORY, every one of which gives an alternate the separation the flag went over the speed of the flag.
capacity. For instance, WITH demonstrates the The Cricket framework is professed to give a
identifications that exists in the sensor region, LOOK precision of up to 2 cm. Be that as it may, when
is utilized to search for an identification by a sensor, confronting NLOS blunder, the Cricket framework
NOTIFY is utilized to tell the BS when the does not give an exact computation of the facilitate.
identification is found and HISTORY demonstrates
the identifications positions over a specific timeframe. Popa [8] proposes joining Cricket with Inertial
Navigation sensors. This blend gives a superior
b. Ultrasonic Positioning Systems precision and relief of NLOS blunder as guaranteed
by Popa. That is on the grounds that Inertial
Ultrasonic reference points are utilized more Navigation sensors have a more extensive inclusion
frequently than infrared innovation. Ultrasonic and their flag isn't influenced by the NLOS mistake.
systems give increasingly exact positions to objects.
Ultrasonic based systems are more exact than Radio c. RF Positioning Systems
Frequency based systems, as we will find in the talk
area. Nevertheless, Ultrasonic systems need a settled The for the most part utilized remote innovation is
structure of the framework [5]. Radio Frequency (RF). This is because of the minimal
effort and the high scope of inclusion of the systems
Instances of the systems created dependent on created dependent on RF innovation. Some RF based
Ultrasonic innovation are the Active Bats [6] created systems that we will examine in this paper are
by AT&T Cambridge as well. This framework has a RADAR, Spot-On, LANDMARC, and UWB systems.
comparative structure to the dynamic identifications.
LANDMARC framework [9] depends on Radio
The Crickets framework [7] created by the Frequency signs and RFID labels. LANDMARC
Massachusetts Institute of Technology (MIT) is decreases the expense of utilizing RFID per users by
another outstanding ultrasonic based indoor lessening the quantity of per users and utilizing
positioning framework. The Cricket framework has reference labels. These reference labels have a well-
two kinds of hubs, signals, and audience members. known area and they transmit to the per users the
Reference points' areas are settled and they are area of the objective items. LANDMARC is a decent
connected to the roof while the audience members framework however its exactness is 1-2 meters.
are appended to the objective items and individuals.
Guides send intermittent data to the BS containing its Guang [10] gave a component to enhance the
ID, scope of inclusion or physical space related to it execution and exactness of LANDMARC. This

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


246
Meenakshi D. Dhobale et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 244-250

component works by lessening the quantity of


applicants of reference labels while ascertaining the B. Indoor Pedestrian Positioning
situation of an article. This diminishes the count
exertion and results in a quicker computation and a The person on foot positioning, as referenced prior in
superior precision. When testing LANDMARC the paper, happens when finding individuals who are
utilizing this component, Guang claims the precision conveying localization sensors while the building is
of LANDMARC was preferred utilizing this system not furnished with an indoor positioning framework.
over utilizing the conventional LANDMARC. In these kinds of systems, Inertial Navigation Systems
Guang's component gives a blunder rate. or dead retribution are for the most part utilized.
Dead Retribution is characterized as a navigation
Jiang [11] proposed a framework that gives open air procedure that begins with an outstanding area [13].
positioning utilizing GPS and UWB to give indoor At that point, includes the position changes in the
positioning. Jiang's framework contains PDA, UWB directions of the beginning stage. It likewise adds the
sensor organize and a Base Station. GPS programming progressions to the heading (course), speed or
is introduced on the PDA just as an interface for the separation. Besides, Pedestrian Dead Reckoning (PDR)
UWB sensor organize. At the point when the client is characterized as assessing the speed of development
moves outside the building the GPS application is and the heading or bearing of development. We will
actuated and when the client moves inside the demonstrate a portion of the intriguing commitment
building the system perceives the gadget and the to this sort of positioning.
UWB application is initiated to empower the client to
explore through the building and enables the Base A few methodologies in Indoor Pedestrian
Station to position this client. The exactness was Positioning use Particle Filters, for example, Bayesian
observed to be inside 10 meters which needs a ton of Filter and Kalman Filter. A Bayesian channel [14] is
upgrades later on. utilized to appraise the progression of the passer by at
a specific time when knowing the past strides of a
d. Optical Indoor Positioning similar walker at a number of times before it. Kalman
Filter [15] is a scientific model which is utilized to
Optical Indoor Positioning is another sort of settled precisely appraise the situation with the presence of
indoor positioning where we have a framework commotion.
introduced in the building and a camera conveyed by
the client. Beauregard [16] utilizes the Pedestrian Dead
Reckoning methodology so as to position. He utilizes
Tilch [12] proposed CLIPS (Camera and Laser based head protector mounted sensors as a novel
Indoor Positioning System). This framework methodology for person on foot positioning. That is
consolidates the two advances to position objects on the grounds that the protective cap is the most
indoor. The camera goes about as the cell phone for astounding position the sensor can reach as expressed
positioning articles. The laser gadget is situated by Beauregard. The calculation utilized in this
towards the roof and laser pillars are on the roof. The framework has two stages, step discovery and
camera tracks the laser shafts and modifies its estimation and heading location and estimation. In
introduction concerning the laser bars area. Optical recognizing and evaluating the progression, they
indoor positioning is utilized more for robot self- distinguish the speed of the development and the
localization inside an indoor domain. length of the progression while in heading location

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


247
Meenakshi D. Dhobale et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 244-250

and estimation the bearing of development. The Project onto way, (iii) Find the objective position,
constraint in this framework is that the sensor on the and (iv) Compute the direction point.
protective cap must be coordinated to the bearing of
the person on foot's development. Woodman [19] proposed a Bat framework. Bat
framework utilizes a foot-mounted inertial unit for
Another methodology given by Robertson [17], he Pedestrian Dead Reckoning and a Wifi innovation to
proposes utilizing foot-mounted inertial sensors position people on foot inside a multi-story building.
which give the person on foot dead retribution. He Bat framework utilizes molecule channel calculation
gave a framework called FootSlam. This framework which is Bayesian channels. Woodman proposed an
utilizes a foot mounted Inertial Measurement Unit adjustment calculation to address the float in inertial
(IMU). In addition, it assembles a 2D guide of the estimations with the person on foot development.
working with no earlier information of the structure Woodman's calculation helps in assessing the
of the building. The learning of all client expectation situation of the person on foot at time k when
assists with shaping the guide of the building and knowing his situation at time k-1.
managing the client through it. Notwithstanding, the
more occasions places are visited inside the building Shao [20] talks about a relative region; he executed
the better data or guide worked in respects of that self-ruling people on foot in a 3D domain. He
put. Recreation results demonstrate a precision of 1-2 demonstrated how his self-sufficient people on foot
meters. associate inside nature and how they fabricate maps
of the earth. He characterized the maps as
Fischer [18] recommends utilizing Ultrasound Topological guide, recognition guide and way outline.
reference points to give better precision and less The topological guide speaks to the ecological
heading mistakes in Pedestrian Dead Reckoning. His districts and the bolts speak to the openness between
methodology joins Ultrasonic and PDR. The PDR them. The discernment maps include stationary
calculation contains two sorts of stages inside the articles, which are the nearby items. Where the
means, a positioning stage, and a swing stage. Fischer versatile articles are the worldwide items. At long last,
saw there are likewise two kinds of mistakes that can the way maps can be either framework, which is for
happen inside PDR, these blunders are the heading short way mapping (e.g., seat) or Quadtree which are
blunder and the separation blunder. As examined for long way arranging. Shao's maps technique for his
before, the heading is the bearing of the development; independent people on foot may be extremely
in this way, the heading blunder is the mistake valuable whenever utilized for walker direction
happens in the appeared of the walker. The through indoor positioning.
separation blunder as expressed by Fischer has fewer
events. Fischer's framework is connected primarily to III.CONCLUSION
the safeguard group specialist on call. The framework
functions as pursues; the specialists on call put the In this paper, we have surveyed indoor positioning
ultrasonic sensors on their way as they continue systems, the standards of positioning and calculations
inside the building. PDR framework is utilized to get utilized in different conditions. We talked about the
the area of the people on foot. Be that as it may, difficulties that can be confronted when planning an
ultrasonic sensors influence the modifications so as to indoor positioning framework, expounded on various
lessen the heading mistake. His calculation has four arrangements proposed to beat these difficulties and
stages (I) Estimate the individual's position, (ii) given a conceivable zone of utilization and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


248
Meenakshi D. Dhobale et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 244-250

improvement inside indoor positioning systems. We Workshop on Software Technologies for Future
saw Indoor Positioning is a tremendous zone with Embedded Systems.
numerous applications and numerous enhancements [9] Lionel M.NI, Yunhao Liu, Iu Cho Lau,
to be conveyed. As referenced before in this paper, Abhıshek P. Patil; LANDMARC: Indoor
explores inside indoor positioning could be Location Sensing Using Active RFID
increasingly valuable whenever conveyed for people [10] Guang-yao Jin, Xiao-yi Lu, Myong-Soon Park.
on foot positioning since the exactness came to isn't 2006. An Indoor Location Mechanism Using
as precise as the precision came to in settled indoor Active RFID. Proceedings of the IEEE
positioning. International Conference on Sensor Networks,
Ubiquitous and Trustworthy Computing.
IV.REFERENCES [11] Lijan Jiang, Lim Nam Hoe, Lay Leong Loon.
2010. Integrated UWB and GPS Location
[1] Solrun Furuholt Pedersen, "Micro Positioning". Sensing System In Hospital Environment. IEEE
Master Thesis. ITEM NTNU, Jun 15, 2014. International Conference of Sensor Networks.
[2] Fhelelboom, Zuher. "Equipment tracking and [12] Tilch, S. Mautz, R. 2010. Current investigations
security systems for hospitals". Master Thesis at the ETH Zurich in optical indoor
Universitiy Teknology Malaysia. May, 2017. positioning. IEEE Positioning Navigation and
[3] I. Guvenc, C.-C. Chong, “A Survey on TOA Communication (WPNC) conference.
Based WirelessLocalization and NLOS [13] S. Beauregard and H. Haas. Pedestrian dead
Mitigation Techniques”, IEEE Communications reckoning: A basis for personal positioning. In
Surveys and Tutorials, vol. 11, no. 3, 2009, Proceedings of the 3rd Workshop on
Pages: 107-124. Positioning, Navigation and Communication,
[4] R.Want, A.Hopper, V.Falcao and J.Gibbons; 2006.
The active Badge location system, ACM [14] V. Fox, J. Hightower, L. Lin, D. Schulz, and G.
Transactions on Information systems Vol. 40, Borriello, “Bayesian filtering for location
No. 1, pp. 91-102, January 2012. estimation,” IEEE Pervasive Computing, vol. 2,
[5] Hazas, M., Hopper, A; A Novel Broadband no. 3, pp. 24–33, 2003.
Ultrasonic Location System for Improved [15] Matthies, L. H., Szeliski, R., and Kanade, T.
Indoor Positioning, IEEE Transactions on (1989). Kalman filter-based algorithms for
mobile Computing, Vol. 5, No. 5, May 2006. estimating depth from image sequences.
[6] Michael Popa, Junaid Ansari, Janne Riihij¨arvi, International Journal of Computer Vision, 3,
and Petri M¨ah¨onen. 2008. Combining 209–236.
Cricket System and Inertial Navigation for [16] Beauregard, S., (2006). A Helmet-Mounted
Indoor Human Tracking. WCNC proceedings. Pedestrian Dead Reckoning System.
[7] Priyantha, N. B; The cricket indoor location Proceedings of IFAWC2006, TZI University
system: PhD Thesis, Massachusetts Institute of Bremen, Germany, pp. 79-89
Technology. 199 p, June 2005. [17] Robertson, P., Angermann, M, Krach, B.,
[8] Fukuju, Y.; Minami, M.; Morikawa, H.; Simultaneous Localization and Mapping for
Aoyama, T.; Dolphin. 2003. An autonomous Pedestrians using only Foot-Mounted Inertial
indoor positioning system in ubiquitous Sensors. In Proc. Ubi Comp 2009, ACM (2009)
computing environment, in Proc of the IEEE 93-96.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


249
Meenakshi D. Dhobale et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 244-250

[18] C. Fischer et al., “Ultrasound-Aided Pedestrian


Dead Reckoning for Indoor Navigation,” Proc.
Int’l Workshop Mobile Entity Localization and
Tracking in GPS-Less Environments, ACM
Press, 2008, pp. 31–36.
[19] O. Woodman and R. Harle. Pedestrian
localisation for indoor environments. In Proc.
of the UbiComp 2008, Seoul, South Korea, Sept.
2008.
[20] SHAO W., TERZOPOULOS D.: Autonomous
pedestrians. In SCA ’05: Proceedings of the
2005 ACM SIGGRAPH/Eurographics
symposium on Computer animation (New
York, NY, USA, 2005), ACM Press, pp. 19–28.

Cite this article as :

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


250
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/ CSEIT195175

A Review Study on Online Job Portal


Aafreen Khan1, Anirudha Wankhade1, Priyanka Pakhide1, Sagar Meshram1, Prof. Sonali Zunke2
1BE Students, Department of Information Technology, JD College of Engineering and Management, Nagpur,
Maharashtra, India
2Assistant Professor, Department of Information Technology, JD College of Engineering and Management,

Nagpur, Maharashtra, India

ABSTRACT

Gaining information and explicit job skills have turned into the fundamental destinations for understudies in
the colleges. Learning is important to settle on educated choices, particularly, in a basic circumstance. Learning
and information the executives in any organization are pivotal to give it an aggressive edge in the present
testing and globalized condition. In this paper we present the plan of different on-line recruitment framework,
that enables businesses to post their job advertisements, which job searcher can allude to, when searching for
jobs. This job portal can catch job prerequisites dependent on industry needs.
Keywords: Knowledge Sharing, Web Portal, Job Portal, Online Recruitment.

I. INTRODUCTION to get another line of work, he/she can present a


resume utilizing word handling programming like
Joblessness is one of the genuine social issues Microsoft Office Word, open an internet browser to
looked by both creating and created nations. For send the resume and get an email. Online
instance, in Europe the rate of joblessness has been recruitment has turned into the standard strategy for
expanding quickly since the 1970's. Dorn and Naz [1] businesses and jobseekers to meet their particular
referenced that one reason for this issue is the targets. The businesses transfer the job contributions
unjustifiable appropriation or absence of data on job in to the job portals. The majority of substantial
openings so individuals cannot presently the new job organizations as well as the little ones has
opportunities. It implies that there are a few jobs acknowledged online recruitment. The organizations
accessible, yet jobseekers do not approach that data. send data or jobs opening for posting on the portals
An effective pursuit of the web may help to and speak with the candidates through the Internet
jobseekers in their job chase. Some web-based and Email.
interfaces give a productive method to look the web
for online data on job opportunities for jobseekers [2]. Gangle [4] characterized the idea of online
recruitment or e-recruitment as the utilization of the
Today, the web has changed numerous parts of our Internet to scan for jobs, which have been publicized
life, for example, the manner in which we search for electronically. Along these lines, the businesses
jobs [3]. On the off chance that one individual needs publicize the job openings, spare the resume and

CSEIT195175 | Received : 01 Feb 2019 | Accepted : 09 Feb 2019 | January-February - 2019 [ 5 (1) : 251-255 ]
251
Aafreen Khan et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 251-255

educational modules vitae (CV) of candidates, and different arrangements they need to make. Securing
contact the jobseekers who are qualified, on the web. all accessible position opening is a principle venture
Today, the Internet has turned out to be one of the at in the job-chasing process.
key strategies for landing data identifying with
position opening. Substantial foundations, similar to The Internet is presently an incredible asset that
colleges incorporate data on vocation prospects in jobseekers can utilize. Today, numerous locales
their sites, which are likewise connected to publicize job positions to be filled by individuals with
recruitment locales. specific skills in different fields. The Internet assumes
an essential job in the region of human asset
II. LITERATURE REVIEW arranging and advancement. Most arranging and
advancement organizations are currently utilizing PC
A. Job Procurement: Old and New Ways innovation and the Internet for staff recruitment. It
Job looking for more often than not includes ought to be noticed that despite the fact that the
distinctive approaches to search for jobs, for example, Internet has encouraged the procedure of job chasing,
through close to home contacts, direct phone calls to it has not supplanted the conventional techniques,
businesses, job organization office, filtering on the totally.
web job postings, and so on [3]. Prior to the Internet,
turned out to be generally utilizes as a strategy for B. Significance of Job Portals
looking for jobs; jobseekers invested a heap of energy
utilizing different strategies to search for job In the period of innovation, the Internet has
openings. Today, jobseekers utilize online techniques, turned into the principle wellspring of data for
which are extremely advantageous and spare a great jobseekers. Extensive partnerships, foundations, and
deal of time. Galanaki [5] records the accompanying colleges incorporate data on vocation prospects on
strategies to be the conventional (old) routes for their sites. As per a review, 70% of the workforce
recruitment: utilizes sites or portals on the Internet to look for jobs
in France. These sites or portals give an internet
• Employment recruitment offices searcher to get to data on job openings [7].
• Job fairs
• Advertising in the broad communications, for Sulaiman and Burke [8] found that most businesses
example, papers are quick to utilize online recruitment strategies for
• Advertisement in TV and radio getting staff. He referenced that online recruitment
• Management Consultants strategies can recognize the best candidates. That is
• Existing worker contacts the motivation behind why progressively created
• Schools universities or colleges understudies nations, for example, Malaysia have begun to utilize
administrations office online job portal as one of the vital method to enlist
• Workers or expert referrals individuals to fill job opening. An examination done
in 2006, found that 21% of web clients in the EU
These old jobs looking for techniques are utilized the web to look for jobs or to send job
excessively moderate, distressing, testing and applications. In 2007, this had expanded to 67% for
furthermore need quality [6]. What's more, the jobless individuals [5].
candidates need to consider the expense and the
measure of time to get the data they need, and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


252
Aafreen Khan et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 251-255

Most organizations distribute their job opening on individuals approach the job data from anywhere on
their site, or utilize online jobsites. These techniques the planet, while with the paper, data is spread at
result in incredible sparing in expenses. Mochol and nearby dimension. One key part of job portal is the
Nixon [9] expressed that the utilization of semantic expense. Organizations spend less to distribute or
web innovation gives advertise straightforwardness, publicize job opportunities on the portals or sites,
higher speed of acquisition yet decreased exchange when contrasted with the utilization of other media,
cost. Today, the Internet is utilized for an expansive for example, paper or job fairs. Besides, online
number of business exchanges. Individuals observe recruitment is quick, and spares time. Once the
the Internet to be a compelling specialized apparatus. managers transfer the job opening on the portal, the
In a report in 2005, it was discovered that 90% of jobseekers can see it and send in their resume.
jobseekers in Germany utilize the web to search for
jobs. A purpose behind this high rate of Internet is Along these lines, cost and efficient are two
that candidates are youthful and very qualified and noteworthy points of interest of job portals. Ganalaki
utilize the web a great deal, and numerous [5], refered to in Rosita and Nadianatra (2007),
organizations distributed their job openings on the expressed that the Internet is an instrument without
web and through their portal. fringes, and is an amazing strategy to come to an
around the world gathering of people.
Job portals are the beginning stage of jobseekers
while scanning for jobs. In this way, some job portals C. Highlights of Job Portals
charge businesses high expenses to distribute data on
job opportunities. Regardless of this, numerous One of the approaches to enhance work portability
businesses still keep on publicizing or distribute data is to give online job offer administrations. Online job
on job openings hands on portal, however portals can help jobseekers as they contain all
constrained to minimize expenses. Numerous required data about accessible opening in a solitary
businesses still trust that a jobseeker will visit job point. Such portals upgrade effectiveness in job
portals while hunting down job opening [10]. recruitment as candidates can coordinate their
capabilities and skills to the prerequisites of bosses.
A great job portal can likewise bolster learning
sharing among the individuals. The quantity of For the most part, looking for jobs on the web
online job portals keeps on expanding. It is trusted includes a procedure of data gathering in light of the
that seventy-five percent of individuals who are fact that the jobseeker accumulates data contained in
looking for jobs, utilize the web and on the web the job portals, amid the inquiry [11]. A great job
portals. Gangle [4] expressed that online recruitment portal imparts data and encounters to its
has the accompanying points of interest: bosses can individuals/clients. This spare time and endeavours
distinguish an extensive number of qualified job and better choices can be made [12]. Job openings
searchers and get their data effectively. It implies that necessities can be coordinated to a candidate's
organizations or organizations can broaden the capability and skills. Along these lines, job portals
inquiry area, subsequently, they have better prospect return the exact matches as well as return the most
of choosing the most qualified applicants. Web gives comparable match. The individuals from the
managers an approach to draw in a higher number of European Commission (EC) expressed that online job
applicants, particularly, the individuals who satisfy portals ought to have very comparable attributes. The
the job necessities. With online recruitment, attributes included are:

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


253
Aafreen Khan et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 251-255

• An online accessible database of positions for job portals to have the necessary information for the
searcher; offices to send cvs to the site; students. Fig. 1 shows the schema of the proposed job
• Email alarms of jobs which coordinate the clients web portal, in illustrating the relationship between
profile; the students and the faculty, and the unemployed
• Additional guidance, for instance, about working graduates and the potential employers.
in remote nations or profession direction;
• The capacity to oversee job applications; Nowadays, most modern universities web portals
• Businesses must be able to distribute and oversee provide information and services to achieve various
job positions, look through the CV database; goals and to improve the quality of education with
• Have online contact with potential jobseekers. these services. Based on information from the
literature review, as well as the students’
III.Discussion and Analysis requirements identified from the survey, a job portal
should be develop. The portal should be a knowledge
Analysis of the data collected from the survey sharing system for the students to create a better
indicates that most of students are not satisfied with academic environment for their studies.
the current faculty portal. The respondents said that
they have problem to get the needed information IV.CONCLUSION
when they are looking for a job. In addition, there is
not enough information and resources related to help In perspective of the prior discoveries, the invalid
them. The students looking for a proper Job want a speculation expressing that there is no huge contrast
knowledge sharing system to get such information in the respondents' assessment on the viability of the
and help them in their decisions. created Online Job Recruitment System regarding
execution, dependability, security, and cost-adequacy
Base the development of the new web portal should was acknowledged. It is inferred that the created
on the feedback from the survey and the programming was successful in choosing qualified
requirements of the students. In recent years, the candidates inside a shorter timeframe. Subsequently,
structure of university portals has changed from static it would turn into a noteworthy supporter of the
information to dynamic and transaction mode. It is value contract. It is additionally inferred that
very important for the faculty to provide the students Performance, Reliability, Security, and Cost-
with online access to different services and adequacy could be used as criteria in assessing web
transactions such as online enrolment, course support recruitment programming.
and job information. Suresh Ram stated that 70% of
students, who finished their studies at tertiary level In light of the ends drawn, it is prescribed that the
education, could not find a job. This shows how product is actualized in the organization. After
important it is to have access to the right information execution, it is recommended that the effect of the
to find a job. It is important for the universities that framework on the recruitment procedures is resolved
provide information to jobless graduates on the jobs and that the adjustments in the recruitment forms
or career opportunities. with the presentation of the online recruitment
programming be examined. It is proposed that
Sulaimun and Burke [8] emphasized that it is reviews on the client experience, reception or non-
necessary for the universities to pay more attention selection of the product, and mistakes submitted in
to their knowledge management system or web

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


254
Aafreen Khan et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 251-255

utilizing the product be directed and that the [8] N. Sulaiman and M. Burke, “A case analysis of
Usability model of FURPS be surveyed. knowledge sharing implementation and job
searching in Malaysia,” International Journal of
As far as an upgrade of the product, it is Information Management, 2009.
unequivocally prescribed that an online test is joined [9] M. Mochol, H. Wache, and L. Nixon,
in the recruitment and that additional security “Improving the accuracy of job search with
highlights, for example, the dimension of access semantic techniques,” in Proc. Business
grouped by the situation in the organization likewise Information Systems, 10th International Conf.,
be fused in the product. BIS 2007, Poznan, Poland, April 25-27, 2007,
Springer, pp. 301-313.
V. REFERENCES [10] M. Mansourvar, Development of a Job Web
Portal to Capture Industry’s Needs, 2011.
[1] J. Dorn and T. Naz, “Integration of Job portals [11] A. Weber and H. Mahringer, “Choice and
by Meta-search,” in Proc. 3rd International success of job search methods,” Empirical
Conf. on Interoperability for Enterprise Economics, vol. 35, no. 1, pp. 153-178, 2008.
Software and Applications, Funchal, Portugal, [12] M. Mansourvar and N. B. M. Yasin,
2007, pp. 401-412. “Knowledge portal: a tool to capture university
[2] S. Bsiri, M. Geierhos, and C. Ringlstetter, requirements,” in Proc. 2011 International
“Structuring job search via local grammars,” Conf. on Graphic and Image Processing,
Advances in Natural Language Processing and International Society for Optics and Photonics,
Applications, pp. 201, 2008. October 2011, pp. 82850F-82850F.
[3] M. Mansourvar and N. Y. Mohd, “Web portal
as a knowledge management system in the Cite this article as :
universities,” World Academy of Science,
Engineering and Technology, vol. 70, pp. 968- Aafreen Khan, Anirudha Wankhade, Priyanka
974, 2010. Pakhide, Sagar Meshram, Prof. Sonali Zunke, "A
[4] M. Gangle, “The only way is up? Employment Review Study on Online Job Portal", International
protection and job mobility among recent Journal of Scientific Research in Computer Science,
entrants to European labour markets,” Engineering and Information Technology
European Sociological Review, vol. 19, pp. 429, (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
2007. 251-255, January-February 2019.
[5] E. Galanki, “The decision to recruit online: a
descriptive study,” Career Development
International, vol. 7, pp. 243-251, 2002.
[6] S. Mauno, U. Kinnunen, and M. Ruokolainen,
“Job demands and resources as antecedents of
work engagement: A longitudinal study,”
Journal of Vocational Behavior, vol. 70, 2007,
pp. 149-171.
[7] A. Doyle, Internet Your Way to a New Job:
How to Really Find a Job Online, Happy about,
2008.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


255
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195158

Voice Based Navigation System for the Blind People


N. Vijayalakshmi 1, K. Kiruthika 2
1 Assistant Professor, Department of Computer Science, Bon Secours College for Women, Bharathidasan
University, Thanjavur, Tamil Nadu, India
2M. Sc Computer Science, Bon Secours College for Women, Bharathidasan University, Thanjavur, Tamil Nadu,
India
ABSTRACT

This main aim of this project to provide voice-based navigation system for blind people using voice recognition
module and it is intended to provide overall measures object detection and real time assistance via Global
Positioning System (GPS) and ultrasonic sensors. This project aims at the development of an Electronic
Travelling Aid (ETA) kit to help the blind people to find obstacle free path. This ETA is fixed to the stick of the
blind people. When the object is detected near to the blinds stick it alerts them with the help of vibratory
circuit. The blind person will give the destination’s name as the input to voice recognition module. GPS module
continuously receives the latitude and longitude of the current location. GPS compares it with the destination’s
latitude and longitude. The blind person receives the pronounced directions which he needs to follow to reach
his destination.
Keywords : Global Positioning System, Electronic Travelling Aid, Navigation System, Latitude and Longitude.

I. INTRODUCTION By the introduction of a basic sensor, the provisions


to the aid of blind people have remarkably increased.
Visual impairment can limit people’s ability to Many universities and companies like IBM have
perform everyday tasks and can affect their quality of given a part of their focus to the development of aids
life and ability to interact with the surrounding for the blind people. Some of the popular ones are
world. Blindness, the most severe form of visual smart canes and obstacle sensors. Recent
impairment, can reduce people’s ability to perform development includes self-driving cars and smart
daily tasks, and move about unaided. Good quality glasses. In the last 30 years, various other strides that
rehabilitation allows people with different degrees of have been developed are the text or speech softwares
visual impairment to fully profit from life, achieve and smartphone apps. Moreover the systems that are
their goals and be active and productive in today’s being developed on robotics and artificial intelligence
society. Ample efforts have been made to aid the will be very advantageous to them as well.
blind by innovating and improving technologies.
According to [23] “The emerging ethics of human One of the major factors in developing these
centric GPS tracking and monitoring”, the main technical aids is the compatibility with the user. He
factors focused while tracking a person are privacy, should not have trouble getting acquainted with the
accuracy and accessibility. product. The features of the product should not be
too difficult to use. Even the notification systems
used to alert the blind should be comfortable and

CSEIT195158 | Received : 01 Feb 2019 | Accepted : 10 Feb 2019 | January-February -2019 [ 5 (1) : 256-259 ]
256
N. Vijayalakshmi, K. Kiruthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 256-259

reliable. Another major factor is the cost of such also in case of an emergency, the location of the blind
products. Since they are already paying for person can be sent to a relative or friend via SMS.
treatments or other nursing cares, the price of the
product should be in the range that is reasonable. II. METHODOLOGY
Other feature of these products should be the
durability. The users might not be able to charge the
system. So, appropriate measure should be taken for
it.

To aid visually impaired and to provide a compact


and complete solution, a smart stick can be developed
which will include a GPS system, obstacle detection
mechanism and an audio module, so that the stick
can instruct the directions to the user. The obstacle
detection mechanism is combined with navigation Figure 1 : Block Diagram
and location detection using GPS-GSM and an audio
module with Bluetooth transceiver to provide a The block diagram of main board is shown in Figure1.
better solution than the already existing solutions. In this diagram the 32-bit ARM processor (LPC2148)
is used, which is the heart of this project. The
The technologies we plan to use are sonar technology, LPC2148 microcontrollers are based on a 16-bit/32-
GPS navigation, GSM network, voice recognition and bit ARM7TDMI-S CPU with real-time emulation and
notification. Sonar technology is used for obstacle embedded trace support, that combine
detection. The principle of the system works in the microcontroller with embedded high speed flash
way that the detector sends and receives a signal. memory ranging from 32 kB to 512 kB. A 128-bit
Upon the reception of the signal, the distance is wide memory interface and unique accelerator
calculated using the time it took for the signal to architecture enable 32-bit code execution at the
reflect back. It is a simple procedure which works maximum clock rate. For critical code size
with the properties of electromagnetic waves. This is applications, the alternative 16-bit Thumb mode
mainly used to find distances in various fields. reduce code by more than 30 % with minimal
Obstacle detection is done by ultrasonic sensors performance penalty.GPS receiver is used to get the
attached to the stick. The sensors are selected to current location in the form of longitude and latitude.
satisfy their criteria of sensor angle, distance and The GPS used here is GR87.The output of GPS
other parameters. When the sensors come across receiver is given to the processor using serial
some obstacle, it sets off a voice announcement from communication. In this system output is in the form
the set of announcements already recorded, of voice hence speaker/headphones are used.
according to their semantics and alerts the person
about the obstacle in the path. In order to reduce the GPS GR87 is a highly integrated smart GPS module
noise and sound dissipation, an earpiece is attached to with GPS patch antenna that is ceramic. The antenna
give the audio announcements. The GPS system is is connected to the module through the LNA. The
used to detect the current location and helps the user module has 51 channel acquisition engine and 14
to reach the destination by announcing the current channel track engine, which is capable of receiving
location name whenever required by the user and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


257
N. Vijayalakshmi, K. Kiruthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 256-259

signals from up to 65 GPS satellites and transferring 3.2 GPS RECEIVER


them into the precise position and timing
information that can be read over either UART port Global Positioning System (GPS) satellites broadcast
or RS232 serial port. Low power consumption is signals from space which are used by GPS receivers,
needed for the small size and high end GPS to provide three-dimensional location (latitude,
functionality. Both the LVTTL-level and RS232 longitude, and altitude) and precise time. Reliable
signal interface are present on the interface positioning, navigation, and timing services are
connector and a supply voltage of 3.6V~6.0V is provided by the GPS receivers to users all around the
supported. The smart GPS antenna module is world continuously in all weather, day or night,
available as an off-the-shelf component which is 100% anywhere on or near the Earth. The GPS receiver
tested. The smart GPS antenna module can be offered used in this project is GR87.Its main features are
for OEM applications. Also, the antenna can be tuned On chip 1Mb SRAM
to the final systems conditions. Low power consumption
Multi path mitigation hardware
III. HARDWARE DESCRIPTION Reacquisition time 0.1 seconds

3.1 MICROCONTROLLER 3.3 VOICE RECOGNITION


A speech analysis is done after the user speaks in a
The Controller used will be ARM LPC2148 which is microphone and inputs are thus taken. The
based on 32/16 bitARM7TDMI-S CPU with real-time manipulation of the input audio signal is done at the
emulation and embedded trace support that combines system level. Different operations are performed at
the microcontroller with embedded high-speed flash different levels on the input signal such as Pre-
memory ranging from 32 kB to 512 kB. It has many emphasis, Framing, Windowing, Mel Cepstrum
important features like. 16/32-bit ARM7 analysis and Recognition (Matching) of the spoken
microcontroller in a tiny package. It has on-chip word. The speech recognition system consists of two
static RAM and on-chip flash program memory. It distinguished phases. The first one is training session,
offers real-time debugging and high-speed tracing of while, the second one is referred to as operation
instruction execution. The features of LPC2148 are session or testing phase. During training phase,
speaker has to provide samples of their speech to
• USB 2.0 Full Speed compliant Device Controller. train the system. During recognition phase, speaker
• 10 bit A/D converters. has to give samples of his speech to match with
• Multiple serial interfaces with two UARTs. existing database and provides exact match. The voice
• Low cost, low consumption, easy handling and recognition IC used here is IC HM2007.
flexibility.
IV. CONCLUSION
These features make the controller reliable for the
project. The world's largest number of blind people and
India’s current population is over 1.22 billion. Earlier
majority of visually impaired people prefer to not use
electronic aids, and use only canes or guide dogs. The
underlying reasons for this include the relatively

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


258
N. Vijayalakshmi, K. Kiruthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 256-259

high costs and relatively poor levels of user sensors. Information and Automation (ICIA),
satisfaction associated with existing electronic 2010 IEEE International Conference on June
systems. So we tried to develop a low cost and user 2010
friendly system for blind people with greatest [7]. Baranski, P.; Polanczyk, M.; Strumillo, P.; A
possible accuracy. This method offers innovative remote guidance system for the blind. e-Health
solutions in order to replace the conventional Networking Applications and Services
methods of guiding visually impaired person. Also, it (Healthcom), 2010 12th IEEE International
can be easily applied anywhere where it can handle Conference .
places like mall, airports etc. In this project we have
used ARM processor which contains more memory Cite this article as :
and its operating speed is high. We guide or navigate
the blind people using voice. N. Vijayalakshmi, K. Kiruthika, "Voice Based
Navigation System for the Blind People",
V. REFERENCES International Journal of Scientific Research in
Computer Science, Engineering and Information
[1]. Amit Kumar,Rusha Patra, M. Manjunatha, J. Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
Mukhopadhyay and A. K. Majumdar An 5 Issue 1, pp. 256-259, January-February 2019.
electronic travel aid for navigation of visually Available at doi :
impaired Communication Systems and https://doi.org/10.32628/CSEIT195158
Networks (COMSNETS), 2011 Third Journal URL : http://ijsrcseit.com/CSEIT195158
International conference on 4-8 jan 2011.
[2]. Shamsi, M.A.; Al-Qutayri, M.; Jeedella, J.; Blind
assistant navigation system Biomedical
Engineering (MECBME), 2011 1st Middle East
Conference on 21-24 Feb. 2011
[3]. Calder, David J.; Curtin .An obstacle signaling
system for the blind ,Digital Ecosystems and
Technologies Conference (DEST), 2011
Proceedings of the 5th IEEE International
Conference on 30 June 2011
[4]. Dunai, L.; Fajarnes, G.P.; Praderas, V.S.; Garcia,
B.D.; Lengua,I.L.; Real-time Assistance
prototype- a new Navigation Aid for Blind
people, IECON 2010 - 36th Annual Conference
on IEEE Industrial Electronics Society
[5]. Shang Wenqin; Jiang Wei; Chu Jian; A machine
vision based navigation system for the blind
China Computer Science and Automation
Engineering (CSAE), 2011 IEEE International
Conference on 10-12 June 2011
[6]. Hashino, S.; Ghurchian, R.; A blind guidance
system for street crossings based on ultrasonic

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


259
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195161

Data Encryption Key Sharing Using Image Pixel Color Value


Vijay Gokul Koli, Raj Kumar Paul
Department of Computer Science & Engineering, Vedica Institute of Technology, Bhopal, India

ABSTRACT

Security is required to transmit confidential information over the network. Security is also demanding in wide
range of applications. Cryptographic algorithms play a vital role in providing the data security against
malicious attacks. Massage communication over internet facing problems like data security, copyright control,
data size capacity, authentication etc. A new idea is to apply reversible data encoding algorithms images by
wishing to remove the embedded data after data receiving with the help of image pixel color value.

Keywords: Network Security, Cryptographic Algorithms, Image Pixel, Data Encryption, Public-Key
Cryptosystem, Encryption, Decryption

I. INTRODUCTION In this technique a public-key cryptosystem (RSA)


using two public key and some mathematical
In the field of networking, role of network security relation. This two public keys are sent
is immense. In the age of information need to keep separately.Security is most important to transmit
information about every aspect of our live. These confidential data over the network, in the today’s
information needs to be hidden from unauthorized world. In wide range of applications, Security is
access (confidentiality), protected from also demanding. For data security Cryptographic
unauthorized change (integrity), and available to an algorithms play a vital role against malicious
authorized entity when it is needed (availability). attacks. In the most popular implementations of
Hence the way of keeping the information securely Public Key Infrastructures, RSA algorithm is
is known as cryptography, which comes from a extensively used. In this paper [1] an algorithm
word with Greek origin, means “secret writing”. has proposed for RSA a method for implementing
Many cryptographic algorithms are developed to a public-key cryptosystem (RSA) using two public
achieve the above said goal. The algorithms should key and some mathematical relation. This two
be such that an opponent cannot defeat its purpose. public keys are sent separately, this makes the
attacker not to get much knowledge about the key
and unable to decrypt the message. Two different
II. METHODS AND MATERIAL keys are used in Public Key cryptography. One key
is used for decryption & only the other
A Modified RSA Encryption Technique Based on corresponding key must be used for encryption.
Multiple public keys Not any other key is possible to decrypt the
message, even the original (i.e. the first) key can’t
used for encryption. Every communicating party

CSEIT195161 | Received : 01 Feb 2019 | Accepted : 10 Feb 2019 | January-February -2019 [ 5 (1) : 260-265 ]
260
Vijay Gokul Koli, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019; 5(1) : 260-265

requires pair of key for communicating with any side, the customer representatives only see the
number of other parties. It is beauty of this names list.
scheme. Once someone obtains a key pair, he can
When operation is needed, software installed on
communicate with anyone else. They have done
the customer representatives’ computer or cell
implementation of RSA algorithm efficiently using
phone were decrypt the data and send them to the
two public key pairs and using some mathematical
call center directly without touching the
logic rather than sending the e value directly as a
representatives.
public key.

Fig 1. RSA algorithm

A new personal information protection approach


based on RSA cryptosystem

With the widespread and rapid development


application of the information technology, the
Fig. 2. The encryption approach of customers’
communication pattern has obviously changed
information
among individuals, corporations and even nations.
However, convenient network-based
Key Distribution by Using Public Key Algorithm
communication method brings not only the benefits
(RSA)
but also some disadvantages such as individual
information leak. This paper [2] introduced that,
personal information can be transformed from plain Authors suggested a new model for quantum key
text into cipher text. Customer representatives can distribution among three parties or more where
be able to contact their clients without seeing the there is a trusted center that providing the
privacy. On the server side, the system necessary secret information of clients to securely
administrator has the permission of authorization communicate to each other.
management. They devolve the authorization to Using the current computing systems classical
database administrators and then database cryptography is based on the computational
administrators input customers’ information into difficulty to compute the secret key. Depending
the system. At the same time, sensitive information only on the difficulty of computational complexity
such as phone number is encrypted. On the client does not provide enough security because finding a
fast method to calculate the secret key. It

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


261
Vijay Gokul Koli, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019; 5(1) : 260-265

compromises the security of the systems. Law of In RSA computation is lengthy and some less
physics is used in Quantum computing for secure. This paper[5] present a new algorithm to
communication. In cryptography and key presents the modified form of new RSA algorithm
distribution quantum theorems and principles are in order to boost up the speed of the
applied. In this paper[3], new model for quantum implementation of RSA algorithm during data
key distribution are introducing among three exchange across the network world. In this method
parties or more where there is a trusted center that keys are stored offline before the process start.
providing the necessary secret information of Thus, the speed of process increased as compared
clients to securely communicate to each other. to original RSA method.

To compare the bases, classical channel is used by Enhancing the Security of The RSA Cryptosystem
quantum key distribution protocols BB84, B92 and
ERP.
This paper [6] increases the security of the RSA
Loop-based RSA Key Generation Algorithm using algorithm, this enhancement use randomized
String Identity parameter to change every encrypted message
block such that even if the same message is sent
more than once the encrypted message block is look
This paper [4] propose i-RSA algorithm that is focus
different. This paper suggests that how to use
on key generation algorithm. user identity is
randomized parameters in the encrypt the data to
Enhancement of this algorithm. it can be used as a
make RSA. By this enhancement it makes the RSA
public key, such as email address. The key
semantically more secure. Means an attacker cannot
certificates are used to authenticate the user's key
distinguish two encryptions from each other, even
pair. So certificate does work as important role in
if the attacker knows (or has chosen) the
secure communication but to issue the certificate is
corresponding plaintexts (original message). In this
a big challenge and it also increases the overhead
Work comparison between the modified RSA and
due to the increasing cost. For public key Previous
the basic RSA version introduced. Enhancement
algorithm has successful used email identity, but all
can easily be implemented on this Work. Also other
type of email can’t be used as a public key. So the
attacks are presented by this paper, also how to
propose i-RSA algorithm that can produces 66.6%
speed up the RSA encryption and decryption
compared to previous algorithm (46.67%) email can
process is an important issue for the RSA
be a string public key. in key generation looping
implementation.
process is the main differences between i- RSA and
previous algorithm, to get new value of p and q Here, RSA is more secure and it may be more
parameter, when value of k is equal to 1, then stronger by applying some techniques. Here They
looping process can stop, and the email can be a have seen that all authors are talking about many
public key. Detail explanations of i-RSA algorithm methods but no one is talking about image pixel for
in propose algorithm section. security purpose. So theyF can add image pixel
technique to make more powerful RSA

algorithm.
Modified RSA Cryptosystem Based on Offline
Storage and Prime Number

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


262
Vijay Gokul Koli, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019; 5(1) : 260-265

Proposed Methodology Algorithm


In this proposed work first of all p and q two prime For both encryption and decryption
number is selected then find p*q after this calculate algorithms are following.
(p-1)*(q-1). Then select e and

after all public key and private key is generated. But


before sharing public key they have to encrypt this.
So that other person who doesn’t belongs to my Encryption
group cannot get public key. (1) Select p and q both prime number, p is not
equal to q.
According to the comparisons and the characteristics (2) Calculate n = p x q.
of RSA, it determined to use RSA cryptography as the (3) Calculate ø (n) = (p -1) x (q-1).
core algorithm for personal information protection in (4) Select integer e whose gcd (Ø (n), e) = 1; 1 < e
information system. It makes users don’t have to <Ø (n).
store a mass of calculated secret keys. The (5) Calculate private key d = ℮-1 (mod Ø (n)).
information owner can easily send messages to the (6) Public key PU = {e, n}.
receiver when he got reliable public key from the (7) Private Key PR = {d, n}.
receiver. This approach makes things easy, only one (8) Im=Load any image img.
pair of keys is necessary. (9) Select pixel position p_id.
(10) For(i=1; i<=P_id; i++)

Encryption {

If(i==P_id)
After generating public key, before sharing to other Px=im(p_id);
people’s sender can be use an image that already
}
have in receiver side. First of all select an image that
already have all receivers then select any pixel color (11) E=Px+e.
value of that image. Then add that color value with (12) Public key to transmit = {E,P_id,n}.
e and save in E. now it doesn’t need to share actual (13) Message (M) Cipher text- C = M^e mod n.
value of e. now it share E and position of that pixel Decryption
in image.

(1) Public key to transmit = {E,P_id,n}. (2)


Decryption For(i=1; i<=P_id; i++)
When receiver get encrypted message then if he {
have same image then they can be select same image
If(i==P_id)
and select pixel position that they have received and
Px=im(p_id);
after all, pick color value of that pixel and minus
that value from E that they have received from }
sender. Means anyone who wants to share public
(3) e=E-Px.
key they only share E and pixel position.
(4) Message M = C^e mod n.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


263
Vijay Gokul Koli, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019; 5(1) : 260-265

Where M is message (Plane text), p and q are prime


numbers, N is common modulus, e and d are public
and private keys, p_id is a pixel position of selected
image, im is a program variable which contain all
pixels color values of selected image, Px is a color
value of pixel position p_id for selected image.

III. RESULT AND ANALYSIS

Figs shows front view of implementation of proposed


work. It has designed using MATLAB. Here sender
side “select image” button used for selecting the
image. Fist text box is using for entering message that
you want to send, second text box is using for giving
pixel position value. And in receiver side first text Fig. 5. Histogram of receiver side image
box is using for giving same pixel position that has
Here, in figs, clearly seen that histogram of sender
given in sender side.
side image is same as histogram of receiver side image
Time Required for Brute force attack because this algorithm don’t modifying image. only
80
data encryption using encrypted image
data encoding using pixel value
pixel color value is using for reference.
70

60
IV. CONCLUSION
Time in Seconds

50

40

30 In this work a new method has proposed for RSA


20 public key sharing. In this method before sharing the
10 public key, e is encrypted with any specified pixel
0
0 50 100 150 200 250 300 350 400 450 500
color value of any particular image. So, it doesn’t
Traffic
need to transfer e. because e is encrypted with any
pixel color value then possibility of attack is very less
Fig. 3. Comparison of brute force attack in existing
as compared to without encrypted that’s why if any
and proposed method using image pixel.
attacker got shared key then they don’t know what is
actual value of e. it is possible if and only if he has
same image. But in this method no one is sharing
image. image is predefined. So after all this method is
complex for attacker to getting.

V. REFERENCES

[1]. Amare Anagaw Ayele, Dr. Vuda Sreenivasarao,


Fig. 4. Histogram of sender side image June 2013, "A Modified RSA Encryption
Technique Based on Multiple public keys",
International Journal of Innovative Research in

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


264
Vijay Gokul Koli, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019; 5(1) : 260-265

Computer and Communication Engineering Cite this article as :


Vol. 1, Issue 4.
[2]. Liang Wang, Yonggui Zhang, 2011, "A New Vijay Gokul Koli, Raj Kumar Paul, "Data Encryption
Personal Information Protection Approach Key Sharing Using Image Pixel Color Value",
Based on RSA Cryptography", IEEE. International Journal of Scientific Research in
[3]. Ammar Odeh, Khaled Elleithy, Muneer Computer Science, Engineering and Information
Alshowkan, Eman Abdelfattah, 2013, Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
"Quantum Key Distribution by Using Public 5 Issue 1, pp. 260-265, January-February 2019.
Key Algorithm(RSA)", IEEE. Available at doi :
[4]. Norhidayah Muhammadi, Jasni Mohamad https://doi.org/10.32628/CSEIT195161
Zaini, Md Yazid Mohd Saman, "Loop-based Journal URL : http://ijsrcseit.com/CSEIT195161
RSA Key Generation Algorithm using String
Identity", 13th International Conference on
Control, Automation and Systems (ICCAS
2013).
[5]. Ms. Ritu Patidar, Mrs. Rupali Bhartiya, 2013,
"Modified RSA Cryptosystem Based on Offline
Storage and Prime Number", IEEE.
[6]. Malek Jakob Kakish, "Enhancing The Security
Of The Rsa Cryptosystem", Ijrras August 2011.
[7]. M. Jason Hinek, Another Look at Small RSA
Exponents, 2006.
[8]. B. A. Forouzan, D. Mukhopadhyay,
Cryptography and Network Security, Tata
McGraw-Hill, 2012.
[9]. Abdullah Darwish Imad Khaled Salah and Saleh
Oqeili, Mathematical Attacks on RSA
Cryptosystem, Journal of Computer Science
(2006).
[10]. J. M. Pollard, A Monte Carlo Method for
Factorization, BIT Numerical Mathematics
(1975).
[11]. H. Riesel, Prime Numbers and Computer
Methods for Factorization, Birkhauser, 1994.
[12]. William Stein, Elementary number theory.
Primes, congruences, and secrets. A
computational approach., New York, NY:
Springer, 2009 (English).

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


265
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195179

A Review on Prediction of Diabetic Retinopathy Using Data


Mining Algorithms
Kajal Sanjay Kothare1, Prof. Kalpana Malpe2
1 M-Tech Student Department of Computer Science and Engineering, Guru Nanak Institute of Engineering &
Technology, Nagpur, Maharashtra, India
2 Assistant Professor Department of Computer Science and Technology, Guru Nanak Institute of Engineering &
Technology, Nagpur, Maharashtra, India

ABSTRACT

The risking components of diabetic retinopathy (DR) were examined broadly in the past investigations, yet it
stays obscure which chance variables were more connected with the DR than others. On the off chance that we
can distinguish the DR related hazard factors all the more precisely, we would then be able to practice early
avoidance systems for diabetic retinopathy in the most high-chance populace. The motivation behind this
examination to study and consider the different predicting mechanisms for the DR in diabetes mellitus utilizing
data mining techniques including the support vector machines, decision trees, artificial neural networks, and
logistic regressions.
Keywords : Data Mining, Artificial neural fuzzy interference system, K-Nearest-Neighbor (KNN), Machine
Learning (ML), Support Vector Machines, Decision Trees

I. INTRODUCTION Diabetic retinopathy is microvascular


inconvenience of diabetes mellitus. It wards on
Diabetic retinopathy (DR) is the most well-known history of diabetes-related difficulties [2]. Diabetic
reason for recently analyzed visual deficiency retinopathy is a profoundly explicit vascular
consistently, particularly in working-age populace. confusion of both kind 1 and type 2 diabetes, with
Review surveys of the United Kingdom Prospective predominance emphatically identified with the
Diabetes Study (UKPDS) and the Diabetes Control length of diabetes [2]. Diabetic retinopathy comprises
and Complications Trial (DCCT) on sort 2 and type 1 of non-proliferative diabetic retinopathy (NPDR) and
diabetes mellitus, separately, both recommended that proliferative diabetic retinopathy (PDR). NPDR is
serious blood glycemic control can adequately lessen otherwise called foundation diabetic retinopathy
the danger of microvascular inconveniences (BDR). NPDR is beginning period of DR, and PDR is
including diabetic retinopathy. Past examinations for late phase of DR. In NPDR status, micro aneurysm,
the most part centered around controlling one discharge, hard exudates, cotton fleece spot,
noteworthy hazard factor which is glucose level, intraregional microvascular variations from the norm,
though few investigations have concentrated on and venous beading are common characters. In PDR
researching diverse hazard variables of DR [1]. organize, there are plate neovascularization, vitreous
discharge, and sinewy scarring. Macular edema is
testimony of hard exudates close macula. Diabetic

CSEIT195179 | Received : 01 Feb 2019 | Accepted : 10 Feb 2019 | January-February -2019 [ 5 (1) : 266-272 ]
266
Kajal Sanjay Kothare, Prof. Kalpana Malpe Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 266-272

retinopathy is the most regular reason for new vision with NPDR. However, when they created
instances of visual deficiency among grown-ups PDR with vitreous drain, they lost the vision all of a
matured 20– 74 years [2]. Around 21% of the sudden. Prior to vision misfortune, photocoagulation
recently determined patients to have type 2 diabetes can evade drain of PDR and diminished vision
(T2D) were likewise found to have co-bleak state of misfortune. In spite of the fact that instruction of
DR, though 60% of the patients with a ceaseless intricacy of T2D in our consideration unit is normal
history of 20 years of T2D were determined to have routine training for patients with T2D, the absence of
diabetic retinopathy [3]. About 20– 40% of patients knowledge is by all accounts the significant reason.
in T2D had diabetic retinopathy and 8% of patients
in T2D had locate compromising diabetic retinopathy One of the significant hazard factors inspected in a
(STDR) in United States [4]. In 2009, Prevalence of pooled examination from populace based
the diabetic retinopathy and poor vision/visual investigations around the globe was the long span of
impairment in Taiwanese patients with T2D were diabetes [8]. Other hazard factors distinguished in
8.91 and 0.62%, separately [5]. this investigation were abnormal state haemoglobin
A1C (HbA1C) and hypertension [8]. As indicated by
There are a few medicines for DR., To begin with, the UKPDS, the rate of diabetic retinopathy is
it is critical to immediately allude patients with any intently connected with the expanding length of T2D,
dimension of macular edema, extreme NPDR, or any and lower dimension of HbA1c can diminish the
PDR to an ophthalmologist who is proficient and danger of experiencing DR in these patients [9]. In
experienced in the administration and treatment of any case, it was seen in clinical practice that a few
diabetic retinopathy [2]. What's more, laser patients with long haul controlled HbA1c levels still
photocoagulation ought to be considered for eyes have dangers experiencing diabetic retinopathy in
with clinically critical macular edema, especially T2D [10]. This proposes the HbA1C level isn't the
when the focal point of the macula is included or main significant hazard factor, and different factors,
unavoidably undermined [6]. Hostile to vascular for example, hypertension, high blood glucose, and
endothelial development factor (against VEGF) length of diabetes may have conceivably played
treatment is additionally demonstrated for diabetic incomplete jobs in the improvement of diabetic
macular edema [2]. retinopathy in T2D. There were realized hazard
components of diabetic retinopathy, for example,
To limit the chances of visual misfortune or the long span of diabetes, poor glycaemic control,
new beginning of visual deficiency of diabetic hypertension, and hyperlipidaemia.
retinopathy, current rules of Taiwan diabetic
association proposes that the screening of fundus In the end, the most well-known hazard factor found
examination in patients with T2D should be by each bit of observational the proof is the term of
performed every year, and performed all the more diabetes. Poor glycaemic control drives a high fasting
much of the time in patients with diabetic glucose level, high, postprandial glucose, and high
retinopathy. It is noticed that the screening rate was HbA1C. Every one of the three previously mentioned
low, for there was just 28.9% of patients with T2D parameters advise diverse parts of diabetes and should
had eye fundus examination in Taiwan in 2009 [7]. all be considered.
One conceivable clarification for the low screening
rate might be that patients with T2D couldn't care
less about retinopathy when they have an ordinary

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


267
Kajal Sanjay Kothare, Prof. Kalpana Malpe Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 266-272

II. RELATED WORK 0.683, which means a decent dimension of


concordance between the watched event of DR and
A few examinations [11– 16] have been created to that anticipated by the model. For the outer approval,
anticipate diabetic retinopathy. A cross-sectional the qualities for C-record and CPE were 0.767 and
investigation of patients with T2D utilized routinely 0.697, separately. The AUC for 1-year survival from
gathered information at outpatient centers of the retinopathy was 0.825. There was no factual
Isfahan Endocrinology and Metabolism Research distinction between the C-record of that determined
Focus (IEMRC), Iran [11]. This investigation in the train informational index versus that
connected collector working trademark (ROC) bends determined on the test informational index (p =
to recognize the ideal estimation of diabetic patients 0.137). At that point, the investigation utilizes the
for deciding DR; affectability and explicitness for classification and relapse tree (CART) examination or
anticipating DR were determined for various cuts of the arbitrary woodland investigation for the train
the score. This examination exhibited the informational collection to confirm how the
consequences of utilizing calculated relapse models outcomes were steady with these distinctive
with DR as the needy variable. The region under the methodologies.
ROC bend (AUC) was 0.704, and furthermore
indicated affectability (60%) and explicitness (69%) Dr. Karim Hashim Al-Saedi et.al.[14] have
of a hazard score ≥ 52.5 for DR. structured and executed a framework to quantify the
effect of diabetic retinopathy utilizing information
Another examination talked about individual mining procedures In this exploration, an exact
hazard evaluation and data innovation to screen the estimation arrangement of diabetic retinopathy was
recurrence of diabetic retinopathy [12]. This created and researched utilizing information mining
investigation utilized a scientific calculation made method in which an early investigation of diabetic
utilizing epidemiological information on hazard retinopathy utilizing a precise and quick system gives
factors for diabetic retinopathy, through a site, the patient enough security treatment time. The
http://risk.is/, in which the calculation gets clinical shading fundus picture was utilized to consequently
information, including type and length of diabetes, identify and understand the different sores of diabetic
HbA1c or mean blood glucose, circulatory strain and retinopathy and its ordinary highlights, separately.
the nearness and grade of retinopathy. The AUC was The determinations of the typical shading fundus
0.76, and this number shows the model predicts the pictures were dissected and arranged by the
likelihood of a patient who creates locate extraction technique into ordinary or strange. The
compromising retinopathy (STR) 76% more right anomalous picture would then be able to be ordered
than who does not create STR. into three dimensions: Mild, moderate, and Severe.
To foresee the obscure class, an association standard
An investigation by Semeraro et al. anticipated and SVM classifier was utilized. The creator indicated
danger of diabetic retinopathy utilizing the c- promising outcomes to help the patients and the
measurement, survival beneficiary working quickening procedure.
trademark, and the Gonen and Heller concordance
likelihood gauge (CPE) for the Cox corresponding Abhilash Bhaisare et. al. [15] have proposed a
peril demonstrate [13]. For the inward approval, the framework to recognize the info picture as ordinary
C-file achieved an estimation of 0.746; the Gonen– or anomalous. At the point when the information
Heller CPE for the Cox relative risk technique was picture is discovered strange then examination for

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


268
Kajal Sanjay Kothare, Prof. Kalpana Malpe Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 266-272

further DR stages is finished. To recognize irregular of multi-display medoid based demonstrating


pictures there are different procedures and strategy approach in a gathering is exhibited to enhance the
utilized in picture mining. Picture mining is an precision of classification. The genuine
augmentation of information mining system. Microaneurysm districts are picked and characterized
Distinguishing proof of the unusual pictures is utilizing a hybrid classifier which is a weighted mix
finished utilizing preprocessing, including extraction of multivariate m-Mediods, GMM and SVM.
and classification calculations. The acquired outcome
is utilized to show the picture as ordinary or strange Vimala Balakrishnan et.al [18] have dealt with
and up to what degree. Integrating association guidelines and case-based
thinking to foresee retinopathy. They have proposed
K. R. Ananthapadmanaban and G. Parthiban [16] a retinopathy expectation framework dependent on
have utilized Naive Bayes and Support Vector information mining, especially association rules
Machine calculations to anticipate the early discovery utilizing Apriori calculation, and case-based thinking.
of eye ailment and DR. Utilizing Rapid Miner device The association rules are utilized to break down
they have assessed that Naive Bayes gives 83.37% examples in the informational index and to ascertain
precision and SVM gives 64.91% exactness. Execution retinopathy likelihood while case-based thinking is
of these strategies was additionally estimated by utilized to recover comparable cases. The creator
particularity as 95% and affectability as 96.65%. They trusted that extraordinary upgrades can be given to
began with a preprocessing task to enhance picture restorative professionals and furthermore to diabetics
quality by killing imperfections brought about by with the usage of this framework.
lighting and procurement forms. In the second step,
the optic plate has upset the programmed recognition. M.Tamilarasi and Dr.K.Duraiswamy [19] have
In the third step, the division of chart slices is utilized played out an overview on Automatic recognition of
so as to identify exudates areas. At long last, the non proliferative DR. Since there are numerous
neural system gave better outcomes with a information mining strategies and calculations that
component extraction of pictures by descriptors and analyze DR in retinal fundus pictures, in this paper,
Hu snapshot of GIST. The last outcomes were creators have checked on, characterized and looked
contrasted quantitatively and a manual exudates at the calculations and methods recently proposed so
division created by a specialist in ophthalmology. as to grow better and increasingly powerful
calculations.
M. Usman Akram, Shehzad Khalid, Shoab A.Khan
[17] have introduced a paper indicating three-arrange Ramon Casanova et.al.[20] concentrated on the
framework for early discovery of Microaneurysm utilization of Random Forest (RF) techniques to DR
utilizing channel banks. The framework extricates all classification investigation dependent on fundus
conceivable applicant districts for Microaneurysm photography information. The creators proposed a
present in the retinal picture. Contingent on a few way to deal with DR hazard evaluation dependent on
properties, similar to shape, shading, power, and metric got from reviewed fundus photography and
measurements an element vector for every area is fundamental information. they proposed that RF
framed to arrange a hopeful district as could be a profitable instrument to analyze DR
Microaneurysm or non-Microaneurysm. A crossover finding and assess its movement.
classifier which consolidates the Gaussian blend show
(GMM), bolster vector machine (SVM) and expansion

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


269
Kajal Sanjay Kothare, Prof. Kalpana Malpe Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 266-272

S.Sagar Imambi and T.Sudha [21] have connected Calculated relapse estimates the connection
content mining classification systems on the between the absolute ward variable and at least one
accessible clinical information to anticipate the autonomous factors by assessing probabilities. The
hazard factor of DR. They demonstrated that first expect a calculated capacity and the second a
moderately brief length of the case the executives standard typical dissemination work. The chances of
initiated before the beginning of clinically the reliant variable measuring up to a case are
recognizable retinopathy, essentially diminish the proportional to the exponential capacity of the direct
danger of creating retinopathy in patients with sort 2 relapse articulation. This shows how the rationale
diabetes. The objective was to build up an adaptable fills in as a connection work between the likelihood
and hearty clinical report classification framework and the direct relapse articulation. We additionally
that could be connected in extensive doctor's facility joined stepwise choice to choose discriminative
settings to encourage the doctors, with the goal that highlights in strategic relapse.
they can direct the patients effectively and diminish
the vision misfortune. SVM classifier is a machine learning calculation
proposed by Vapnik dependent on basic hazard
Hayrettin Evirgen, Menduh Çerkezi [22], have minimization rule of insights learning hypothesis. It
utilized Naive Bayes classification calculation to very well may be utilized to take care of classification
examine the genuine dataset so as to assembled and relapse issues. As a forecast of diabetic
prescient framework for DR. A sum of 385 diabetes retinopathy is a paired classification issue, SVM
patients' information was utilized to prepare the would be helpful for our motivation. During the time
expectation framework. With cross-approval creators spent model improvement, we utilize outspread
demonstrated that gullible Bayes calculation can be premise work (RBF) as the bit work in SVM.
utilized for diabetic retinopathy forecast with an
enhanced precision of 89%. An artificial neural Network is a group of
measurable learning models propelled by organic
III. MACHINE LEARNING ALGORITHMS TO neural systems and are utilized to evaluate or inexact
PREDICT DIABETIC RETINOPATHY capacities that can rely on countless and are
commonly obscure. ANNs are for the most part
DR expectation can be viewed as a multi-class exhibited as frameworks of interconnected neurons
classification issue. We consolidated choice trees, which send messages to one another. Positioned
calculated relapse, fake neural systems, and bolster forecast execution of choice trees dependent on 2-
vector machines to anticipate DR. The SAS year, 10-year, and a 15-year term of diabetes be
Enterprise Miner adaptation 12.1 programming of tuned dependent on experience, making neural nets
was utilized to produce expectation models. versatile to sources of info and fit for learning.

To help basic leadership forms, we utilized choice IV. CONCLUSIONS


trees to create interpretable guidelines for clinical
practice. We build choice trees and produce rules for Forecast and conclusion of Diabetic Retinopathy
clinical basic leadership, arranging dependent on rely upon the nearness of hemorrhages and
information gathering and unmitigated examination, microaneurysms in fundus pictures. There are
and producing choice trees as foreseeing models to numerous calculations which have been proposed
help with clinical basic leadership. and produced for the programmed recognition of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


270
Kajal Sanjay Kothare, Prof. Kalpana Malpe Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 266-272

diabetic retinopathy from highlight extraction. In angiographic and other retinal characteristics at
this paper condensed perspective of different baseline: ETDRS report no. 19", Early treatment
information mining methods is displayed which diabetic retinopathy study research group. Arch
demonstrates that KNN and SVM have given the best Ophthalmol. 1995; 113(9):1144–55.
correctnesses. This survey paper can go about as an [7] Chang TJ, Jiang YD, Chang CH, Chung CH, Yu
asset for the future scientists for the forecast of NC, Chuang LM. Accountability, utilization and
diabetic retinopathy utilizing information mining providers for diabetes management in Taiwan,
procedures. This will likewise be helpful for scientists 2000–2009: an analysis of the National Health
to get a diagram of this zone so as to grow Insurance database. J Formos Med Assoc.
increasingly proficient calculations. 2012;111(11):605–16. Tsao et al. BMC
Bioinformatics 2018, 19(Suppl 9):283 Page 120 of
V. REFERENCES 121
[8] Yau JW, Rogers SL, Kawasaki R, Lamoureux EL,
[1] Chew EY, Klein ML, Ferris FL 3rd, Remaley NA, Kowalski JW, Bek T, Chen SJ, Dekker JM,
Murphy RP, Chantry K, Hoogwerf BJ, Miller D., Fletcher A, Grauslund J, et al. Global prevalence
"Association of elevated serum lipid levels with and major risk factors of diabetic retinopathy.
retinal hard exudate in diabetic retinopathy. Diabetes Care. 2012; 35(3):556–64.
Early treatment diabetic retinopathy study [9] UK Pospective Diabetes Study Group. Tight
(ETDRS) report", 22. Arch Ophthalmol. 2016; blood pressure control and risk of macrovascular
114(9):1079–84. and microvascular complications in type 2
[2] American Diabetes Association. Standards of diabetes: UKPDS 38. UK Prospective Diabetes
medical care in diabetes–2014. Diabetes Care. Study Group. BMJ. 1998; 317(7160):703–13.
2014;37(Suppl 1):S14–80. [10] Kowall B, Rathmann W. HbA for diagnosis of
[3] Fong DS, Aiello L, Gardner TW, King GL, type 2 diabetes. Is there an optimal cut point to
Blankenship G, Cavallerano JD, Ferris FL 3rd, assess high risk of diabetes complications, and
Klein R, "American Diabetes A. Retinopathy in how well does the 6.5% cut-off perform?
diabetes. Diabetes Care", 2004;27(Suppl 1):S84–7. Diabetes Metab Syndr Obes. 2013; 6:477–91.
[4] Kempen JH, O'Colmain BJ, Leske MC, Haffner [11] Hosseini SM, Maracy MR, Amini M, Baradaran
SM, Klein R, Moss SE, Taylor HR, Hamman RF. HR. A risk score development for diabetic
"The prevalence of diabetic retinopathy among retinopathy screening in Isfahan-Iran. J Res Med
adults in the United States", Arch Ophthalmol. Sci. 2009;14(2):105–10.
2004; 122(4):552–63. [12] Aspelund T, Thornorisdottir O, Olafsdottir E,
[5] Huang YY, Lin KD, Jiang YD, Chang CH, Chung Gudmundsdottir A, Einarsdottir AB, Mehlsen J,
CH, Chuang LM, Tai TY, Ho LT, Shin SJ. Einarsson S, Palsson O, Einarsson G, Bek T, et al.
"Diabetes-related kidney, eye, and foot disease in Individual risk assessment and information
Taiwan: an analysis of the nation wide data for technology to optimise screening frequency for
2000-2009", J Formos Med Assoc. 2012; diabetic retinopathy. Diabetologia.
111(11):637–44. 2011;54(10):2525–32.
[6] Early Treatment Diabetic Retinopathy Study [13] Semeraro F, Parrinello G, Cancarini A, Pasquini
Research Group. "Focal Photocoagulation L, Zarra E, Cimino A, Cancarini G, Valentini U,
treatment of diabetic macular edema. Costagliola C. Predicting the risk of diabetic
Relationship of treatment effect to fluorescein

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


271
Kajal Sanjay Kothare, Prof. Kalpana Malpe Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 266-272

retinopathy in type 2 diabetic patients. J [21] S. Sagar Imambi and T. Sudha - Building
Diabetes Complicat. 2011;25(5):292–7. Classification System to Predict Risk factors of
[14] Dr. Karim Hashim Al-Saedi, Dr. Razi Jabur Al- Diabetic Retinopathy Using Text mining, (IJCSE)
Azawi, Rasha Asaad Kamil, - Design and International Journal on Computer Science and
Implementation System to Measure the Impact Engineering Vol. 02, No. 07, 2010, 2309-2312
of Diabetic Retinopathy Using Data Mining [22] Hayrettin Evirgen, Menduh Çerkezi- Prediction
Techniques, International Journal of Innovative and Diagnosis of Diabetic Retinopathy using
Research in Electronics and Communications Data Mining
(IJIREC) Volume 4, Issue 1, 2017, PP 1-6 Technique,ijaedu.ocerintjournals.org/tojsat/issue
[15] Abhilash Bhaisare, Sagar Lachure, Amol Bhagat, /22642/241874, Sep, 2014
Jaykumar Lachure - Diabetic Retinopathy
Diagnosis Using Image Mining, International Cite this article as :
Research Journal of Engineering and Technology
(IRJET), Volume: 03, Issue: 10, Oct -2016 Kajal Sanjay Kothare, Prof. Kalpana Malpe, "A
[16] K. R. Ananthapadmanaban and G. Parthiban. - Review on Prediction of Diabetic Retinopathy Using
Prediction of Chances - Diabetic Retinopathy Data Mining Algorithms", International Journal of
using Data Mining Classification Techniques. Scientific Research in Computer Science, Engineering
Indian Journal of Science and Technology, Vol and Information Technology (IJSRCSEIT), ISSN :
7(10), 1498–1503, October 2014 2456-3307, Volume 5 Issue 1, pp. 266-272, January-
[17] M. Usman Akram, Shehzad Khalid, Shoab A. February 2019.
Khan - Identification and classification of Journal URL : http://ijsrcseit.com/CSEIT195179
microaneurysms for early detection of diabetic
retinopathy, Pattern Recognition, Vol. 46, No. 1,
2013, pp. 107–116.
[18] Vimala Balakrishnan , Mohammad R. Shakouri
and Hooman Hoodeh - Integrating association
rules and case-based reasoning to predict
retinopathy, Maejo Int. J. Sci. Technol. 2012,
6(03), 334-343, ISSN 1905-7873
[19] M. Tamilarasi and Dr. K. Duraiswamy -A Survey
for Automatic Detection of Non- Proliferative
Diabetic Retinopathy, International Journal of
Innovative Research in Computer and
Communication Engineering, Vol. 2, Issue 1,
January 2014
[20] Ramon Casanova , Santiago Saldana, Emily Y.
Chew, Ronald P. Danis, Craig M. Greven,
Walter T. Ambrosius - Application of Random
Forests Methods to Diabetic Retinopathy
Classification Analyses, Published: June 18,
2014https://doi.org/10.1371/journal.pone.009858
7

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


272
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT195180

A Review on Predicting User Behaviour through Session Using


Web Mining
Veena Moharle1, Sonali Waghade1, Divyani Dhoke1, Sheetal Kolhe1, Snehal Narule1, Prof. Merajul Haque2
1BE Students, Department of Information Technology, JD College of Engineering and Management, Nagpur,
Maharashtra, India
2Assistant Professor, Department of Information Technology, JD College of Engineering and Management,
Nagpur, Maharashtra, India

ABSTRACT

Web usage mining is driving examination zone in Web Mining worried about the web client's behavior.
Weblog mining is one of the ongoing regions of research in Data mining. Web Usage Mining turns into an
imperative perspective in the present time because the amount of information is persistently expanding. We
manage the web server logs which keep up the historical backdrop of page demands. Web log records are the
documents, which contain total data about the clients, peruse exercises on the web server Web mining is the
utilization of information mining strategies to find designs from the World Wide Web. This paper gives a
consideration on Web usage mining to anticipate the behavior of web clients dependent on web server log
records. Clients utilizing web pages, a successive access way's and continuous access pages, joins are put away in
web server log records. Contingent on the recurrence of clients visiting each page mining is performed. By
finding the session of the client we can break down the client behavior when spending on a specific page. Web
log alongside the distinction of the client catches their perusing behavior on a website and talking about with
respect to the behavior from the investigation of various algorithms and diverse strategies.
Keywords : Knowledge Sharing, Web Portal, Job Portal, Online Recruitment.

I. INTRODUCTION between web pages through the web structure. Web


structure mining is again classified into hyper
Web Usage mining applies the information mining interface mining and inside structure mining. Web
method to extricate learning from weblog records usage mining examinations weblog records for
naturally. Web mining can be arranged into web discovering program examples of the client. Web
content mining, web structure mining and web usage usage mining is additionally called weblog mining.
mining. Web content mining is the way toward
separating data from an expansive measure of Understanding client get to examples can help a
gathered information. This strategy extricates the webmaster in tweaking the website content so as to
data from the substance of web pages. Web content expand the client's perusing knowledge. Web route
mining is again arranged into interactive media example might be portrayed by breaking down web
mining and web printed mining. Web structure server logs that contain nitty-gritty data of client
mining is the way toward dissecting the connections perusing exercises. An occasion in a weblog is made

CSEIT195180 | Received : 01 Feb 2019 | Accepted : 10 Feb 2019 | January-February -2019 [ 5 (1) : 273-277 ]
273
Veena Moharle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 273-277

out of IP address of the asking for customer, date and algorithms for playing out every one of these
time of the demand, asking for technique utilized, procedures [1].
status code, the asked for the record, and so forth. It
is conceivable to sort the arrangement of snap Anshul Bhargav and Munish Bhargav likewise the
exercises into sessions, and after that utilization them individuals from IEEE proposed edge work for web
for further investigation to get the predicted client usage mining. The edge work is to perform client’s
behaviour. Breaking down clients' Web watching characterization dependent on found examples and to
behaviours is one of the essential and testing research discover the qualities of clients. The casing work
subjects of Web usage mining. On the off chance that comprises of three principal steps: pre-processing,
clients' advantages can be naturally recognized from design disclosure and client arrangement. The pre-
clients' Web log information, they can be utilized for processing stage incorporates log document cleaning,
data suggestion and showcasing which are valuable client ID and session recognizable proof. The
for the two clients and Web webpage engineers. example disclosure process finds a new example from
Research for dissecting Web log information has the informational collection. In the client
been finished by numerous specialists in the field of arrangement process, every client is grouped by his
Web usage mining; finding incessant examples of log or her attributes. The groupings depend on nation-
information. The principle motivation behind the based arrangement, site section based order and
investigation of the client's visit record is to break access time-based characterization [2].
down the client's most worried about the outcomes
from the mining results. By breaking down the client S. S. Patil and H.P Khandagale proposed subjective
access to assets of the time, recurrence, etc., alter the client demonstrate which determine the foreseen
structure and plan of the site to anticipate that more usage behaviour dependent on the examples found in
clients should remain and better serve clients. Client past pre-processing, stage. The edge work
behaviour investigation has turned into another incorporates information cleaning, client
examination hotspot. recognizable proof, session distinguishing proof, way
finish, exchange ID and example revelation and
The work of this research mainly contemplates the extraction forms. Way culmination alludes to
weblog Mining technology in client behaviour creating rules for missing references dependent
investigation and assembles the client premium available structure, referrer, and other heuristic data.
model dependent on the client premium data, lastly By breaking down the found example, the behavior
draws the client's advantage. of the client is recognized and route refreshes are
given in the web page [3].
II. LITERATURE REVIEW
Virendra R. Rathod and Govind V. Patel proposed an
G. Neelima and Dr. Sireesha Rodda individuals from edge work dependent on FCM grouping and Markov
IEEE have proposed a thought for anticipating client show for foreseeing client behaviour utilizing weblog.
behavior through sessions utilizing weblog mining. Subsequent to pre-processing fluffy c implies (FCM)
They manage web server logs of NCSA normal log calculation is connected for example disclosure and
record design for mining. The proposed methodology examination. Markov demonstrate is utilized to next
comprises of information handling, client page expectation and better web page forecast
distinguishing proof and session recognizable proof precision. Fluffy c implies bunching calculation gives
procedures. They are utilizing three unique the preferred outcome over k-implies grouping.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


274
Veena Moharle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 273-277

Fluffy c implies bunching is a standout amongst the client's advantage dependent on the M5 display tree
most generally utilized fluffy grouping algorithms [4]. to break down the explanatory occasions.

In this exploration [5] propose the structure and In this examination [9], we present the "errand trail"
usage of a half and half framework by consolidating a to comprehend client seek behaviours. We
few information mining methods to catch the client's characterize an errand to be a nuclear client data
web perusing behavior. Client route sessions that need, while an assignment trail speaks to all client
speak to the association with a given website are exercises inside that specific undertaking, for
utilized to develop Hypertext Probability Grammar example, question reformulations, URL clicks. he
(HPG). Break down the HPG with the idea chain of considers investigations and correlations with assess
command show and apply the network bunching the viability of undertaking trails in a few inquiry
procedure to separate the usual portrayal of client applications: determining client fulfillment,
route behavior. anticipating client seek premiums, and
recommending related questions. Analyses on
This paper [6] portrays a technique for clearing up expansive scale datasets of a business internet
clients' interests dependent on an examination of the searcher demonstrate that:
site-catchphrase diagram. The technique is for (1) Task trail performs superior to anything session
removing sub-diagrams speaking to clients' primary and question trails in determining client
advantages from a website Keyword chart. It is fulfillment;
produced from increased Web crowd estimation (2) Task trail builds web page utilities of end clients
information (Weblog information). Our contrasting with session and inquiry trails;
commitments are 1) Web structure mining strategy (3) Task trails are equivalent to question trails yet
(PageRank) is appropriate for assessing the more touchy than session trails in estimating
characteristics of chose sub-diagrams and 2) two sorts distinctive positioning capacities;
of information (destinations and catchphrases) can be (4) Query terms from a similar assignment are more
spoken to and handled in a similar way utilizing a topically predictable to one another than
webpage catchphrase diagram. Our technique is basic question terms from various undertakings;
and it has capacities of separating sub-diagrams (5) Query proposal dependent on errand trail is a
identified with clients' principle advantages. Our decent supplement of question recommendations
work is required to give bits of knowledge for dependent on session trail and navigate
preparing boisterous heterogeneous chart structures. bipartitely.

This investigation of work [8] for the most part The discoveries in this paper check the need for
presents the technique for weblog mining, which can removing errand trails from web look logs and
find the method of web pages by burrowing web log improve applications in hunt and suggestion
records. By dissecting and investigating the standards frameworks.
of weblog records, we can recognize the potential
clients of the website and enhance the nature of data III.Discussion and Analysis
administrations to clients. In the phase of client
behaviour investigation, this examination investigates Research for breaking down Web log information has
the distinctions in client perusing behaviour in been finished by numerous analysts in the field of
various sorts of access occasions and figures the Web usage mining; finding regular examples of log

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


275
Veena Moharle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 273-277

information, demonstrating clients' route designs, At present, there is no standard meaning of session in
grouping clients of explicit Web webpage, and the International scholarly network. The meaning of
finding client networks. the session made by W3C Union is that it's a point of
confinement set of pages produced by clients' one or
A. Web Log Mining a few demand click. In the Weblog, the demand
records have a place with the clients who have
Web Usage Mining is the use of information mining distinctive IP addresses positively not in a similar
strategies to find intriguing usage designs from Web session. A client may peruse a similar site a few times
information to comprehend and better serve the and the time interim between each peruse
requirements of Web-based applications. Usage moderately expansive, at that point that is not in a
information catches the character or beginning of similar session. Session distinguishing proof's
Web clients alongside their perusing behaviour at a undertaking is to partition clients' peruse records into
Web webpage. a suitable session. We in every case initially do
distinctive presumptions on the client's perusing
behaviour on the customer, and after that distinguish
the session. These sessions can be utilized as a piece
of info information in grouping, bunching,
expectation and different errands. In light of a
uniform settled timeout, a conventional session
recognizable proof calculation is utilized.

Fig 1. Phases of Web Usage Mining IV.CONCLUSION

A server log is a log document (or a few records) Web usage mining is without a doubt one of the
consequently made and kept up by a server rising regions of research and imperative sub-space of
comprising of a rundown of exercises it performed. information mining and its systems. So as to take the
Web server logs are utilized to bunch web clients full favourable position of web usage mining and its
having comparative interests. It is likewise everything methods, it is essential to do pre-
characterized as adjusting administrations and data processing stage proficiently and viable. This paper
which was accessible on a website to the necessities endeavours to convey territories of pre-processing,
and the desires for an objective client, the dynamic including information purifying, session ID, client
client; the personalization undertaking by might recognizable proof. When the pre-processing stage is
profit by the information picked up from an all around performed, we can apply information
investigation of the client's navigational behaviour mining strategies like bunching, affiliation, order and
joined with different highlights which are curious to so on for utilization of web usage mining, for
a Web setting, to be specific its structure and content example, business insight, web-based business, e-
[10]. learning, personalization, and so on. Weblog mining
is one of the ongoing regions of research in Data
B. Session Identification mining. Web Usage Mining turns into an imperative
perspective in the present time in light of the fact
that the amount of information is ceaselessly
expanding. We manage the web server logs which

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


276
Veena Moharle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 273-277

keep up the historical backdrop of page demands Efficient Web Mining Algorithm for Web Log
Weblog record examination started with the reason Analysis: E-Web Miner ",1st International
to offer to Web webpage directors an approach to Conference on Recent Advances in Information
guarantee sufficient transfer speed and server ability Technology (RAIT) 2012
to their association. By examining these logs, it is [8] Xipei Luo, Jing Wang, Qiwei Shen, Jingyu
conceivable to find different sorts of learning, which Wang, Qi Qi ,"User Behavior Analysis Based on
can be connected behavior investigation of clients. User Interest by Web Log Mining ", 27th
International Telecommunication Networks
V. REFERENCES and Applications Conference (ITNAC),2017
[9] Zhen Liao, Yang Song, Yalou Huang, Li-wei
[1] G. Neelima , Dr. Sireesha Rodda, “Predicting He, Qi He,"Task Trail: An Effective
user behavior through sessions using the web Segmentation of User Search Behavior",IEEE
log mining”, Conference on Advances in Transactions on Knowledge and Data
Human Machine Interaction (HMI), R. L. Engineering,2014.
Jalappa Institute of Technology, Doddaballapur, [10] C., Ventura, S., Zafra, A., de Bra, P.: Applying
Bangalore, India, March 2016. Web usage mining for personalizing hyperlinks
[2] Anshul Bhargav , Munish Bhargav, “Pattern in Web-based adaptive educational systems
discovery and users classification through web (received January 8, 2009) (received in revised
usage mining”, International Conference on form May 4, 2009) (accepted May 4, 2009)
Control Instrumentation, Communication and
Computational Technologies, IEEE 2014.
[3] S. S. Patil, H. P. Khandagale, “Survey paper on
enhancing web navigation usability using web
usage mining techniques”, International journel
of modern trends inenginnering and
research.2016.
[4] Virendra R. Rathod and Govind V Patel,
“Prediction of user behaviour using web log
mining in web usage mining”, International
journel of computer application vol. 139- No. 8,
April 2016.
[5] . M. -S. Kao, T. Ozyer, R. Alhajj,"Hybrid
approach for predicting the behavior of Web
users", IEEE International Conference on
Information Reuse and Integration, Conf, 2005.
[6] Tsuyoshi Murata, Kota Saito,"Extracting Users'
Interests from Web Log Data ",
IEEE/WIC/ACM International Conference on
Web Intelligence (WI 2006 Main Conference
Proceedings) (WI'06)
[7] Mahendra Pratap Yadav, Pankaj Kumar
Keserwani, Shefalika Ghosh Samaddar,"An

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


277
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195169

Elimination of Impulse Noise using Mean Median filter for Image


Enhancement
Trupti Arun Jangale, Raj Kumar Paul
Department of Computer Science & Engineering, Vedica Institute of Technology, Bhopal, Madhya Pradesh,
India

ABSTRACT

In this method, we've got introduced a new technique for the improvement of gray scale images, when images
are corrupted by salt and pepper noise that's additionally referred to as an impulse noise. Our suggested
phenomena show a better output for Medium density impulse noise as compare to the opposite renowned
filters like standard Median Filter (SMF), a decision based mostly Median Filter (DBMF) and modified decision
based Median Filter (MDBMF), Nonlinear filter (NLF) and so on. Our projected technique worked on two steps,
within the beginning is that the detection of noisy pixels and within the second step is that the removal of
noisy pixels. For detection of noisy constituent apply condition pixels values lies in between 0 to 255 it noisy
it's noisy free pixels. In our second step that's the removal of noisy pixel recommended technique that's replaces
the noisy pixel by alpha trimmed mean median value. Different grayscale pictures are tested via proposed
technique. The experimental result shows higher Peak Signal to Noise ratio (PSNR) values and with higher
visual and human perception.
Keywords : Gray Scale, Impulse Noise, Trimmed Mean, Median, Unsymmetricness

I. INTRODUCTION degree economical de-noising technique to catch up


on such knowledge corruption.
Digital images play a very important role each in Image De-noising: -
existence applications like satellite television,
resonance imaging, laptop pictorial representation Image de-noising is a vital image process task, each as
also as in areas of analysis and technology like a method itself, and as a part in alternative processes.
geographical data systems and physical science. There ar some ways to de-noise a picture or a group
Knowledge sets collected by image sensors are of knowledge exists. the most properties of an honest
typically contaminated by noise. Imperfect image de-noising model ar that it'll take away noise
instruments, issues with the information acquisition whereas protective edges. historically, Image de-
method, and intrusive with natural phenomena will noising are often thought-about as a element of
all degrade the information of interest. what is more, process or as a method itself. within the 1st case, the
noise are often introduced by transmission errors and image de-noising is employed to enhance the
compression. Thus, de-noising is usually a necessary accuracy of varied image process algorithms like
and also the start to be taken before the image registration or segmentation. within the second case,
knowledge is analyzed. it's necessary to use associate

CSEIT195169 | Received : 01 Feb 2019 | Accepted : 10 Feb 2019 | January-February -2019 [ 5 (1) : 278-288 ]
278
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

the noise removal aims at rising the image quality for


visual examination.

The de-noising technique aims to attenuate noise


through 2 phases particularly,
1. Noise detection
2. Noise removal

Classification of De-noising Algorithm: -

As shown in Figure 2.2 below, there are two basic


approaches to image denoising, spatial filtering
Figure 1 (b): Introduction of noise into the clean
methods and transform domain filtering methods.
signal
• Spatial Filtering
• Non-Linear Filters
These embrace photographic grain noise, electrical
• Linear Filters
device noise, and transmission errors. Any globe
• Linear and nonlinear Filtering Approach
device is suffering from a particular degree of noise,
• Median Filter (M.F.)
whether or not it's thermal, electrical or otherwise.
• Mean Filter (M.F.)
Additive noise, in all probability the foremost
• LMS adaptative Filter
common kind, may be expressed as:
• Center Weighted Median Filter (CWM)
• Adaptative Mean Filters (AMF)
I (t) = S (t) + N (t)
• Progressive shift Median (PSM)
Where
• Switching Median Filter (SMF)
I (t) = ensuing knowledge measured at time t,
• Adaptative Weighted Median Filtering
• Adaptative Median Filter (AMF) S (t) = original signal measured, and
• Dision based Unsymmetrical Trimmed Mean N (t) = noise introduced by the atmosphere and
Filter (DBUTMF) different sources of interference.
• Modofied Dision primarily based
Unsymmetrical cut Mean Filter (MDBUTMF) kinds of noise
The types of Noise are following:-
Idea of noise:- I. Electronic equipment noise (Gaussian noise)
An image passes through many media and II. Impulse noise (Salt-and-pepper noise)
transmission channels before it reaches the process III.Shot noise (Poisson noise)
part. throughout this point, the image is also subject IV.Speckle noise
to interference from many sources. V.Brownian Noise

Image noise removal victimization Filters:-


Filtering whereas not Detection:-
this kind of filtering a window mask is affected across
the discovered image. The mask is usually of size
Figure 1 (a) : Introduction of noise into the system (2N+1)2, where N is also a positive range. usually the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


279
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

centre part is that the constituent of interest. once signal, the foremost obvious window is simply the
the mask is affected starting from the left-top corner primary few preceding and following entries,
of theimagetothe right-bottom corner, it performs whereas for 2nd (or higher-dimensional) signals like
some mathematics operations whereas not images, additional complicated window patterns are
discriminating any constituent. potential (such as "box" or "cross" patterns). Note that
if the window has associate degree odd range of
Detection followed by Filtering:- entries, then the median is easy to define: It's simply
This type of filtering involves two steps. In the center price on balance the entries within the
commencement it identifies screeching pixels and in window are sorted numerically. For an excellent
second step it filters those pixels. Here in addition a range of entries, there's quite one potential median.
mask is affected across the image and a number of
mathematics operations unit of measurement 1-D median Filter:-In 1D median filter, we have a
distributed to look at the screeching pixels. Then tendency to take into account the filtering window
filtering operation is performed exclusively on those slippy over a 1D array (either horizontal or vertical)
pixels that unit of measurement found to be of pixels. Assume the 5 pixels presently within the
screeching inside the previous step, keeping the non- windows are:
noisy intact.

Mean filter:-
Mean filtering may be a easy, intuitive and simple to Where the center picture element with price two
implement technique of smoothing images, i.e. hundred is associate isolated picture element and it's
reducing the quantity of intensity variation between out-of-range (noise). The median of those 5 values is
one pel and also the next. It's usually accustomed cut found by sorting the values (in either ascending or
back noise in images. The thought of mean filtering is falling order). the center price is that the median:
just to interchange every pel price in a picture with
the mean (`average') price of its neighbours, as well The original picture element price two hundred is
as itself. This has the impact of eliminating pel values replaced by the median a hundred and ten.
that square measure atypical of their surroundings.
Thearithmeticmean filter may be a terribly easy one 2-D median Filter:
and is calculated as follows: Replace every pel by the
typical of pels during a sq. window close this pixel

Median filter:-
The median filter is often accustomed scale back
noise in a picture, somewhat just like the mean filter.
However, it usually will a much better job than the
mean filter of protective helpful detail within the
image. the most plan of the median filter is to run Neighbouring pixel values = 115,119,120,123,
through the signal entry by entry, commutation 124,125,126,127,150
every entry with the median of close entries. The Median value = 124.
pattern of neighbours is named the "window", that Figure 2 (a) : Selection of filtering window of 2-D
slides, entry by entry, over the whole signal. For 1D

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


280
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

Input a 2-D Image (x by y) process simultaneously. In the suggested method


employ the smallest window dimension which
conserve the very well information of image.We
Preallocate another metrix of size (x+2 divide these pixels into matrix form and called it
by y+2) with zeroes window. Then we analyse every pixel by sliding that
window over the entire image. The window of size
Copy the input marix into the 3x3 choose for noise recognition and noise
Preallocate matrix elimination. The windowpane contains total 9
elements which are as follows:
Form a window matrix of size 3 x3
with the element of input matrix Z1, Z2, Z3, Z4, Z5, Z6, Z7, Z8, Z9.

Copy the window matrix (3 by 3) into Table 1 Filtering window of size 3x3
an array and sort it. Column 1 Column 2 Column 3
Row 1 Z1 Z2 Z3
Row 1 Z4 Z5 Z6
Find the median element. Here 5th
element (total elements are 9, middle Row 3 Z7 Z8 Z9
element will be 5).
Different blocks are shown for different steps. At the
starting there is filtering window of size 3x3 has been
Place 5th element into the output selected and calculated gray value of pixels according
matrix. Do the process for the to the strength of intensity of every pixel.
complete input matrix ALGORITHM

Convert the image into an image of 0- Step-1:-


255 colour range type.
In the first step select a image for image processing
than apply pre-processing task in this image byusing
Display the image without noise
pre-defined command in MATLAB. The prepressing
Figure 2 : Flow chart o 2-D Median Filter task are first convert our target image into a gray
scale level. Second convert image into a strand image
Another concern within the application of the that is 255X255. Now selected image is prepared for
median filter is that the form of the neighbourhood dealing out. For the pre doling out assignment and all
mask. A square-shaped mask will erode the corners of other things are implemented in MATLAB. In matrix
rectangular objects, whereas a cross-shaped lab. A large no. function file available for the
neighbourhood mask can leave corners intact. performing the initial task in the digital image
processing. Also done the padding of zeros and ones
II. PROPOSED METHOD in image.

We have develop the trouble-free algorithm in which Select a target image - [x] = imread(‘Test image.jpeg’)/
we execute the noise finding and noise removal for the read of image

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


281
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

Step-2:- Condition 1- If Pixels are between (0) Zero and two


In the second step apply impulse noise or can say salt five five (255) then they are noise free and shift to
and pepper noise in this image with the help Matrix reinstatement digital image pixels.
laboratory function. In this step apply noise with the
help of add-noise function in on the image. Condition 2- digital images pixel values are not
between in the range then they are moved to step 3.
[y] =imnoise(x,’Salt and Pepper’, Present of noise)
Step 3: In the step third check the window. Here two
Step-3:- cases are arise
In the third step for the preservation of corner, in 1. In a small 3X3 window contain all elements are
this step preserve the corner we the help of zeros and zeros (0’s) and two fifty five (255’s). In this
ones padding in the boundary side of the image. This condition apply mean filter. Take a mean of all
is use for the protection of corner of the digital image pixel in the small 3x3 windows all element aspect
pixels. centerpixel , and replace this value mean value by
the targeted pixel.
Step-4:-
If
W(ij) = [0] & [255] condition true
Noise elimination stage. In this stage first we divide a
{Mean (W (ij)) all pixels]=Replace by W(ij)
whole image into a small 3X3 window. Now relate
(Centervalue)
the trimming situation. This condition isknown
else
asnoise identification stage. In this stage check the
3x3 window pixel values combination of 0’s and
pixel values.In which pixels are stuck between 0 to
255’s. Now move to the case 2
255 ranges or not. Here two cases are generating.
end
If
2. In case 2, In a small 3x3 window pixels are the
X(ij) = 0<Y(ij)<255 / condition true Pixels Noise
combination of zeros and two fifty five. At this
Free
condition apply median filter for this case. Take a
else
median of all zeros and two fifty five and replace
X(ij) ≠ 0<Y(ij)<255 / condition Pixels are Noise
this value by centre pixel value.
End
If
Where X(ij) is the image small 3x3 windows pixels W(ij) = [0’s & 255’s ] / Combination in a 3X3
values.And Y(i,j) is the center pixel of targeted pixel window
of the small 3X3 window. {Median(W (ij) of all pixels}=Replace by W (i,j)’s
Center pixel value
Table 2 3x3 windows pixels value end
Step 4: Repeat all steps, all pixels in the whole
Column 1 Column 2 Column 3 255X255 image are 65,536 pixels are not
Row 1 Z1 Z2 Z3 processed.Hence a better de-noised image is obtained
Row 1 Z4 Z5 Z6 with improved PSNR MSE MAE and also shows a
Row 3 Z7 Z8 Z9 better image improvement. Therefore the overall
procedure of Modified unsymmetric trimmed median

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


282
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

Mean Filtering (MUTMMF), will next part of this the image, Y means the original image, 𝑌̂ shows the
report. Shows the result and simulation. de-noised image.
Graphic User Interface (GUI) in MATLAB:-
III. RESULT
In this preceding section, the image Tool GUI and
Mean Square Error (MSE): Measure GUI (Graphic User Interface) will be
introduced. Both of them were developed with the
The mean square error (MSE) dealings the value Matlab tool GUIDE (GUI Design Environment). In
modify or can say the changes to pixel intensity this GUI part shows the whole phenomena of
values between the original image (Xi)and de-noised proposed work like that image selection, impulse
image (Yj) and is given by noise addition, removal of noise and edges
preservation of the images in this research, in a user-
1 𝑁−1
𝑀𝑆𝐸 = ∑ (𝑋𝐽 ̅𝑗 )2
−𝑌 (5.1) friendly environment. First of all, the images are
𝑁 𝑗−0
selected by clicking the browse and Select Images. A
The MSE has been widely used to quantify image browser will appear afterwards to allow the user to
quality and when used alone, it does not correlate select the images, which can be any of the following
strongly sufficient with perceptual excellence. It file formats:
should be use therefore, with other quality check
parameter like SNR, PSNR, RMSEand visual Layout implementation of image enhancement based
perception. fusion techniques is done by following steps.

Peak signal-to-noise ratio (PSNR): ▪ First create simple Graphical User Interface (GUI)
window.
The PSNR is computed as - ▪ Second add impulse noise with the help of
𝑆2 imnoise command in this GUI.
PSNR=10log10 (𝑀𝑆𝐸 ) (5.2)
▪ Third step apply our proposed method that is
removal of noise.
Where 𝑆 is the maximum intensity in the original
▪ Fourth stepcalculated the image enhancement
image. The pear signal to noise ratio is higher for a
parameters like peak signal to noise ratio, mean
good resultant image and poorer for a reduced quality
square error, mean absolute error.
image. It calculate image reliability, that is, nearer
▪ Fifth step shows edges preservation of tested
the transformed image resembles the original
image.
image.In this proposed work on the source of our
▪ Sixth step graphical representation of PSNR and
image dimension 255×255, we mentioned peak signal
MSE values of the images. There are two
to noise ration and mean square error are as follows:
important push button design in this GUI these
are RESET and END.
The PSNR is expressed as:
(255)2
𝑃𝑆𝑁𝑅 = 10 𝑙𝑜𝑔10 (5.3)
(𝑀𝑆𝐸)
Where MSE acronym of Mean Square Error stands
for image enhancement factor, 𝑚 × 𝑛 is the size of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


283
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

Graphical representation of result of our proposed


method is shown in the blow the graph. Here in this
plot we have compare with different algorithms they
are - median filter,adaptive median filter, decision
based algorithms, modified decision based
unsymmetric trimmed median filter, Non-linear
Filter algorithm. This method is tested on the ‘Lena’
image of standard sizeshown plotsof PSNR values of
different noise. Our proposed method provide a good
result in low and medium noise density. Here
compression with different filters against noise
Fig. 3 – Shows the GUI of Proposed Method densities for Lena image is shown in Figure .

TABLE 4 CALCULATED PSNR OF PROPOSED In the figure 5 put side by side our proposed method
ALGORITHM FOR LENA IMAGE visual results compare with different algorithms they
are - median filter,adaptive median filter, decision
based algorithms, modified decision based
unsymmetric trimmed median filter, Non-linear
Filter algorithm. This method is tested on the ‘Lena’
image of size 256×256 in the 80% noise density ‘Lena’
image result with our planned algorithm. Shown
plotsof PSNR values of different noise. Our proposed
method provide a good result in low and medium
noise density.

(a)Original Image (b) 70% Noise (c) MF (d) AMF (e) PSMF (f) DBA

(g)MDBA (h )MDBUTMF (i) Proposed

Figure 4: Proposed outcome for ‘Lena’ test image

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


284
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

Table 5 Different values of Barbra image


Result of BARBRA Image
Noise Density PSNR MSE(dB) MAE (dB) Time (s)
3% 44.9155 2.0967 0.1486 2.24
5% 42.3732 3.765 0.2541 2.67
10% 38.5745 9.0288 0.5632 2.97
20% 35.1368 19.9249 1.1749 4.01
30% 32.5039 36.5334 1.9553 4.82
40% 30.3439 60.0745 2.8946 6.03
50% 28.7536 86.641 3.9566 6.17
60% 26.5334 144.4569 5.582 7.17
70% 24.1005 252.9446 7.8463 8.15
80% 20.8724 531.9078 12.599 9.63
90% 16.9621 1308.8 23.8651 11.49

Original Image Noise Image 10% Restored Image

Original Image Noise Image 20% Restored Image

Original Noise Image Restored


Image 30% Image

Figure-6 Shows the outcome of proposed method for ‘girl’ image

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


285
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

Table 6 different values of Girl image

Result of GIRL Image


Noise Density PSNR MSE(dB) MAE (dB) Time (s)
3% 44.2093 2.4669 0.1751 2.5
5% 43.2871 3.0505 0.2328 2.5
10% 40.7521 5.4685 0.3715 2.99
20% 37.2585 12.2244 0.7707 3.83
30% 35.1066 20.0642 1.1863 4.67
40% 32.5532 36.1213 1.7769 5.66
50% 30.2181 61.8405 2.5712 6.65
60% 28.1567 99.406 3.6035 7.6
70% 25.1084 200.5603 5.5773 8.89
80% 21.8142 428.2142 9.117 9.55
90% 17.5649 1139.2 18.7688 9.26

IV. CONCLUSION MSE and PSNR. Indifference to mean filter and


further accessible algorithms, the new algorithm
The new proposed algorithm has been introduced exploits a tiny 3x3 window containing only eight
to manage the problems, namely, poor image neighbors of the corrupted pixel that have
enhancement or we can say enhancement at high advanced bond this provides a good edge
noise density that is frequently enhanced in the protection , also better edge protection as well as
mean absolute error (MAE), mean square error more better person and illustration prescription.
(MSE) and peak signal to noise ratio. In this The proposed filter demonstrates trustworthy
proposed workModified unsymmetric trimmed and conventional performance among various
median Mean Filtering (MUTMMF) is used for range of noise densities increasing from 10%-
enhancing both the peak signal to noise ratio. 90%. The performance of the introduced
The performances of proposed Modified methodology has been tested at low and medium
unsymmetric trimmed median Mean Filtering densities on different level. In case of high noise
(MUTMMF) are quantitatively vies as well as the density levels the new proposed algorithm gives
individual perception show better outcome as better presentation as evaluate with other
compare last decade algorithms. Results reveal presented noise removal algorithms.
that the proposed filter exhibits improved
performance in comparison by means of Median V. REFERENCES
Filter (MF) [5], Adaptive Median Filter (AMF) [7],
Progressive Switching Median Filter (PSMF) [1] Digital Image Processing, Rafael C. Gonzalez,
[11],Decision Based Algorithms (DBA) [16], Richard E. Woods, Prentice Hall, Upper
Modified Decision Based Algorithms (MDBA), Saddle River, New Jersey 07458.
Modified Decision Based Unsymmetric Trimmed
Median Filter [29], Modified Non-Linear Filter
Algorithm (MNF)in terms of improved MAE,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


286
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

[2] Fundamentals of Digital Image Processing, S. [12] A. Ben Hamza, P. Luque, J. Martinez, and R.
Annadurai, R. Shammugalakshmi, Pearson Roman, “Removing noise and preserving
Education India. details with relaxed median filters,” J. Math.
[3] Castleman Kenneth R, Digital Image Imag. Vision, vol. 11, no. 2, pp. 161–177, Oct.
Processing, Prentice Hall, NewJersey, 1979 1999.
[4] T.A. Nodes and N.C. Gallagher, Jr., “The [13] V. Strela. “Denoising via block Wiener
output distribution of median type filters,” filtering in wavelet domain”. In 3rd European
IEEE Trans. Communication., 32(5): 532-541, Congress of Mathematics, Barcelona, July
1984. 2000. BirkhäuserVerlag.
[5] A.K.Jain,Fundamentals of digital image [14] S. Zhang and M. A. Karim, “A new impulse
processing. Prentice-Hall, 1989. detector for switching median filters,” IEEE
[6] David L. Donoho and Iain M. Johnstone., Signal Process. Lett., vol. 9, no. 11, pp. 360–
“Adapting to unknown smoothness via 363, Nov. 2002.
wavelet shrinkage”, Journal of the American [15] T. Yahagi. "A Random-valued Impulse
Statistical Association, vol.90, no432, NoiseDetector Using Level Detection", 2005
pp.1200-1224, December 1995. National IEEEInternational Symposium on Circuits
Laboratory, July 27, 2001. and Systems, 2005.
[7] H. Hwang and R. A. Hadded, “Adaptive [16] K. S. Srinivasan. "A New Fast and Efficient
median filter: New algorithms and results,” Decision-Based Algorithm for Removal of
IEEE Trans. Image Process., vol. 4, no. 4, pp. High-Density Impulse Noises", IEEE Signal
499–502, Apr. 1995. Processing Letters, 3/2007
[8] E. Abreu and S.K. Mztra,” A Signal- [17] JafarRamadhan Mohammed. "An Improved
Dependent Rank Ordered Mean (Sd-Rom) MedianFilter Based on Efficient Noise
Filter –ANew Approach For Removal Of Detection for HighQuality Image
Impulses From Highly Corrupted Images”, 0- Restoration", 2008 Second Asia
7803-2431- 5/95$ 4.000 1995 IEEE. InternationalConference on Modelling &
[9] R. Yang, L. Yin, M. Gabbouj, J. Astola, and Y. Simulation (AMS).
Neuvo, “Optimal weighted median filters [18] V.R. Vijaykumar. "Detail preserving median
understructural constraints,” IEEE Trans. basedfilter for impulse noise removal in
Signal Processing vol. 43, pp. 591–604, Mar. digital images",2008 9th International
1995. Conference on Signal Processing.
[10] Scott E Umbaugh, Computer Vision and [19] S. Saudia. "Salt & pepper impulse detection
Image Processing, Prentice HallPTR, New andmedian based regularization using
Jersey, 1998. Adaptive Median Filter", TENCON 2008 -
[11] Zhou Wang and David Zhang, “Progressive 2008 IEEE Region 10 Conference, 11/2008
Switching Median Filter for the Removal of [20] DENG Xiuqin, XIONG Yong PENG Hong , A
Impulse Noise from Highly Corrupted new kind of weighted median filtering
Images", IEEE Transactions On Circuits And algorithm used for image Processing,
Systems—Ii: Analog And Digital Signal International Symposium on Information
Processing, Vol. 46, No. 1, January 1999. Science and Engieering, 2008.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


287
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

[21] Ashraf Aboshosha. "Image denoising based on


spatialfilters, an analytical study", 2009
InternationalConference on Computer
Engineering & Systems, 12/2009.
[22] V.R.Vijaykumar, P.T.Vanathi,
P.Kanagasabapathy and D.Ebenezer, “Robust
Statistics Based Algorithm to Remove Salt
and Pepper Noise in Images”, International
Journal of Information and Communication
Engineering 5:3 2009.

Cite this article as :

Trupti Arun Jangale, Raj Kumar Paul,


"Elimination of Impulse Noise using Mean Median
filter for Image Enhancement", International
Journal of Scientific Research in Computer
Science, Engineering and Information Technology
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue
1, pp. 278-288, January-February 2019. Available
at doi : https://doi.org/10.32628/CSEIT195169
Journal URL : http://ijsrcseit.com/CSEIT195169

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


288
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195145

E - Tax Information System


M. Florence Dayana*1, B. Vinithamani 2
1 Head & Assistant Professor, Department of Computer Applications, Bon Secours College for Women,
Thanjavur, Tamil Nadu, India
2M. Sc., Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India

ABSTRACT
For that purpose, we are developing this application, which can collect the tax information from different
authorities based on the house no’s and owner name by using Web Services and stores the data in our own
database. For getting the information, regarding taxes citizens need to register first. The bill is calculating on
different modes like annual, half-yearly, quarterly etc. By using the user, id and password citizens can log into
the system and collect the necessary information from the home page. For this every user need to provide his
house no and owner name. By clicking on the search button, they will get their own.
Keywords: E- TAX, Web Services

I. INTRODUCTION web service is only capable to get the information


from other sources.
Local bodies like municipalities need collect different
kind of taxes from the citizens. Every citizen needs II. METHODS AND MATERIAL
to pay Water Bills and Electricity Bills to the
particular authorities. Actually, a citizen wants to Local bodies are classified into Municipal
pay his municipal tax means he should go to Corporations, Municipalities of different grades
municipal office and collect the necessary bill, pay depending upon the population and income of the
there only. Same as water and electricity bills. This Urban Local Body. These Urban Local Bodies have
is not an easy job to get all this tax information in been constituted with the objective of discharging
different authorities now days. By using the user id certain.
and password citizens can log into the system and Obligatory functions like
collect the necessary information from the home Supply of Drinking Water
page. For this every user need to provide his house Providing Water connection to houses
no and owner name. By clicking on the search Sending water bills to houses
button, they will get their own Public Street Lighting
• House Details Calculating and sending Electric Bills
• Electricity Bill Details Maintaining sanitation and hygiene of public places
• Water Bill Details Municipal Library maintenance.
• Municipality Bill Details Calculate and sending Municipal tax and house taxes.
Web services play vital role here. Why because this The development of this new system contains the
system get the information from other sources. A following activities, which try to automate the entire
process keeping in the view of database integration

CSEIT195145 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 289-291 ]
289
M. Florence Dayana, B. Vinithamani. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 289-291

approach. User Friendliness is provided in the PROPOSED METHODOLOGY


application with various controls provided by system To debug the existing system, remove procedures
Rich User Interface. The system makes the overall those cause data redundancy, make navigational
project management much easier and flexible. It can sequence proper. To provide information about users
be accessed over the Intranet. The employee on different level and to reflect the current work
information can be stored in centralized database status depending on organization. To build strong
which can be maintained by the system. This can password mechanism
give the good security for user information because
data is not in client machine. Authentication is IV. RESULTS AND DISCUSSION
provided for this application only registered Users
can access. There is no risk of data management at
any level while the project development is under
process. The automated system will provide to the
employees for reliable services. The proposed system
using web services, a web service can get the
information from other sources also. There no burden
of calculating water, electric, house tax bills in this Administrator
system, by sending a request to the particular Citizen’s
authority via web service, we can get all the Web Services
information regarding these issues. Citizen’s Bill Information
Search
III. EXISTING PROCESS Reports
Authentication
The existing system is a manual system. Here the
employees needs to save the information in the form Administration
of excel sheets or Disk Drives. There is no sharing is
possible if the data is in the form of paper or Disk Administration is the main person of this system. He
drives. The manual system gives us very less security can have all the privileges to do anything in this
for saving data; some data may be lost due to is system. Admin can get the information regarding
management. It’s a limited system and fewer users citizen’s water bills, electric bills, and municipal taxes
friendly. Searching of particular information is very from various sources. For that administrator need to
critical it takes lot of time. Gathering information of send a request for that particular authority throw
different sources is not an easy job, data will be web services.
mismanaged. Calculating different bills manually
going to be as a process of mistake. There is no Citizens
interface to provide various bills in the existing People who are living in that municipality area are
system. In the existing system citizen get the bills called as citizens. Citizens need to pay water bill,
information manually, in various authorities like electricity bill, property tax for a certain of intervals.
water board, electricity department, property tax By using this system interface citizens know their
department etc. different bill amounts, pending bills, due dates etc.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


290
M. Florence Dayana, B. Vinithamani. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 289-291

Web Services taxpayers who are using the e-tax system. With these
Citizen Bill Information limitations the result of our study gives an insight to
The system interface is going to show the Bill e-tax quality evaluation criteria in an emerging
information of different authorities. The bills mainly economy like Nigeria and also provide the basis for
going to these types further research in these areas.
Water Bills
Electricity Bills VII. REFERENCES
Property Taxes
These interfaces also show the [1]. Productivity of the Nigerian Tax System: 1970-
Pending Bills 1990 AERC Research Paper 67. Nairobi: African
Penalties Economic Research Consortium. Black. (1999).
Due Dates Black’s Law Dictionary (Seventh Edition). West
Group. Centers for Medicare and Medicaid
V. CONCLUSION Services Office of Information Service. (2008,
March 27).
[2]. Selecting a development approach. United States
It has been a great pleasure for me to work on this
Department of Health and Human Services (HHS).
exciting and challenging project. This project proved
Clarke, J. M. (1998).
good for me as it provided practical knowledge of not
[3]. Automated test generation from a Behavioral
only programming in ASP.NET andC#.NET web-
Model. Pacific Northwest Software Quality
based application and no some extent Windows
Conference. IEEE Press. Federal Republic of
Application and SQL Server, but also about all
Nigeria. (1997).
handling procedure. It also provides knowledge about
[4]. Report of the Vision 2010 Committee. Abuja,
the latest technology used in developing web enabled Nigeria. Federal Republic of Nigeria. (2002).
application and client server technology that will be [5]. Personal Income Tax Act 1993(1993 No. 104):
great demand in future. This will provide better [6]. Operation of Pay as You Earn (PAYE) Regulations
opportunities and guidance in future in developing 2002 Official Gazette. Federal Ministry of Finance
projects independently. Abuja. Geoffrey, E. (2004).
[7]. Global Business Information Technology: An
VI. FUTURE ENHANCEMENT Integrated Systems Approach. Pearson Education.
Grady. (2005).
This paper examines the taxpayers’ e-tax system [8]. Unified Modelling Language User Guide the
implementation by validating the psychometric Second Edition. Addison-Wesley. International
properties and using the instrument validated for e- Organization for Standardization. (2015, February
tax service evaluation criteria in the context of 26).
Nigeria. It is concluded that all the six variables,
Cite this article as :
compatibility, complexity, reliability, affordability, M. Florence Dayana, B. Vinithamani, "E - Tax
ease of use and tangible are important in the Information System", International Journal of Scientific
evaluation of e-tax system implementation. The items Research in Computer Science, Engineering and
used should enable the tax authority to bench mark Information Technology (IJSRCSEIT), ISSN : 2456-3307,
their routine and plan for enhancement. The first Volume 5 Issue 1, pp. 289-291, January-February 2019.
limitation of these study is small sample size and the Available at doi : https://doi.org/10.32628/CSEIT195145
population been limited to only self-employed Journal URL : http://ijsrcseit.com/CSEIT195145

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


291
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195154

Children Tracking Classification Control Using Android


M. Juno Isabel Susinthra*1, S. Praveena 2
1Department of Computer Applications, Bon Secours College for Women, Thanjavur, Tamil Nadu, India
2M. Sc., Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India

ABSTRACT

Children tracking system based on android terminals is proposed. Recently, all over the world crime against the
children in the age of 14 to 17 years is more popular. Parent’s always worry about their children whenever they
are outside from the home. In this paper, the proposed system consists of two sides out of them one is parent
module and another is the child module. The child module consists of ARM7 Microcontroller (LPC 2148), GPS
(Global Positioning System), GSM (Global System for Mobile Communication) and voice chip where the parent
module consists of android mobile phone. This paper gives the information about missing child from school
campus. There are two android mobile phones for the safety of the both modules. The system tracking the child
from source to destination i.e. from home to school or anywhere.
Keywords : ARM Based Children Tracking System, arm7microcontroller, GSM, GPS and Android Mobile
Phone.

I. INTRODUCTION services of a smart phone which is GSM more


specifically SMS in one system.
Children tracking system is widely used all over the
world to assure their parent‘s that their wards are safe II. METHODS AND MATERIAL
from suspicious actions in today‘s world over 80% of
the world population in the age of seven or eight LITERATURE SURVEY
owns smart phone. The proposed system includes
tracking the child movements outside from the home. Monitoring mobile devices has been a concern for
This is due to reason one of them is the remarkable not only individual users, but also for organizations,
features and capabilities that new smart phone offers communities, and scientific researchers. The
especially android based smart phones. GPS offers challenge of using monitoring applications is about
outstanding capabilities in locating positions and this how the collected data will be handled and the how
can be used to developed resourceful application that to build a macroscopic pattern from the collected
helps in locating missing or lost children. This project data. This part of the survey discussed the summary
is designed to be used by parents and aimed to help of the selected research papers related to monitoring
locating missing or lost children. It takes advantages mobile devices. In, the authors discussed the
of the fact that many of today ‘s children bring smart importance of having mobile phone sensing to
phone which is convenient for this kind of situation. develop micro and macro views of mobile usage. This
In this paper GPS is combined with one of the basic importance increases as the mobile devices have
become the main computing and communication

CSEIT195154 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 292-295 ]
292
M. Juno Isabel Susinthra, S. Praveena Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 292-295

devices. Therefore, research areas on data collection, among different apps, which increase the difficulty in
analysis and patterns learning are becoming rich monitor single flow between an app and the
discussion area, especially with all the cheap features operating system. The proposed platform only tracks
and technological advances available on phones data flows but not control flows. Data flows are the
allowing for a variety of users experiences and usage. explicit flowing of data through the apps. Control
The framework architecture proposed in is built upon flows are the implicit flowing of data that require
using mobile phone sensing and the cloud computing. analytical analysis.
The paper discussed the different challenges with
gathering data techniques using mobile and EXISTING PROCESS
unexpected environment (mobile context), as well as,
the privacy issues on gathering personal information Android terminals have wireless LAN and Bluetooth
from a third-party application. The study discussed device. It results in lack of individual attention
the three different scales mobile sensing considered towards the children. It offers less security. Studies
beneficial on: individual, group, and community data conducted by Cyber Travel Tips, showed that in
gathering. The scalability feature of gathering Malaysia, missing children are basically classified into
individual behaviour to learn from community two categories. The first category is disappearance,
patterns may help improve communities in social, which includes running away from home. Children
healthy and environmental studies. The study also tracking system is also developed on mobile ad hoc
explained the two sensing paradigms: participatory networks. System developed in says that in GPS
sensing, which involves users who lunch sensors system and tag-based system, each parent cannot
manually, and opportunistic sensing, where sensors obtain group information on the vicinity of the child.
automatically collect data. The main focus of the A self -configurable new generation children tracking
introduced work in is about understanding how system. Hiroshima City Children Tracking System is
personal data travels through selected third-party a safety support system for children based on ad hoc
applications. The authors raised the issue about network technologies. Field experiments have been
violating privacy barriers when applications access conducted in cooperation with an elementary school
the personal information by only ask for users’ in Hiroshima. In this paper, propose a new
permissions to access the data with no explanation of generation child tracking system which is based on
how the data will be used. The research proposed a experiences and findings of the field experiments for
monitoring platform based on Android mobile phone Hiroshima children tracking system. Existing
to track the data flow through 30 selected technologies, however, are not powerful to prevent
applications. After testing the proposed the platform, crimes against children and helpful parent ‘s since it
the research concluded that 20 apps miss used the is difficult to take information of children as a group.
users’ private data and 15 apps used users’ locations to
support marketing services. The main challenges in PROPOSED METHODOLOGY
monitoring the use of mobile apps include the
shortage of Smartphone used data as a resource The children information is transmitted and received
constrained, as well as, the dynamic nature of data using GSM technology. The child module acts as a
flow using mobile devices. This dynamicity causes transmitter which includes ARM7 microcontroller
another context-based challenge because data could (lpc 2148), GSM module, GPS module and voice chip
be sent at any time and any place. Further, mobile as well as android mobile phone. The receiver
apps allow for across sharing data and information module includes Android mobile phone and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


293
M. Juno Isabel Susinthra, S. Praveena Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 292-295

monitoring database. The position of the moving defining the location of their children (indoor or
child is tracked by is tracked by GPS and is sent to outdoor) and to give the safety distance for each child.
ARM7 microcontroller. This controller forwards the
GPS data (latitude and longitude) to GSM board. Detection of Safety Distance When parent starts the
GSM will in turn send the position of the moving system, she/he should select where they stay; indoor
child to two receivers. It allows the parent to get or outdoor; in order to calculate the distance between
their child ‘s location. them and each child. Indoor option limit child range
to 3m by reducing child transmitter source so signal
III. RESULTS AND DISCUSSION attenuated will not be detected. But if parent selects
outdoor option, distance will be calculated by taking
IMPLEMENTATION RESULTS child and parent locations from GPS chips. Child
module will send its GPS reading to parent module
which calculates distance and compares it with
distance threshold; this threshold value should be
entered by parent after selecting outdoor option.

IV.CONCLUSION

All parent wishes to shield their children from real


dangerous they will inevitably undergo hard times.
Child Module The child module is attached to child. Normal behaviour of any child depends on the child's
Its primary role is to periodically receive messages age, personality, and at the end the physical and
and in response send messages to the parent module emotional development. A child's behaviour may be a
and alert them if the child is in danger. The child problem if it doesn't match the expectations of the
module also has a buzzer alarm that sounds whenever family or if it is disruptive. So Parents may face many
the child is in an alarmed state. This allows a parent problems with their children, when they let them
to more easily locate the troubled child. The sensors free without any observation especially if they have
that used with this module are described below. many children that may face many dangers.

Parent Module The parent module is the V. FUTURE ENHANCEMENT


computational brain of the entire system. It is
responsible for communication with all the children There are many tracking system available in present
and calculating their distance. It also analyzes time, these systems use different technologies but
incoming packets to see if any child has been in an each of them has one or more limitation such as not
alarmed state or not. Finally, if any child stops suitable for many children or monitor how far the
sending packets, the parent assumes it is out of range child from their parent without monitoring the
or in some other danger and also triggers an alarm. environment surrounded the children or can monitor
Every time the parent module cycles through all the only one state at a time. While the designed tracking
children, it packs the status of every child and system allows parent to monitor multiple children
displays information on the LCD. Also, parent and they will be alarmed if any child be in danger
module has a keypad which is used to let the parent state. The danger states that were taken in the
consideration are: missing far away from their parent,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


294
M. Juno Isabel Susinthra, S. Praveena Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 292-295

falling into a swimming pool, leaving in a hot car, Cite this article as :
leaving behind gas heater. The flexibility of the
designed system makes it easy to add new sensors for M. Juno Isabel Susinthra, S. Praveena, "Children
any child. Tracking Classification Control Using Android",
International Journal of Scientific Research in
VI.REFERENCES Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
[1] Druin A., "The role of children in the design of 5 Issue 1, pp. 292-295, January-February 2019.
new technology", Behavior and Information Available at doi :
Technology", vol21, No.1,pp1-25, 2002. https://doi.org/10.32628/CSEIT195154
[2] Morris Williams, Owain Jones, Constance Journal URL : http://ijsrcseit.com/CSEIT195154
Fleuriot and Lucy Wood, "Children and emerging
wireless Technologies", Conference on Human
Factors in Computing systems, 2005.
[3] Deborah L. Ruf, "Losing Our Minds", 2005, Great
Potential Press, Inc.
[4] Stephen S. Intill, James W. Davis and Aaron F.
Bobick, "Real-Time Closed-World Tracking",
Proceedings of the IEEE Computer Sciety
Conference on Computer Vision and Pattern
Recognition, pp. 697-703, June 1997.
[5] GregoryT.FRENCH "Understanding the GPS,
1996
[6] Katin Michael, "The Emerging Ethics of Human
Centric GPS Tracking and Monitoring",
International Conference on Mobile Business
(ICM'06), 2006.
[7] A. (Angela) Canny, Anne E. Green and Malcolm
J. Maguire, "Keeping Track", Published 2001, The
Policy Press.
[8] PIC16F87XA Datasheet 28/40-Pin Enhanced
Flash
[9] Lovine, John, “PIC Microcontroller Project
Book“, Mc Graw-Hill, 2004

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


295
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195163
Voice Based Information Sharing System for Visually Impaired People
R. Gomathijayam1, A. Jenifer2
1 Assistant Professor, Department of Computer Applications, Bon Secours College for Women, Thanjavur, Tamil
Nadu, India
M. Sc., Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India
2

ABSTRACT
In today’s world communication has become so easy due to integration of communication technologies
with internet. However, the visually challenged people find it very difficult to utilize this technology
because of the fact that using them requires visual perception. Even though many new advancements have
been implemented to help them use the computers efficiently no naïve user who is visually challenged can
use this technology as efficiently as a normal naïve user can do that is unlike normal users they require
some practice for using the available technologies. This paper aims at developing an email system that will
help even a naïve visually impaired person to use the services for communication without previous
training. The system will not let the user make use of keyboard instead will work only on mouse operation
and speech conversion to text. Also this system can be used by any normal person also for example the
one who is not able to read. The system is completely based on interactive voice response which will make
it user friendly and efficient to use.
Keywords : Visually Challenged People, IVR, Speech to Text Converter, Mouse Click Event, Screen
Reader

I. INTRODUCTION impaired it become necessary to make internet


facilities for communication usable for them also.
Internet is considered as a major storehouse of Therefore we have come up with this project in
information in today’s world. No single work can be which we will be developing a voice based email
done without the help of it. It has even become one system which will aid the visually impaired people
of the de facto methods used in communication. And who are naive to computer systems to use email
out of all methods available email is one of the most facilities in a hassle free manner. The users of this
common forms of communication especially in the system would not need to have any basic information
business world. However not all people can use the regarding keyboard shortcuts or where the keys are
internet. This is because in order to access the located. All functions are based on simple mouse
internet you would need to know what is written on click operations making it very easy for any type of
the screen. If that is not visible it is of no use. This user to use this system.
makes internet a completely useless technology for
the visually impaired and illiterate people. Even the II. LITERATURE SURVEY
systems that are available currently like the screen
readers TTS and ASR do not provide full efficiency to It is estimated that there are a total of 4.92 billion
the blind people so as to use the internet. As nearly email accounts existing in 2017 and there will be
285 million people worldwide are estimated visually approximately 5.59 billion accounts by the end of

CSEIT195163 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 296-298 ]
296
R. Gomathijayam, A. Jenifer Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 296-298

2019[4]. It is also estimated that there are a total of perception. Even though many new advancements
340.2 million smart phone users in India in the year have been implemented to help them use the
2017 [5]. This makes emails the most used kind of computers efficiently no naïve user who is visually
communication. The prevailing email systems don’t challenged can use this technology as efficiently as a
give any means of feedback or Talkback service. The normal naïve user can do that is unlike normal users
most common mail services that we tend to use in they require some practice for using the available
our day to day life cannot be used by visually technologies.
challenged people. This is as a result of they do not
offer any facility in order that the person in front will IV. PROPOSED METHODOLOGY
listen the content of the screen. As they cannot
visualize what is already present on screen they The proposed system relies on voice command based
cannot build out where to click in order to perform system unlike the existing mail systems. The most
the required operations. For a visually impaired important thing that has been kept in mind while
person employing a computer or smart phone system developing the proposed system is accessibility. A
for the first time isn't that convenient as it is for a system is accessible solely if it may be used
standard user even though it is user friendly. Though expeditiously by all varieties of individuals whether
there are several screen readers offered then also or not disabled. The current systems do not offer this
these individuals face some minor difficulties. Screen accessibility. Therefore the system we are developing
readers speak out whatever content is there on the is completely different from the current system.
screen and to perform the particular actions the Unlike current system that emphasizes a lot on user
person will have to use keyboard shortcuts because friendliness of traditional users, our system focuses a
mouse location cannot be detected by the screen lot on user friendliness of all varieties of individuals
readers. This means 2 things; one that the user cannot as well as normal people visually impaired individuals.
make use of mouse pointer as it is fully inconvenient The complete system is primarily based on speech to
if the pointer location cannot be derived and second text commands. Once using this system the
that user should be versed with the keyboard on application will be prompting the user to speak
wherever each and every key is placed. A user who is specific commands to avail respective services and if
new to computer will therefore not use this service as the user wants to access the respective services then
they're not conscious of the key locations. Also there he/she needs to speak that command.
are some difficulties faced by visually impaired
people when using smartphone systems. All these are V. IMPLEMENTATION RESULTS
some drawbacks of the present system that we are
going to overcome within the application we are
developing.

III. EXISTING PROCESS

In today’s world communication has become so easy


due to integration of communication technologies
Speech-To-Text: Input to the System is provided by
with internet. However the visually challenged
the means of speech-to text. Speech-to-Text API is
people find it very difficult to utilize this technology
used for conversion of speech-totext.HTML5’s Built
because of the fact that using them requires visual

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


297
R. Gomathijayam, A. Jenifer Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 296-298

in Speech Recognition API is used for this purpose keyboard usage is eliminated. The user solely has to
hence using this input is provided to different follow the directions given by the system and use
modules of the system as intended user cannot voice commands consequently to get the several
manually enter the data required. services offered. Other than this the user may have to
Text-To-Speech is used for conversion from text to be requested to feed info through voice inputs
speech. It is used to Speak menu to the user which whenever required.
can be used to provide list of available options a
particular user can do on a given module or page. It is VII.FUTURE ENHANCEMENT
also used to read message that is been received by the
intended user. As intended user cannot directly read. Voice could be extended to image attachments and
ResponsiveVocic.js is API that is used for conversion other options such as indentation, fonts etc., that are
of Text to speech. available with normal E-Mail.
Php Mailer and PhP-IMAP PhP Mailer is a library
which is provided by PhP which can be used to send VIII. REFERENCES
e-mail. PhP Mailer is used in proposed system which
is used to send emails. In order to extract the mails [1]. T. Shabana, A.Anam, A.Rafiya, K.Aisha, “Voice
from the system Php Imap is used to fetch the user’s based Email System for Blinds”, “IJARCCE”, Jan
mail from the IMAP server. 2015.
Knuth-Morris-Pratt Algorithm Pattern Matching is [2]. P. Ingle, Harshada Kanade, Arti Lanke, “Voice
very key component when it comes to searching mail based Email System for Blind”, “IJRSCSE”,
in inboxes. System provides three ways to search by 2016.
keyword, date and today’s date out of search by [3]. Jagtap Nilesh, Pawan Alai, Chavan Swapnil,
keyword requires KMP for pattern matching. Given a “Voice based System in Desktop and Mobile
string ‘S’, the problem of string matching deals with devices for blind people”, “IJETAE”, Feb 2014.
finding whether a pattern ‘p’ occurs in ‘S’ and if ‘p’ [4]. Prof Umesh A Patil, Pranouti Patil, “A survey
does occur then returning position in ‘S’ where ‘p’ on voice based mails system for physically
occurs. Knuth, Morris and Pratt proposed a linear impaired peoples”, “IJIRCCE”, Jan 2016.
time algorithm for the string matching problem as
compared to other approaches. Cite this article as:

VI.CONCLUSION R. Gomathijayam, A. Jenifer, "Voice Based


Information Sharing System for Visually Impaired
In this paper we have planned a system which can People", International Journal of Scientific Research
facilitate the visually impaired individuals to access in Computer Science, Engineering and Information
email services efficiently. This system can help in Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
overcoming some drawbacks that were earlier faced 5 Issue 1, pp. 296-298, January-February 2019.
by the blind individuals in accessing emails. We've Available at doi :
eliminated the thought of using keyboard altogether https://doi.org/10.32628/CSEIT195163
with screen readers which can help reducing the Journal URL : http://ijsrcseit.com/CSEIT195163
cognitive load of remembering keyboard operations.
Conjointly any user who will not grasp the location
of keys on the keyboard would like not worry as

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


298
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195149
A Review on Big Data Analytics Via Social Media
T. K. Anusuya*1, A. Preethi2
1 Head & Assistant Professor, PG Department of Department of Computer Applications, Bon Secours College for
Women, Thanjavur, Tamil Nadu, India
M. Sc., Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India
2

ABSTRACT

Analytics is very important in all fields in order to make decisions over certain facts. Social media analytics is
the process of collecting information from various social media platforms, websites and blogs. These analytics is
done to make effective business conclusions. The usage of social media has become the latest trend in today’s
world. Social data analytics is not about just collecting likes and comments shared by individuals but it has
become the platform for many trademarks to bring out promotion. Applications such as marketing, elections
widely used social data to make predictive decisions. Some of the approaches followed are forming hypothesis,
getting deep into the data, mapping events etc. These analytics can also be done in applications such as business,
Change in amendments, Education, Demonetization etc. The challenges faced are metrics formed by social
media should reach the right people, unstructured data being difficult to priestship paper discusses about the
model, theme, performance evaluation, advantages and disadvantages under literature survey.
Keywords : Social Media, Insights, Datasets, Analysis, Challenges

I. INTRODUCTION various applications in which Big data is being used.


The foremost is the Banking in which Securities
Big data is nothing but huge amount of data which Exchange Commission is being used to monitor
cannot be processed using traditional methods. The various activities. At present, they are using the
data to be processed may be structured, unstructured network analytics and natural language process as
or semi structured. These data come from various trend in the markets. Retail traders, big banks and
sources such as sensor data, transactional data, many others are using analytics for making decisions.
cameras, microphones and lot more of resources. On Moving on to the next application, it is the media.
these large data sets analytics is performed in order to Big data is widely used in media, since it has now
obtain undiscovered insights. Based on facts, reached the users through much hand held devices.
everyday petabytes of data are being generated and Here analysis is done on the consumer insights, social
now it has come up to zeta data. All social media media content and understanding pattern. Real time
platforms in which billions of users are connected examples of application using Big data are Spottily,
generate huge data which can be of any type such as Amazon Prime etc. Next widely used sector is
text, images, audio, video, gif, etc. Big data is highly Healthcare. This sector has access to huge sets of the
used in many applications because each and every data as they handle much number of patients and
field now generates tones of data which need to be their respective data.
processed at one point of time. Let us discuss the

CSEIT195149 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 299-302 ]
299
T. K. Anusuya, A. Preethi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 299-302

II. METHODS AND MATERIAL EXISTING PROCESS

LITERATURE SURVEY Big data is nothing but huge amount of data which
cannot be processed using traditional methods. The
A. Can Twitter save lives? A Broad scale study on data to be processed may bestructured, unstructured
Visual Social Media Analytics for Public Safety or semi structured. These data come from various
Theme: According to Dennis Thom et al. presents sources such as sensor data, transactional data,
that nowadays the social media is being used for cameras, microphones and lot more of resources. On
commercial purpose rather than safety. The paper these large data sets analytics is performed in order to
gives a clear understanding on the crisis intelligence obtain undiscovered insights. Based on facts,
field study via Twitter data during the flood on everyday terabytes of data are being generated and
German 2013. Scatter Blogs systems was introduced now it has come up to zeta data. All social media
to implement other techniques. The second phase platforms in which billions of users are connected
sketched out a system based on the feedback about generate huge data which can be of any type such as
Scatter Blogs and a comparison is made between both text, images, audio, video, gif, etc.
the systems. B. Big Data Analytics in Healthcare
Theme: Based on the field of healthcare Big Data PROPOSED METHODOLOGY
analytics and its benefits with the methodologies
used and its outcomes and the challenges faced are It has been observed that each paper has used
discussed here according to Wullianallur Raghupathi different techniques which have their own pros and
et al. C. Big Social Data Analytics for Public Health: corns. The techniques overall used are Scatter Blogs,
Face book Engagement and Performance Theme: SoLoMo, IBM Big Insights each used for a purpose
analysis of Facebook data using unsupervised learning such as analysis, storage, streaming real time data etc.
indicates recent tendency of user engagement in We have decided to take the positive features of all
public health has been increased as observed by the algorithms, tools and techniques used in the
Nadia Straton et al. D. SoLoMo analytics for Telco Big above discussed papers. The social Media analytics is
Data monetization Theme: The reference has the main resource to perform analysis. It is actually
observed that the world can be accessed using the the process of developing informative tools to collect
mobile internet. This has helped widely in Business. and summarize the data. The usage of social media
The algorithms and technologies used in this paper has been evolving more as the days pass by. It has
are used to find out the insights with the help of been used for many positive as well asnegative usage.
social, location and mobile data of the individual. E. It all depends upon the attitude of the users.
Impact of Demonetization on Indian Economy Facebook, YouTube and Twitter are in the first, third
Theme: India has the highest rate of currencies. 87% and tenth position on account of the usage. In general
of the currencies are rotating in the form of Rs.500 the analysis process comprises of three stages Capture,
and Rs.1000. These currencies become the source of Understand and Present. Topic modeling is also one
income either by working under organizations or of the important techniques being used in above
through illegal enterprise. This paper by Dr. Pratap papers. Hence Social media analytics is highly used in
Singh et al. presents the impacts after the process of the analysis that is performed on demonetization.
demonetization occurred in India.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


300
T. K. Anusuya, A. Preethi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 299-302

III. RESULTS AND DISCUSSION the API (Application Programming Interface) of the
respective platform. Sometimes it is requested to pay
in order to retrieve the data as they might keep their
resources secure. The next step processes it to
perform analysis on the retrieved data. This process
can also make use of R studio and flume platform to
store the data and perform analysis. The outcome is
derived in any of the formats such as Graph, line or
pie chart.

VI.REFERENCES

[1]. Dennis Thom, Robert Kruger, Thomas Ertl,


“Can Twitter Save Lives? A Broadscale Study on
Visual Social Media Analytics for Public
Safety”, IEEE Transactions on Visualization and
Computer Graphics, 2015.
IV. CONCLUSION [2]. Wullianallur Raghupathi and Viju Raghupathi,
“Big data Analytics in Healthcare: promise and
This paper presents about the introduction, potential”, Raghupathi and Raghupathi Health
characteristics, applications and limitations of Big Information Science and Systems, 2014.
data. Then about the introduction about social data [3]. Nadiya Straton, Kjeld Hansen, Raghava Rao
analytics which is the source for performing analysis. Mukkamala, Abid Hussain, TorMorten Grønli
The literature study includes the proposed work, Henning Langberg and Ravi Vatrapu, “Big
used tools and techniques in each respective paper. It Social Data Analytics for Public Health:
also contains the overall observations with the Facebook Engagement and Performance”, IEEE
futuristic solutions where it has been discussed about 18th International Conference on e-Health
the steps involved in analysis and the applications. Networking, Applications and Services
The conclusion and the reference for the above (Healthcom), 2016.
surveyed paper are also included. [4]. H. Cao, W.S. Dong, L.S. Liu, C.Y. Ma, W.H.
Qian, J.W. Shi, C.H. Tian, Y. Wang, D.
V. FUTURE ENHANCEMENT Konopnicki, M. Shmueli-Scheuer, D. Cohen, N.
Modani, H. Lamba, A. Dwivedi, A. A. Nanavati,
According the literature study that we did on the M. Kumar, “SoLoMo analytics for telco Big
paper, there were many techniques for analysis that Data monetization”, IBM J. RES. & DEV. VOL.
were implied on social data. This was especially to 58 NO. 5/6 Paper 9 September/November,
gather insights and make efficient decisions so that it 2014.
becomes useful for the users who wish to do research [5]. Dr. Partap Singh, Virender Singh, “Impact of
based on the problem. The first step process is to Demonetization on Indian Economy”, 3rd
collect data from the social networks such as Twitter, International Conference on Recent
Facebook. The data can be easily retrieved through Innovations in Science, Technology,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


301
T. K. Anusuya, A. Preethi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 299-302

Management and Environment, 18th


December, 2016
[6]. Weiguo Fan, Michel D. Gordan, “Unveiling the
Power of social Media Analytics”
[7]. Dr. Sunitha V Gangier, Ranganatha B,
“Demonetization and its impact on Social
Development”, INDIAN JOURNAL OF
APPLIED RESEARCH, January 2017

Cite this article as :

T. K. Anusuya, A. Preethi, "A Review on Big Data


Analytics Via Social Media", International Journal of
Scientific Research in Computer Science, Engineering
and Information Technology (IJSRCSEIT), ISSN :
2456-3307, Volume 5 Issue 1, pp. 299-302, January-
February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195149
Journal URL : http://ijsrcseit.com/CSEIT195149

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


302
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195147

Automatic Railway Gate Control System Using RFID with


High Speed Alerting System
M. Florence Dayana *1, R. Bhuvaneshwari 2
1 Head & Assistant Professor, Department of Computer Applications, Bon Secours College for Women,
Thanjavur, Tamil Nadu, India
2M. Sc., Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India

ABSTRACT

The main objective of this paper is to avoid the railway accidents happening at the level crossings. In India
railway is the major mode of the transportations and it is the cheapest way for travelling. So there are more
numbers of rail users. And it is not easy to stop railway anywhere to prevent accident or for any other reason.
Due to that there are major drawbacks of that. At present manned system is available at level crossings. Hence,
many accidents occur at such crossings, since there is no one to take care of the functioning of the railway gate
when a train approaches the crossing. The objective of this paper is to manage the control system of railway
gate using the microcontroller. This model includes IR sensors, RFID, LCD, buzzer, light led, motor driver and
microcontroller PIC16F877A. In the automatic railway gate control system, at the level crossing the arrival of
the train is detected by the IR sensor and RFID placed near to the gate. In case of RFID it detects only arrival of
train. Hence, the time for which it is closed is less compared to the manually operated gates and reduces the
human labour. As the entire system is automated errors occurring due to manual operation are prevented
because the accuracy of automated operation is more than the manned operation.
Keywords : Automatic Railway Gate, IR Sensors, RFID, PIC 16F877A, Motor Driver, Level Crossing.

I. INTRODUCTION the train by both sides. RFID is also used for the
detection of specific train only. RFID consisting of
We can’t take care of ours while in running by less tag and reader. Tag is given to the train and it will be
Railways being the cheapest mode of transportation read by that RFID reader only. When reader reads
are preferred over all the other means. But there are that tag which is placed on train then the gate will be
so many accidents happening at the manned control closed. When the last carriage of train passes having
level crossings. This is mainly due to the carelessness tag then the gate will be opened. LCD will display
in manual operations or lack of workers. This project information for the road users. This process Reduces
utilizes two pairs of IR sensor and RFID placed at the the traffic and time for which users have to for long
either sides of gate. Each pair of sensor consist of a time at level crossing.
transmitter and receiver, in our case light emitting
diode is used as a transmitter and a p-n junction
photodiode is used as receiver. It detects the arrival of

CSEIT195147 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 303-306 ]
303
M. Florence Dayana, R. Bhuvaneshwari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 303-306

Concept
At the present, India having world’s largest railway
network which having manned gate control system.
Over hundreds of railways running on track every
day. It is definitely impossible to stop the running
train at immediate in some critical situation or
emergency arises. Train accidents having serious
drawbacks in terms of loss of human life, injury,
damage to Railway property. This model uses
automatic control or operation of gate and reduces
the human labors. The concept of the model is to
control the railway gate using microcontroller
PIC16F877A. Figure 1. Block Diagram of System

Objective - To avoid an unwanted accident, Circuit Diagram


happening at the railway level crossing due to less It shows the circuit diagram of Automatic Railway
awareness of drivers and poor work of gate keeper Gate Control by using PIC 16F877A microcontroller.
and also time saves. This system includes the various components for the
operation of the system. This system includes two
II. METHODS AND MATERIAL components for the detection of the train i.e. IR
sensor and RFID. Initially IR emits the light. When
Block Diagram train cuts the IR sensor then it gives the input to the
microcontroller PIC16F877A, then microcontroller
Fig. 1 shows over all block diagrams for railway gate gives command to the gate driver so gate is closed
control system by using microcontroller PIC and buzzer will be ON. When the last carriage of the
16F877A. train cut the second IR sensor then gate is open and
buzzer is OFF. In case of RFID, the specific tag is
PIC1F877A performs some operations related to the given to the train. When that train is crosses the
system. IR and RFID detects the arrival the of train RFID, then that tag is read by the RFID reader. If the
and they gives the signal to the microcontroller. specific tag is read by train then and only
Accordingly microcontroller gives the command to microcontroller gives the command to the gate driver
the gate driver, LCD, buzzer. If train cut the IR then gate is closed and buzzer is ON. On the
sensor and RFID then microcontroller gives the departure of the train the gate is open and buzzer will
command to gate driver, it closed the gate and buzzer be OFF. Accordingly the red and green led is
is ON. After sometime, departure of train gate is operating. Also the LCD will display some message
opened and buzzer is OFF. Accordingly LCD will for the road users for informing the information
display the message to the road users. But in case of about the arrival and departure of the train.
RFID, it detects only the arrival of the train as
compared to the IR sensor. So RFID is the
advantageous than IR sensor for the detection of the
train.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


304
M. Florence Dayana, R. Bhuvaneshwari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 303-306

Algorithm large quantity. This system can contribute a lot of


• Step1: Start benefit either to the road users or to the railway
• Step2: Initialize timer, LED, interrupt management. Since the design is completely
• Step3: Display on LCD “WELCOME TO MET automated it can be used in remote villages where no
BKC IOE” station master or line man is present. The RFID
• Step4: Wait for 3sec placed at the either side of gate so they can detect the
• Step5: Clear LCD arrival and departure of train. Now a day’s automatic
• Step6: Display on LCD “AUTOMATIC system occupies each and every sector of applications
RAILWAY GATE CONTROL SYSTEM” as it is reliable and accurate.
• Step7: wait for 4sec
• Step8: if IR sensor1=1 or IR sensor2=1 or PROPOSED MODEL
RFID reader o/p=1 or RFID reader o/p=1
• Step9: if yes then display on LCD “Gate close
please wait”
• Step10: Relay1=ON
Relay2=OFF
Relay3=ON
Relay4=OFF
• Step11: Delay 1sec
• Step12: Red led on, green led off
• Step13: if IR sensor1=0 or IR sensor2=0 or
RFID reader o/p1=0 or RFID reader2=0
• Step14: Relay1=OFF
• Relay2=ON
• Relay3=OFF
• Relay4=ON
• Step15: Delay 1sec Figure 2. Proposed model of system
• Step16: Display on LCD “Gate open please go”
Step17: Stop. IV.REFERENCES

III.RESULTS AND DISCUSSION [1]. Hnin Ngwe Yee Pwint, ZawMyoTun,


HlaMyoTun, “Automatic Railway Gate Control
CONCLUSIONS System Using Microcontroller,” Volume 3,
Issue 5, May 2014.
Automatic gate control system using RFID is the [2]. Sayali R. More, Ruchira J. Raut, Rasika K.
efficient and easy way of reducing the accidents at Tandel,
the level crossings. This gives an secure model for the [3]. Snehal D. Yendhe, “Intelligent Railway
railway track system. Using RFID tag and reader, Crossing Gate Control with High Speed Anti-
microcontroller PIC16F877A detects only the arrival Collision Alerting System,” .
of the specific train so only at that time gate is open [4]. Krishna, Shakshi Yadav, Nidhi, “Automatic
and close. Due to that there is no any traffic jam and Railway Gate Control Using Microcontroller,”
wastage of time. This will reduce the accidents in

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


305
M. Florence Dayana, R. Bhuvaneshwari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 303-306

ISSN: 0974-6471 December 2013, Vol. 6, No.


(4).
[5]. R. BoopalaVenkatesh, K. Kamalanathan,
“Microcontroller Based Automatic Railway
Gate Control With Zero Accident,” Vol. 3,
Special Issue 2, March 2016.
[6]. Vikash Kumar, Prajit Paul, Nishantkumar,
Pratik kumar Sinha, Sumant Kumar Mahato,
“Automatic Railway Gate Controller with High
Speed Alerting System,” Vol. 4, Issue 5, May
2015.

Cite this article as :

M. Florence Dayana, R. Bhuvaneshwari, "Automatic


Railway Gate Control System Using RFID with High
Speed Alerting System", International Journal of
Scientific Research in Computer Science, Engineering
and Information Technology (IJSRCSEIT), ISSN :
2456-3307, Volume 5 Issue 1, pp. 303-306, January-
February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195147
Journal URL : http://ijsrcseit.com/CSEIT195147

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


306
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195141
Design of Robot for Scrambling on Tree and Clutching Fruit
Using Foundary Design
V. Manochithra1, S. Karthika2
1Head, Department of Information Technology, Bon Secours College for women, Thanjavur,
Tamil Nadu, India
2 M.Sc., (Computer Science), Bon Secours College for women, Thanjavur, Tamil Nadu, India

ABSTRACT

The advancement in nature has prompted the presentation of very effective organic systems. Emulating these
instruments offers gigantic possibilities for the change of our life and the apparatuses we utilize. Nowadays, the
robot is biggest mechanism that is used for each and every process as it has the gripper assembly designed uses a
flexible casing which grasps the tree trunk of almost all diameter within the predefined range, basically it uses
passive compliance. This mechanism can be used for climbing trees which are almost height like coconut trees
and poles. The main drawback is being slow and there is a possibility of damaging the tree trunk. Hence this
system propose a new tree climbing mechanism for multi tree climbing robot that can be used to climb trees
that are almost straight and height.
Keywords : Robot, Grippers, Climbing trees, Clutching

I. INTRODUCTION mechanism that can be used to climb trees that are


almost straight, like coconut trees and poles with
Most commonly used design for tree climbing is inch greater agility.
worm design. These models are very slow. The main
body of this type of robot is divided into two parts The main locomotion of the robot is caused due to
and each part has a gripper. These models using inch the two motors which facilitates simultaneous
worm mechanism are continuous in their motion and motion of two consecutive links. This robot grasps
can make over in complex tree environments the tree with the help of grippers which is acted by a
involving multiple branches. Next common type of spring. When the motor is ON it unclamps and
design is wheeled robot where in instead of grippers clamps while the motor is OFF by spring action while
wheels are used for climbing up. The robot engulfs harnessing the spring energy for gripping. When one
the tree and locks against it and as the motor rotates, gripper grasps the tree, the whole body makes a
it moves up. The sizes of these robots are comparable revolution such that the next gripper comes up and
to the diameter of tree. These models are suitable for grips while the other un grasps and the whole process
trees with straight and plain trunks. The main continues. Attachments like pesticide sprayer, weed
drawback is it being slow and there is a possibility of remover or some harvesting mechanism can be
them damaging the tree trunk. Hence most of the mounted on the Robot. With these attachments this
models are slow, less agile, have less load carrying robot can act as harvesting or maintenance
capacity. As a result, we propose a tree climbing equipment by the farmers.

CSEIT195141 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 307-311 ]
307
V. Manochithra, S. Karthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 307-311

These models using inch worm mechanism are pesticide sprayer, weed remover or some harvesting
continuous in their motion and can maneuver in mechanism can be mounted on the Robot. With
complex tree environments involving multiple these attachments this robot can act as harvesting or
branches. Next common type of design is wheeled maintenance equipment by the farmers.
robot where in instead of grippers wheels are used for
climbing up. The robot engulfs the tree and locks Coconut picking is a tough, arduous work. The
against it and as the motor rotates, it moves up. The farmers in popular coconut cultivating countries are
sizes of these robots are comparable to the diameter facing severe shortage of skilled workers for
of tree. These models are suitable for trees with harvesting. In recent years, the number of coconut
straight and plain trunks. tree climbers has declined sharply, as those men who
do this rigorous job of climbing 30 to 80 feet high
The main drawback is it being slow and there is a above the ground, bare handed or with the aid of a
possibility of them damaging the tree trunk. Hence rope, have broken away from their traditional role to
most of the models are slow, less agile, have less load take-up better paid jobs. Since the Green Revolution
carrying capacity. As a result, we propose a tree of the late 1960s, the harvesting of coconuts today
climbing mechanism that can be used to climb trees has little changed from how it used to be done for
that are almost straight, like coconut trees and poles generations. The job of coconut picking, which is
with greater agility. The main locomotion of the perceived as risky and unglamorous, does not attract
robot is caused due to the two motors which youngsters either, even though the remuneration is
facilitates simultaneous motion of two consecutive fairly good in local standards. Though there exist few
links. tree climbing machines today, by any measure,
climbing for coconuts is hard work. Not only does a
climber have to scale the tree and reach out among
the fronds and cut the coconut, but they also have to
instinctively learn which nuts are ripe enough to be
harvested. Besides, those few climbers who still
remain in the business of coconut picking, claims
higher wages.

III. PROBLEM STATEMENT


Fig.1. System Model

In traditional harvesting method on reaching the top


II. LITERATURE SURVEY
of a coconut tree, the climber taps the nut in the
lowermost bunch with its harvesting knife to test its
This robot grasps the tree with the help of grippers
maturity. If he is satisfied, he cuts the bunch at the
which is acted by a spring. When the motor is ON it
base of the stalk when it drops down to the ground.
un-clamps and clamps while the motor is OFF by
The climber also cleans the crown and removes the
spring action while harnessing the spring energy for
dry leaves, sheaths and spathes. In some places where
gripping. When one gripper grasps the tree, the
the trees are not tall, cutting the coconut bunches
whole body makes a revolution such that the next
with a knife, attached to a long bamboo pole, does
gripper comes up and grips while the other ungrasps
the harvesting. It seems quite simple to climb the
and the whole process continues. Attachments like
coconut palm in order to stay beneath the leaf crown

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


308
V. Manochithra, S. Karthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 307-311

to harvest fruits and cut old leaves. But it proves


much more challenging to climb up the coconut leaf
crown in order to reach the young inflorescences for
making controlled pollinations with bagging for
research purposes, and for harvesting the toddy (the
sweet sap from coconut inflorescences), which serves
to produce sugar, vinegar, wine and spirits. This
paved the way to the proposed robotic harvesting
method, where it can peep and reach out to any
remote nook and corner at the top of the tree, of the Fig.2. Robot Climbing Model
operator’s choice.

V. LOCOMOTION OF ROBOT
IV. METHODOLOGY
There are three types of motion. These are classified
In traditional harvesting method on reaching the top as continuous, discrete and serpentine. As we get
of a coconut tree, the climber taps the nut in the closer to continuum type, energy consumption is
lowermost bunch with its harvesting knife to test its reduced and speed is increased. But it is difficult to
maturity. If he is satisfied, he cuts the bunch at the achieve continuous motion, so discrete form is
base of the stalk when it drops down to the ground. followed in the construction.
The climber also cleans the crown and removes the
dry leaves, sheaths and spathes. In some places where First, the bot is attached to the tree trunk with the
the trees are not tall, cutting the coconut bunches help of grippers which grips on to the tree trunk and
with a knife, attached to a long bamboo pole, does provides adequate friction such that it does not slip.
the harvesting. The motors used for climbing overcome the moments
caused by its weight.
❖ Ease of use - They wanted use of the robot to be
something that could be easily learned, so it The movement takes place as follows:
would not slow down ALB inspection rate.
❖ Work safe - They wanted the system to have fail 1. First the arm with first motor will hold the tree.
safes implemented to ensure no potential harm 2. Then the link with second motor will release the
to USDA operators in the field. grip and the whole setup will rotate about motor
❖ Minimal damage to tree - They did not want the one as shown in figure.
robot to inflict any serious harm on the tree it 3. After half a rotation, the arm with second motor
was climbing. will grip the tree and the process continues. Once
❖ Navigation around branches - They needed to the robot reaches the topmost point of the tree
robot to be able to find its way around branches, trunk, the whole body is gripped on to the trunk
so inspections could be complete. without any further movement, i.e., the body is
grounded, thereafter miscellaneous action like
harvesting or maintenance work can be
performed.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


309
V. Manochithra, S. Karthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 307-311

device, hence conventional climbing robots cannot


be used for tree climbing applications.

Fig.3. Robot Climbing Model


Graph.1. Efficiency Analysis

This mechanism can be achieved by using a push


plate in the gripper which will be connected to the
shaft of linear motor, when the linear motor actuates
the push plate pushes the claws and unclamps from
the tree.

VI. RESULT AND DISCUSSIONS Graph.2. Robot View

The evolution in nature has led to the introduction of VII. CONCLUSION


highly efficient biological mechanisms. Imitating
these mechanisms offers enormous potentials for the In this work a tree climbing mechanism for coconut
improvement of our life and the tools we use. trees has been proposed this mechanism can also be
used to climb structures and other trees which are
We have devised this mechanism inspired from the almost straight like palm trees, poles etc. The
locomotion of a worm called Stomata pod which uses mechanism used in this robot is a bio inspired
its whole body as distributed foot facilitating it to mechanism which is derived from an organism called
move on any terrain and have high agility on a Stomata pod. This mechanism appears to be more
regular tree environment. The gripper assembly agile compared to other available mechanisms but the
designed uses a flexible casing which grasps the tree main limitation is that it can be used only for a
trunk of almost all diameter within the predefined smaller category of trees.
range, basically it uses passive compliance.
A prototype of robot was made and the efficiency of
This mechanism can be used for climbing trees which the slinky type gripper was tested on few trees. Also,
are almost straight like coconut trees and poles. software simulation was done for this mechanism.
Hence this paper presents a new climbing mechanism Our future works includes the addition of
for coconut tree climbing robot. It also requires agricultural equipment’s like harvester sprayer etc.
greater agility and high maneuverability to be used as also to develop upon the gripping materials as
a product. Also the bark of some trees may not be discussed.
strong enough to bear the weight of the climbing

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


310
V. Manochithra, S. Karthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 307-311

VIII. REFERENCES [8]. Appeltants W and Ilsbroux W (1995)


Automatische sturing van een plukwagen
[1]. Juan Carlos Grieco,Manueul Prieto,Manuel (Autonomous guidance for a harvester). MA
Armada and Pablo Gonzalez de Santos, “A Six Thesis, Katholieke Hogeschool Limburg,
Legged Climbing Robot For High Payloads” Belgium
Proc. of the 1998 IEEE International [9]. Bulanon DM, Kataoka T, Okamoto H, Hata S
Conference on Control Applications ,Trieste, (2005) Feedback control of manipulator using
Italy,pp.446- 450,1998 machine vision for robotic apple harvesting. In
[2]. Yili Fu, Zhihai Li, Hejin Yang and Shuguo Proceedings of ASAE, Paper No. 053114,
Wang.” Development of a Wall Climbing Tampa, USA
Robotwith Wheel-Leg Hybrid Locomotion [10]. Bulanon DM, Kataoka T, Ukamoto H, Hata S
Mechanism”Proceedings of the 2007 IEEE (2004) Development of a realtime machine
International Conference on Robotics and vision system for the apple harvesting robot. In
Biomimetics, Sanya, China,pp.1879-1881,2007 SICE Annual Conference in Sapporo, pages
[3]. Michael P. Murphyand andMetin Sitti.” 595-598, Hokkaido Inst. of Techn., Japan
Waalbot: An Agile Small-Scale WallClimbing [11]. Hutchinson S, Hager G, Corke P (1996) a
RobotUtilizing Dry Elastomer Adhesives”, tutorial on visual servo controls. IEEE Trans. on
IEEE/ASME TRANSACTIONS ON Robotics and Automation, 12:651-670
MECHATRONICS, VOL. 12, NO. 3, pp.330- [12]. Jimenez A, Ceres R, Pons J (2000) a survey of
338,2007. computer vision methods for locating fruits on
[4]. Ozgur Unver, Ali Uneri, Alper Aydemir and trees. Transactions of the ASAE, 43(6):1911-
Metin Sitti,” Geckobot: A Gecko Inspired 1920
Climbing Robot Using Elastomer Adhesives.” [13]. Katypitiya J, Eaton R, Cole A, Meyer C, Rodnay
Proceedings of the 2006 IEEE International G (2005) Automation of an agricultural tractor
Conference on Robotics and for fruit picking. In IEEE Int. Conf. on Robotics
AutomationOrlando, Florida, pp.2329-2335, and Automation, pages 3212-3217, Barcelona,
2006. Spain
[5]. Tin Lun Lam and Yangsheng Xu, A Flexible [14]. Mateus D, Avi˜na G, Devy M (2005) Robot
Tree Climbing Robot: Treebot - Design and visual navigation in semi-structured outdoor
Implementation,’ 2011 IEEE International environments. In IEEE Int. Conf. on Robotics
Conference on Robotics and Automation and Automation, pages 4702-4707, Barcelona,
Shanghai International Conference Center, Spain
Shanghai, China,pp.5849-5854,2011 [15]. MVTec Software GmbH, M¨unchen, Germany
[6]. Y. Kushihashi, et al., “Development of Tree (2006) Halcon 7.1, Quick Guide, 2b edition
Climbing and PruningRobot,Woody-1- Cite this article as : V. Manochithra, S. Karthika, "Design
Simplification of Control using adjust Function of Robot for Scrambling on Tree and Clutching Fruit Using
of Grasping Power” .Proceedings of JSME Foundary Design", International Journal of Scientific
Research in Computer Science, Engineering and
Conference onRobotics and Mechatronics, pp.
Information Technology (IJSRCSEIT), ISSN : 2456-3307,
1A1- E08, 2006.
Volume 5 Issue 1, pp. 307-311, January-February 2019.
[7]. H. Kawasaki, et al., “Novel climbing method of
Available at doi : https://doi.org/10.32628/CSEIT195141
pruning robot”, Proceedings of the SICE Journal URL : http://ijsrcseit.com/CSEIT195141
Annual Conference, Tokyo, pp. 160-163,2008.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


311
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195142
Spinal Muscular Atrophy Type B Awareness and Symptoms
Prediction Using 3D Segmentation Process in MRI Images
V. Manochithra, G. Sumithra
1Head, Department of Information Technology, Bon Secours College for women, Thanjavur,
Tamil Nadu, India
2 M.Sc., (Computer Science), Bon Secours College for women, Thanjavur, Tamil Nadu, India

ABSTRACT

Spinal muscular atrophy (SMA) describes a group of disorders associated with spinal motor neuron loss. In this
review we provide an update regarding the most common form of SMA, proximal or 5q SMA, and discuss the
contemporary approach to diagnosis and treatment. Electromyography and muscle biopsy features of
denervation were once the basis for diagnosis, but molecular testing for homozygous deletion or mutation of
the SMN1 gene allows efficient and specific diagnosis. In combination with loss of SMN1, patients retain
variable numbers of copies of a second similar gene, SMN2, which produce reduced levels of the survival motor
neuron (SMN) protein that are insufficient for normal motor neuron function. Despite the fact that the
understanding of how ubiquitous reduction of SMN protein leads to motor neuron loss remains incomplete,
several promising therapeutics are now being tested in early phase clinical trials. This proposed model
investigates the symptoms and scans readings from the initial MRI scan images of babies with mutation
progress and SMN proteins formation benchmark values for this particular disorder SMA and further this
segmented parameters are acquitted into the K-means clustering technique that predict the report with the
disorder symptoms with MSE (mean square error) values that helps the babies in future to take prevention
measures to overcome this problem.
Keywords : Spinal Muscular, K-Means Clustering, Neuron, Disorder Symptoms, MRI Images.

I. INTRODUCTION legs first and then the arms and, in due time, they
will become quadriplegic and even more disabilities
Spinal Muscular Atrophy (SMA) is a progressive can follow including speech impairment. Assistive
neuromuscular disorder. Usually, this condition is Technology support for people with such disabilities
considered genetically induced with no known cure often requires identification of the best residual
to date. Children are born with the condition and muscular function so that this can be utilized as a
develop muscular weakness progressively as they means of voluntary control.
grow. The weakness ultimately encompasses the
whole muscular function rendering the limbs Spinal muscular atrophy (SMA) refers to a group of
dysfunctional or paralyzed. Many children with SMA, inherited diseases of the motor nerves that cause
if they do not have the weakness from the beginning, muscle weakness and atrophy (wasting). The motor
will start having the disease manifesting itself on the nerves arise from the spinal cord and control the

CSEIT195142 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 312-318 ]
312
V. Manochithra, G. Sumithra Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 312-318

muscles that are used for activities such as breathing, muscles at disease duration of >10 years compared to
crawling, walking, head and neck control, and disease duration of ≤10 years. Longitudinal follow-up
swallowing. SMA is a rare disorder affecting of four patients with two MRI studies eight years
approximately 1 out of every 10,000 individuals apart showed evidence of slow progression. Clinical-
worldwide. radiological comparison revealed a good correlation
for most of the muscles. This muscle MRI study in a
SMA affects muscles throughout the body. In the relatively large number of patients showed a specific
most common types, weakness in the legs is generally pattern of involvement in SMA 3b and revealed that
greater than in the arms. Sometimes feeding, muscle MRI can be an objective and powerful tool in
swallowing, and respiratory function (e.g., breathing, showing disease progression with implications for
coughing, and clearing secretions) can be affected. clinical trials.
When the muscles used for breathing and coughing
are affected and weakened, this can lead to an Clinical Features Of SMA
increased risk for pneumonia and other respiratory
infections, as well as breathing difficulty during sleep. Since only α motor neurons are lost progressively,
The brain’s cognitive functions and the ability to feel only motor function is compromised and sensory
objects and pain are not affected. People with SMA neurons are unaffected. This loss of function leads to
are generally grouped into one of four types (I, II, III, weakness and to progressive symmetrical atrophy of
IV) based on their highest level of motor function or the proximal voluntary muscles of the legs, arms and,
ability. sometimes, the trunk, as the disease progresses. A
number of unusual clinical features are observed in
There are only a few reported muscle magnetic SMA. One of these is the distribution of muscle
resonance imaging (MRI) studies of genetically- weakness, which is more compatible with a
confirmed spinal muscular atrophy (SMA). The aim myopathic disorder than with a neurogenic
of this study was to search for the presence of disorder.12 Proximal muscles are more involved than
selective muscle involvement and of disease distal muscles, legs are more affected than arms and
progression by MRI in SMA 3b. MRI of proximal arms are more affected than the face anddiaphragm.8,
upper and lower extremities was performed on a 1.5- 12 In other words, muscle weakness and atrophy does
Tesla MR scanner using conventional T1 and T2 not have a homogeneous distribution. The severity of
weighted axial images in twenty-two patients with muscle weakness is related to age at onset and
genetically-confirmed SMA. Patients had detailed children with the most severe form of the disease
clinical examinations within one month of the (Type I SMA) can appear normal at birth, but present
radiological investigation. MRI revealed a selective muscle weakness few months later. Furthermore, the
involvement of muscles. Triceps, iliopsoas and clinical course followed by SMA patients.
quadriceps femoris were the most affected muscles.
In contrast, biceps brachii, deltoid and gluteus Who survive beyond childhood demonstrates that
maximus were well-preserved. Gracilis, short head of loss of muscle strength is normally most evident at
biceps femoris, semimembranosus, and Sartorius and diseases onset and that, after this, residual muscle
adductor long us were also relatively spared. Slow, power can remain stable for months or years.
but definite progression with a predictable sequence
and magnitude was documented. Median MRI grade II. LITERATURE SURVEY
was higher (worse) usually by one grade in the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


313
V. Manochithra, G. Sumithra Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 312-318

One of the major challenges facing researchers poses an extra challenge precisely due to their
engaged in the study of neurodegenerative and complicated genetics. An attractive alternative would
neuron development diseases is to explain why therefore be to address these issues in diseases with a
defects in ubiquitously expressed proteins have such simple Mendelian inheritance pattern. Spinal
a selective effect on the nervous system and its muscular atrophy provides researchers studying
constituent cell types. Explaining the molecular neurodegenerative diseases with this opportunity.
mechanisms underlying the disease phenotypes has This review highlights why (1) SMA might be
been hampered by their multifactorial nature and considered a prototypical neurodegenerative disease
high incidence of sporadic cases. Alzheimer’s disease, in which to address certain general questions facing
Parkinson’s disease, and amyotrophic lateral sclerosis the field, and (2) why a thorough understanding of
(ALS) are among the most common this disorder might shed light on the mechanisms
neurodegenerative diseases in humans and good that have a specific effect on the development, health,
examples of disorders that exemplify these challenges. survival, and causes of motor neuron degeneration,
Two less common neurological diseases which are which ultimately leads to a disease phenotype.
monogenic in origin but are the result of defects in Proximal spinal muscular atrophy, commonly
ubiquitously expressed proteins involved in referred to as SMA, is a common autosomal recessive
housekeeping functions are Rett syndrome and 5q neuromuscular disease that affects the anterior horn
spinal muscular atrophy (SMA). cells of the spinal cord, resulting in atrophy of the
proximal muscles of the limbs and trunk. There are
The former is characterized by mutations in the gene numerous other forms of spinal muscular atrophy
encoding methyl – CpG binding protein 2 (MeCP2), which share certain characteristics with proximal
which likely regulates gene expression and chromatin SMA; however, they are genetically distinct and
structure; the latter by a deficit of the survival of often affect different subsets of neurons and muscle.
motor neurons (SMN) protein, whose best They include autosomal dominant forms of the
characterized function is in snRNP biogenesis and disease (Sambuughin et al., 1998; Van der Vleuten et
pre-mRNA splicing. Yet, despite the apparent need al., 1998), X-linked forms (La Spada et al., 1991;
for these proteins in all tissues, Rett syndrome, a Kobayashi et al., 1995), recessive forms that affect the
prototypical neurodevelopment disorder, is distal muscles (Viollet et al., 2002), and a severe form
characterized by a specific defect in maintaining of SMA (SMARD) with respiratory distress
proper function of post mitotic neurons in the (Grohmann et al., 2001) (Table 1).
forebrain, hippocampus, and brainstem, while SMA
has a particularly profound effect on lower motor After cystic fibrosis, SMA is the most common
neurons. autosomal recessive disorder in humans, with a
carrier frequency of approximately 1 in 35 and
This begs the question: do the proteins implicated in therefore an incidence of 1 in 6000 in the human
these diseases have functions in neurons other than population. It is also the most common genetic cause
their proposed housekeeping roles? Do they modulate of infant mortality. Despite the high incidence of the
downstream pathways specific to the tissues affected? disease in the human population, SMA has gained
Alternatively, are neuronal cells simply more relatively little attention among researchers studying
susceptible than other tissues to genetic, biochemical, neurodegenerative diseases. However, it has a
and environmental insults? Answering these fascinating biology, which includes two major players:
questions using multifactorial diseases as paradigms the highly homologous SMN1 and SMN2 genes.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


314
V. Manochithra, G. Sumithra Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 312-318

some labels cannot be adjacent to each other, then a


A splicing defect in SMN2 is a key factor in causing higher penalty is applied for the neighbouring pixels
the disease phenotype, while the SMN protein, that are assigned those labels for the given
expressed by both genes, is very likely neighbourhood system. Finally, the 3-D graph cut
multifunctional. However, low levels of the protein structure is formed and segmentation is performed.
have a particularly detrimental effect on one tissue
type, the lower motor neurons. One of SMN’s
functions is essential to cell survival. These
characteristics, coupled with a relatively simple
Mendelian inheritance pattern, make a compelling
case that study of this disorder in more detail will
provide a prototype that might shed considerable
light on motor neuron biology and disease. Such
studies may eventually lead to a better understanding
of SMA and other similar diseases and may accelerate
progress toward rational therapeutics.

III. METHODOLOGY

Magnetic Resonance Imaging is a fast and high


resolution technique that is commonly used to
investigate the anatomy and physiology of the body
in health and disease. High resolution imaging of the
important heart elements has become possible
through Multi-slice MRI technology. There has been
a very large body of studies using MRI scans to
segment body parts for research purposes and also to
increase the efficiency of the clinician’s work.

There are three major parts of the proposed approach


to segment 3-D MRI SMA Data with minimum user
interaction.

In the first step, a pre-processing operation is


employed to remove noise from individual slices.

Second, the user is required to provide soft


constraints to incorporate information about spatial
distance of regions to a reference point.

Fig.1. Work Flow


In the third step, the smoothness term can be
adjusted based on the input data. For example, if

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


315
V. Manochithra, G. Sumithra Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 312-318

employing new drugs and drugs that are already


STEPS INVOLVED: well-known. It cannot be expected that the
pharmacological treatments currently being studied
⚫ Respondents were invited by Cure SMA to can recover motor neurons or muscle cells that have
participate in the survey via email containing a already been lost to atrophy.
link to the survey.
Rather, the objective is to retard the progress of the
⚫ Responses were collected, tallied and analyzed
disease and improve residual muscle function. Sadly,
by Silicon Valley Research Group.
paralysis can be halted, but it cannot be reversed.
⚫ No PHI was collected from survey respondents However, through a combination of medical care and
through the survey or by any other means. rehabilitation, many patients with SMA can have
fulfilling and productive lives and often have normal
⚫ The survey was submitted to the IRB for life expectancy.
approval prior to data collection and is IRB
approved and compliant. Finally, there is one other point it is important to
emphasize, which is that since SMA is a recessive
⚫ The total sample size for the study was 298
genetic disorder, genetic counselling is an essential
completed responses, yielding a margin of error
component forthe families of these patients. Through
of 5.68%.
genetic counselling,
⚫ A possible explanation for the consistency in
most tolerable and least tolerable risks across all
treatments is that SMA sufferers and their
caregivers consider any of the benefits of
treatments presented as equally important, i.e.
they are not trading off risks for different
treatments.

⚫ Survey respondents also appeared to weigh risks


against their probabilities of occurrence.
Consistently, high probabilities of occurrence
made a risk less tolerable. Graph.1. Efficiency Analysis

IV. RESULT AND DISCUSSIONS

Both medical follow-up and palliative care are


important throughout SMA patients’ entire lives.
This care includes respiratory and nutritional support,
and orthopaedic and physiotherapeutic care to avoid
postural disorders.

In addition to this we can cite pharmacological


treatments that are still being studied, both

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


316
V. Manochithra, G. Sumithra Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 312-318

Graph.2. Diagnosed SMA of severity of spinal muscular atrophy. Am J


V. CONCLUSION Hum Genet. 2002;70:358‑68.
[6]. Chang JG, Hsieh-Li HM, Jong YJ, Wang NM,
It is obvious that significant challenges remain Tsai CH, Li H. Treatment of spinal muscular
toward an eventual treatment for SMA. These stem atrophy by sodium butyrate. Proc Natl AcadSci
from the many unanswered questions this review has U S A. 2001;98:9808-13.
raised. Readers may come away thinking that SMA [7]. Campbell L, Potter A, Ignatius J, Dubowitz V,
research is still in its infancy. However, it should be Davies K. Genomic variation and gene
noted that the tools to answer these questions, i.e., an conversion in spinal muscular atrophy:
excellent set of animal models and a basic implications for disease process and clinical
understanding of the disease, already exist. It should phenotype. Am JHum Genet. 1997;61:40-50.
also be reiterated that there are a number of [8]. Wang CH, Finkel RS, Bertini ES, Schroth M,
compelling reasons investigators outside the field Simonds A, Wong B,et al. Consensus statement
may find problems in SMA an attractive challenge to for standard of care in spinal muscular atrophy.
take on; for example, SMA is a neurodegenerative J Child Neurol. 2007; 22:1027-49.
disease in which a defect in a ubiquitously expressed [9]. Oskoui M, Levy G, Garland CJ, Gray JM,
protein affects a very specific tissue type. The O’Hagen J, De Vivo DC,et al. The changing
genetics of the disease are also relatively simple: SMA natural history of spinal muscular atrophy type
is autosomal recessive and involves a loss of function 1. Neurology. 2007;69:1931-6.
in an essential protein in certain cells. In essence, [10]. Wirth B, Brichta L, Hahnen E. Spinal muscular
SMA results from an insufficient amount of the SMN atrophy: from geneto therapy. Semin Pediatr
protein. Neurol. 2006;13:121-31. Review.
[11]. Zerres K, Rudnik-Schoneborn S. Natural
VI. REFERENCES history in proximal spinal muscular atrophy.
Clinical analysis of 445 patients and suggestions
[1]. Araujo AP, Ramos VG, Cabello PH. for a modification of existing classifications.
Dificuldades diagnósticas naatrofia muscular ArchNeurol. 1995; 52:518-23.
espinhal. Arq Neuropsiquiatr. 2005;63:145-9. [12]. Sumner CJ. Molecular mechanisms of spinal
[2]. Prior TW. Spinal muscular atrophy diagnostics. muscular atrophy. JChild Neurol. 2007;22:979-
J Child Neurol2007; 22:952-6. Review. 89.
[3]. Russman BS. Spinal muscular atrophy: clinical [13]. Swoboda KJ, Kissel JT, Crawford TO, Bromberg
classifications and disease heterogeneity. J MB, Acsadi G,D’Anjou G, et al. Perspectives on
Child Neurol. 2007; 22:946-51. clinical trials in spinal muscular atrophy. J
[4]. Shanmugarajan S, Swoboda KJ, Iannaccone ST, Child Neurol. 2007;22:957-66.
Ries WL, Maria BL,Reddy SV. Congenital bone [14]. Heier CR, Gogliotti RG, DiDonato CJ. SMN
fractures in spinal muscular atrophy: functional transcript stability: couldmodulation of
role for SMN protein in bone remodelling. J messenger RNA degradation provide a novel
Child Neurol.2007; 22:967-73. therapy for spinal muscular atrophy? J Child
[5]. Feldkotter M, Schwarzer V, Wirth R, Wienker Neurol. 2007;22:1013-8.
TF, Wirth B. Quantitative analyses of SMN1 [15]. Meldrum C, Scott C, Swoboda KJ. Spinal
and SMN2 based on real-time light Cycler PCR: muscular atrophy geneticcounseling access and
fast and reliable carrier testing and prediction

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


317
V. Manochithra, G. Sumithra Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 312-318

genetic knowledge: parents’ perspectives. J [24]. Oskoui M, Kaufmann P. Spinal


Child Neurol. 2007; 22:1019-26. muscularatrophy.Neurotherapeutics.2008;5:499
[16]. Lefebvre S, Bürglen L, Reboullet S, Clermont O, -506. Review.
Burlet P, ViolletL, et al. Identification and
characterization of a spinal muscular atrophy-
determining gene. Cell. 1995;80:155-65. Cite this article as :
[17]. Sumner CJ, Huynh TN, Markowitz JA, Perhac
JS, Hill B, CoovetDD, et al. Valproic acid V. Manochithra, G. Sumithra , "Spinal Muscular
increases SMN levels in spinal muscular Atrophy Type B Awareness And Symptoms
atrophy patient cells. Ann Neurol. 2003;54:647- Prediction Using 3d Segmentation Process In Mri
54. Images", International Journal of Scientific Research
[18]. Burghes AH, Beattie CE. Spinal muscular in Computer Science, Engineering and Information
atrophy: why do lowlevels of survival motor Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
neuron protein make motor neurons sick? Nat 5 Issue 1, pp. 312-318, January-February 2019.
Rev Neurosis. 2009;10:597-609. Available at doi :
[19]. Monani UR, Sendtner M, Coovert DD, Parsons https://doi.org/10.32628/CSEIT195142
DW, AndreassiC, Le TT, et al. The human Journal URL : http://ijsrcseit.com/CSEIT195142
centromeric survival motor neuron gene
(SMN2) rescues embryonic lethality in SMN (-
/-) mice and results in a mouse with spinal
muscular atrophy. Hum Mol Genet.2000;9:333-
9.
[20]. Burghes AH. When is a deletion not a deletion?
When it is converted. Am J Hum Genet.
1997;61:9-15. Review.
[21]. McAndrew PE, Parsons DW, Simard LR,
Rochette C, Ray PN, MendellJR, et al.
Identification of proximal spinal muscular
atrophy carriers and patients by analysis of
SMNT and SMNC gene copy number. Am J
Hum Genet. 1997; 60:1411-22.
[22]. Coovert DD, Le TT, McAndrew PE,
Strasswimmer J, Crawford TO,Mendell JR, et al.
The survival motor neuron protein in spinal
muscular atrophy. Hum Mol Genet.
1997;6:1205-14.
[23]. McWhorter ML, Monani UR, Burghes AH,
Beattie CE. Knockdown of the survival motor
neuron (SMN) protein in zebra fish causes
defects in motor axon outgrowth and path
finding. J Cell Biol.2003;162:919-31.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


318
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1183818
A Survey of Telemedicine Services using Datamining
T. K. Anusuya1, P. Maharajothi 2
Assistant Professor, Department of Computer Science, Bon Secours College for Women, Thanjavur, Tamil
1

Nadu, India
2M.Sc Computer Science, Department of Computer Science, Bon Secours College for Women, Thanjavur, Tamil
Nadu, India

ABSTRACT

The transfer of the medical care services to the patient, rather than the transport of the patient to the medical
services providers is aim of the project. This is achieved by using web-based applications including Modern
Medical Informatics Services which is easier, faster and less expensive. The required system implements the
suitable informatics and electronics solutions efficiently for the Tele-medicine care. We proposed an approach
to manage different multimedia medical databases in the telemedicine system. In order to be efficiently and
effectively manage, search, and display database information, we define an information package for both of
doctor and patient as a concise data set of their medical information from each visit. The methodology for
accessing various types of medical records will be provided, also we will design two web-based interfaces, high-
quality data and display for many medical service purposes.
Keywords: Datamining, Telemedicine, Medical Services, Databases.

I. INTRODUCTION can find that the integrated telemedicine system


consists of the following five subsystems:
There are shortages of medical resources in rural
areas or geographically isolated regions, so many 1) Acquisition subsystem;
physicians may be reluctant to serve in these areas. 2) Viewing subsystem;
Therefore, people who live there will receive lower 3) Teleconferencing subsystem;
medical care than those who live in urban areas. 4) Communication subsystem;
There is an important need to develop a telemedicine 5) Database management subsystem.
system to improve the quality of medical services
there and provide more educational opportunities to The first subsystem is the acquisition subsystem
the physicians in these areas [1] [4].Telemedicine can which collects multimedia information [12] then
be defined as the providing of medical services over a converts it to a standard format (e.g., DICOM 3.0
distance. The Archiving and Communication System [13]). The second one isthe viewing subsystem which
(PACS) will be used in the telemedicine process as displays and manipulates the images and other
this service requires patient history, medical images, medical information [14] [15]. The third one is the
and related information. By using PACS [5] [11], we teleconferencing subsystem which allows face-to-
face interactive conference between physicians in

CSEIT1183818 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 319-321 ]
319
T. K. Anusuya, P. Maharajothi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 319-321

rural areas and medical centers [16][18],this web management services, namely—
database
subsystem is not included in a PACS. fragmentation, data distribution, websites clustering,
distributed caching, and database scalability.
II. LITERATURE SURVEY
IV. PROPOSED METHODOLOGY
Recently, many researchers have focused on
designing web medical database management systems Our approach integrates three enhanced computing
that satisfy certain performance levels. Such services techniques namely, database fragmentation,
performance is evaluated by measuring the amount of network sites clustering and fragments allocation.
relevant and irrelevant data accessed and the amount We propose an estimation model to compute
of transferred medical data during transactions’ communications cost which helps in finding cost-
processing time. The medication reminder effective data allocation solutions. We perform both
synchronization system provides a patient with external and internal evaluation of our integrated
medications as prescribed by medical staff. In approach.
addition, medical staff can remotely send messages to
the system in order to change the medication In our proposed system we develop a fragmentation
schedules or device configuration settings embedded computing service technique by splitting
in the medication reminder. The proposed system telemedicine database relations into small disjoint
supports the OMA (open mobile alliance) DS (data fragments. This technique generates the minimum
synchronization) protocol, which was originally number of disjoint fragments that would be allocated
proposed as a DS standard protocol that synchronizes to the web servers in the data distribution phase. This
data between mobile devices and a central server. In in turn reduces the data transferred and accessed
this study, the OMA DS protocol is redefined and through different websites and accordingly reduces
extended to transmit the patient’ s medication status the communications cost.
data and the device configuration data.
V. IMPLEMENTATION RESULTS
III. EXISTING PROCESS

Recently, many researchers have focused on


designing web medical database management systems
that satisfy certain performance levels. Such
performance is evaluated by measuring the amount of
relevant and irrelevant data accessed and the amount
of transferred medical data during transactions’
processing time.

Several techniques have been proposed in order to


improve telemedicine database performance, • Doctor Registration
optimize medical data distribution, and control • Patient Registration
medical data proliferation. These techniques believed • Telemedicine services Activation
that high performance for such systems can be • Doctor Manage telemedicine services & their
achieved by improving at least one of the database setting

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


320
T. K. Anusuya, P. Maharajothi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 319-321

• Doctor Manage booking requests [3]. R. K. Taira, B. K. Stewart, a nd U. Sinha, PACS


• Patient add basic information & medical history database architecture and design, Comput.
• Patient makes booking &reviews his booking list Med.
[4]. Imag. Graph., vol. 15, no. 3, pp. 171-176, 1991.
VI. CONCLUSION [5]. S. Badaoui, V. Chameroy, andd F. Aubry, A
database manager of biomedical images, Med.
We perform both external and internal evaluation of Inform., vol. 18, no. 1, pp. 23-33, 1993 .
our integrated approach. In the internal evaluation, [6]. F. Pinciroli, C. Combi, and G. Pozzi,
we measure the impact of using our techniques on ARCADIA: A system for the integration of
WTDS and web service performance measures like angiocardiographic data and images by an
communications cost, response time and throughput. object- oriented DBMS, Comput. Biomed.Res.,
In the external evaluation, we compare the vol. 28, pp. 5-23, 1995.
performance of our approach to that of other [7]. Karmouch, Multimedia d istributed cooperative
techniques in the literature. The results show that system, Comput. Commun., vol. 16, pp. 568–
our integrated approach significantly improves 580, Sept. 1993.
services requirement satisfaction in web systems. [8]. R. E. Rakel, Textbook of Family Practice, 5th
This conclusion requires more investigation and ed. Philadelphia, PA: Saunders, 1995.
experiments. [9]. Digital Imaging and Com munications in
Medicine (DICOM) Version 3.0, A mer. College
VII. FUTURE ENHANCEMENT Radiologists/Nat. Elect. Manufacturers Assoc.,
1993.
As future work we plan to investigate our approach [10]. C. T. Liu, C. C. Lin, J. M. W ong, S. K. Chiou, R.
on larger scale networks involving large number of S. Chen, J. H. Chen, S. M. Hou, and T. Y. Tai,
sites over the cloud. We will consider applying [11]. Design and evaluation of a telediagnosis system,
different types of clustering and introduce search Biomed. Eng. Applicat., Basis Commun., vol. 9,
based technique to perform more intelligent data pp. 52-60, Apr. 1997.
redistribution. Finally, we intend to introduce [12]. H. S. Chen et al., Integrated medical
security concerns that need to be addressed over data informatics with small group teaching in
fragments. medical education

Cite this article as :


VIII. REFERENCES
T. K. Anusuya, P. Maharajothi, "A Survey of
[1]. T. Kitanosono, Y. Kurashita, M. Honda, T. Telemedicine Services using Datamining",
Hishida, H. Konishi, M. Mizuno, and M. A nzai, International Journal of Scientific Research in
The use of multimedia in patient care, Comput. Computer Science, Engineering and Information
Methods Programs Biomed., vol. 37, pp. 259– Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
263, 1992. 5 Issue 1, pp. 319-321, January-February 2019.
[2]. S. T. C. Wong and H. K. Huang A hospital Available at doi :
integrated framework for multimodaliity image https://doi.org/10.32628/CSEIT1183818
base management, IEEE Trans. Sy st., Man, Journal URL : http://ijsrcseit.com/CSEIT1183818
Cybern. A, vol. 26, pp. 455–469, July 1996.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


321
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195171

Medical Product Verify and Complaint Management System


for Ngo’s
R. Gomathijayam *1, J. Ayeesha Parveen2
*1Department of Computer Application, Bon Secours College for Women, Thanjavur, TamilNadu, India
2Department of Computer Science, Bon Secours College for Women, Thanjavur, TamilNadu, India

ABSTRACT

Today’s world belongs to information technology. Information is worthwhile only if, it can be accessed at the
right time, by the right person & is useful for the purpose defined. The purpose of Medical Product Complaint
Management System (MPCMS) is to provide their prescription details and complaint registration easier one.
This is an important tool to control and effectively manage the entire operation of patient details management
and complaint registration in a very efficient manner and take it to the path of growth. The System deals with
the collection of patient details and the prescription details and outputting it to the user through web
application. The prescription details are stored into the database and when the user logs into the system they
can view their prescription details. The user has been given a login to view the details and he can able to
register the complaint in case of severe problems occurred due to the wrong prescription of medicines. The
complaint will be submitted and it will be viewed by the NGO. The NGO is having separate login to view the
complaints and he will be given the authority to accept/reject the complaint.
Keywords : Medical Product, Complaints, NGO, Prescription, Patient, MPCMS (Medical Product Complaint
Management System).

I. INTRODUCTION governments, foundations, businesses, or private


persons. Some avoid formal funding altogether and
Effective complaint system can improve the are run primarily by volunteers. NGOs are highly
patient’s safety and leads to reduction in doing diverse groups of organizations engaged in a wide
mistakes. The complaint management systems are range of activities, and take different forms in
therefore important to help improving the safety different parts of the world. Some may have
and livelihood of patients. This system is for charitable status, while others may be registered for
complaint registering and acceptance of complaints tax exemption based on recognition of social
by the NGO’s therefore the users can be able to purposes. Others may be fronts for political,
register the case easily and therefore the necessary religious or other interest groups.
action be taken. A non-governmental organization
(NGO) is not the term commonly used for an The system includes separate module for user and
organization that is a part of neither a government NGO logins. Each login is given authentication as
nor a conventional for-profit business. Usually set the system involves handling medical data. In the
up by ordinary citizens, NGOs may be funded by routine functioning of a hospital, various types of

CSEIT195171 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 322-326 ]
322
R. Gomathijayam, J. Ayeesha Int J Sci Res CSE & IT. January-February-2019 ; 5(1) 322-326

investigations are carried out. Carrying out number his login. The online storage can be implemented
of tests and making the results available promptly is using cloud computing which provides a shared and
very crucial for assessing the patient's medical status secured access to all the resources shared. If a
and deciding on the further course of action. patient is currently admitted then that patient’s
Patient’s medical record data is critical for the current status (whether he is undertaking tests or in
analysis and research purposes and therefore it has OPD etc.). Doctors can have this customized version
to be stored and manipulated in an efficient manner. of application on their tablets based on their
This data includes patient history, observation, speciality that can give them all the patient details
diagnosis and therapeutic conclusions along with in hand which can help manage their schedule by
the tests details and serves as a source of prioritizing each patient.
information for any analysis and research. The
purpose for this storing the medical data is for Doctor can also send pictures and videos of a
utilizing the patient’s medical information and uses patient’s condition to other doctors for expert
it for analysis thereby improving patient care. opinion if required; doctors can give online
prescriptions to pharmacy directly specific to
When the user logs into the system, he can view all particular patient with their patient id. We can
his medical data through ‘medical prescription develop this application using android programming,
details module’. Since the data are stored in case of android programming being open source the entire
severe problems after the treatment, the user can OS is open for usage and extension. Every person
register the complaint as he has the electronic proof. can view the website and should be able to register
Further if the concerned doctor is not available and him/her providing all the necessary details required ,
if there comes the need to take care of the person by the system will provide a Patient’s ID to the patient
some other doctor he can make use of the stored and further actions can be taken based on the this
proof instead of searching the prescriptions and patient ID .A patient can take the appointments
medical data from the form. Even if third party tries online and know the availability of the doctor.
to login / alter the data he will be prompted to the Notifications about regular health check-ups and
error page as there were strong validations in the vaccinations can be sent to the patients and doctors.
system. Website also contains page for online journals
II. LITERATURE SURVEY which include all the latest medical advances and
also medical breakthroughs, unique cases for study
A. Functional Description of Online Medical and videos featuring several surgeries and research.
Management System Using Modern The user can also check for tips. Using web
Technologies: Priyanka Patil, Sruthi technology it can be easy to provide all the above
Kunhiraman, Rohini Temkar, VES Institute of functionalities and making the management of
Technology, Chembur, Mumbai patients efficient.

As many web services are made available online,


B. Handling e-complaints in customer complaint
Almost every field is made online, web based
management system using FMEA as a
applications can provide a boon to hospital
qualitative system:
management. The system should incorporate many
things online that include the following Most Negative customer perception often manifest as
importantly maintaining the patient’s records in formal companies; while unknown percentages do
details including his disease, history, reports etc. not become so visible. Within a customer complaint
which the doctor can access from anywhere using management system, it is possible through careful
Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com
323
R. Gomathijayam, J. Ayeesha Int J Sci Res CSE & IT. January-February-2019 ; 5(1) 322-326

analysis to identify the flaws in the service causes of online complaints; (2) seek effective ways
standards, in the complaint handling procedure of handling customer complaints by examining
itself, and the finer points which help to formulate different product types; and (3) provide guidelines
resolution methods. When major key problems for successful e-CRM. 1000 customer complaints
persist, complaints may spread across a range of from three different publicized e-business customer
company activity, from operational employee service centres and 500 complaints from online
performance to managerial handling of both policy feedback systems were analyzed in this study. The
and practice. A poorly managed and/or designed research findings suggest that e-businesses should (1)
complaint system will inevitably impact on the provide excellent online customer services because
company's profit and loss status, through excessive customer service is the most important factor in
expenditures, damage to market reputation and online customer satisfaction; (2) respond to
declining income. As the ultimate goal of a customers' requests/complaints quickly because the
company is to content customers, acquire an response speed is more important in online
efficient profit margin, and have a long-lasting customer satisfaction than offline; and (3) employ
relationship with customers and improving market strategies that are appropriate for the product
shares. They should handle the complaints within category in question.
complaint management system and this system
alone must be under the monitoring of customer D. A Multimedia Medical Report Management
relationship management system. The aim of this System:
study is to identify these irregularities and to
The goal of the MrBrAQue (Medical Report Browse
propose a strategy to minimize them. For improving
and Query) project is to develop a system that
reliability of the company we benefited from failure
enables effective management of a large collection
mode and effect analysis technique (FMEA) which
of multimedia documents. The tools for consulting
lets preventive analysis of failures that may
the report collection are presented in this paper.
compromise credibility of the company. As a
Our proposal is based on a paradigm, which we call
methodology model set of 30 questionnaires have
orthogonal navigation that allows a user to
been assigned for this reason and a case study of an
dynamically build a hypermedia network of
airline has been chosen. Ultimately, using, 5-point
documents by means of queries over the document
likert scale and SPSS software, we were managed to
collection
administer the data and conclude the results.

E. Design of Medical Management Information


C. An analysis of online customer complaints:
System Based On Soa:
implications for Web complaint management:
Medical retailers face the challenge of integrating
How businesses resolve customer-complaining
their legacy systems within their stores and at the
behaviour effectively has been considered a
enterprise (or corporate) headquarters. This
"defensive marketing" strategy or a "zero-
challenge becomes even more critical due to
defections" strategy, which diminishes customer
demanding customers who expect a rich experience
dissatisfaction. Handling customer dissatisfaction
both in the store and in other channels. In this
accompanies Web customer complaint management,
paper, a few common scenarios and use cases in
which might be the critical issue for online
retail have been discussed. The paper puts forward a
customer service solutions and e-CRM (electronic
new design method of medicine management
customer relationship management). In this paper;
information system based on SOA. This new system
the authors (1) investigate the current sources and
Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com
324
R. Gomathijayam, J. Ayeesha Int J Sci Res CSE & IT. January-February-2019 ; 5(1) 322-326

model makes full use of legacy applications and is • Less Time constraint for registering complaints.
flexible for future extensions. Experiment shows • Once the complaint is accepted, the patient can
that real-time, intelligent performances and register the case so that they can be given some
efficiency of operation are obviously improved. fund in return to the mistakes done.
F. Clinical Database Management Software
(CDMS) for medical, diagnostic and research
centres:

The Clinical Database Management Software


(CDMS) is a state-of-the-art concept in integrated
medical report recording and analysis for the
medical and research fraternity. It revolutionizes
the archiving and analysis of patient data and
replaces the traditional passive and manual method
of collecting patient data in text format with an
active, online, easy-to-use pop-up and pull-down
series of menus in an interactive mode in a very
structured format. This new concept is the outcome
of years of dedicated and continuous interaction
with leading medical practitioners, research
scientists and software specialists. Lacunae in the
collection of patent clinical and research data and Figure 1 : System Architecture
its management should now be a matter of the past.
Flexibility and cost-effectiveness are the most IV. CONCLUSION
important aspects of CDMS. The software is easily
adaptable to a variety of database requirements and, Thus the above mentioned system has been
with a minimum of modification of the source code, developed to satisfy all the requirements of the user.
it can be configured to meet the very specific needs Since it gives the user an easy way to view
of the medical and research fraternity. prescription details and to register complaints it
takes less time constraint to make the needful. The
III. METHODOLOGY system is simple and easy to implement. Since there
is high level of authentication, only right person can
In the proposed system, everything will be online. view and edit the details. By implementing this
The patients will have separate login to avail system, users can attain maximum benefit with less
services. The prescription details for that concerned time constraint and energy.
patient will be available to them in their Medical
Prescription Details module. The system provides V. REFERENCES
high security by granting access rights to change the
sensitive data. [1] An analysis of online customer complaints:
• The prescribed data will be secured since implications for Web complaint management
authentication has been provided to view it. -Yooncheong Cho ; Rutgers Univ.,
• Medical complaint forms are there for Piscataway, NJ, USA ; Il Im ; Hiltz, R. ;
registering the complaints in case of severe side Fjermestad, J.
effects.
Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com
325
R. Gomathijayam, J. Ayeesha Int J Sci Res CSE & IT. January-February-2019 ; 5(1) 322-326

[2] Handling e-complaints in customer complaint qualitative system, Faed, A., DEBII, Curtin
management system using FMEA as a Univ. of Technol., Perth, WA, Australia
qualitative system - Faed, A. ; DEBII, Curtin
Cite this article as :
Univ. of Technol., Perth, WA, Australia
[3] A multimedia medical report management
R. Gomathijayam, J. Ayeesha Parveen, "Medical
system ,Merialdo, P. ; Dipartimento di Inf. e
Product Verify and Complaint Management System
Autom., Rome Univ., Italy ; Sindoni, G.
for Ngo's", International Journal of Scientific
[4] Functional Description of Online Medical
Research in Computer Science, Engineering and
Management System Using Modern
Information Technology (IJSRCSEIT), ISSN : 2456-
Technology - Priyanka Patil, Sruthi
3307, Volume 5 Issue 1, pp. 322-326, January-
Kunhiraman, Rohini Temkar,VES Institute of
February 2019. Available at doi :
Technology, Chembur, Mumba
https://doi.org/10.32628/CSEIT195171
[5] Handling e-complaints in customer complaint
Journal URL : http://ijsrcseit.com/CSEIT195171
management system using FMEA as a

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


326
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195191

A Survey on Child Safety Wearable Device to Prevent Child


Trafficking Using Arduino
Elakiya M.1, S. Radhika2
1Master of Engineering (ME), Department of Computer Science and Engineering, RMK Engineering College,

Chennai, India
2 Assistant Professor, Department of Computer Science and Engineering, RMK Engineering College,
Chennai, India
ABSTRACT

Currently a child’s security is a crucial space of concern. This paper is developed to rectify the troubles parents
relating to their child security. In keeping with a story from 2014, about 135,000 kids are calculable to be
trafficked in India annually. Trafficked children are sold-out into slavery, domestic slavery, beggary, and
therefore the sex trade. This paper provides the conception of sensible wearable devices for our little ones. And
to stop kid trafficking. One amongst the most precedence over this device is that it provides us the tracing
details of our little ones. And it doesn’t require a smart phone or a technician person to know and use it. The
motive of this device is to assist find their kids with none difficulties. At the moment there are several tracking
device that helps us to note our children’s activity with ease and additionally notice the kid using Wi-FI and
Bluetooth. However Wi-Fi and Bluetooth seem to be unpredictable medium of communication between the
parent and kid. So the main focus of the paper is to possess an SMS text enabled communication between the
child’s wearable and also the parent because the GSM mobile communication is nearly present everyplace. The
parents will send a text with specific keywords like “LOCATION” “TEMPERATURE” “UV” “SOS” “BUZZ” ETC.
The tracking device can reply back providing the correct location of the kid and it will navigate through google
maps. So this paper provides parents with a way of security for his or her kid in today’s world.
Keywords : Child Safety Wearable Device, Arduino, Trafficked children, Lynceus2Market, RFID, MASS, RSSI,
DCC, ATNAC

I. INTRODUCTION the other describing the issues faced by these


fishermen in their day to day commute to the oceans.
Fishing is one of the chief sources of food and income In order to solve this issue, the governments of these
for almost all coastal lands irrespective of its countries decided to have a common territory of
geological location in the earth. Since it has an ocean as international waters and that it would act as
important role to play in the economy of a country, the region common as well as a separation between
there is no doubt that neighboring countries sharing the two lands. But even this did not prevent the
the same oceans frequently engage in disputes in fishermen from unknowingly wandering off into the
regards to ownership of the area. This has resulted in other country’s waters. Hence, there is a strong need
deep problems to the fishermen community residing to device methods to prevent this from happening
in the coastal regions of these countries. Often, we and save the fishermen from severe punishments and
hear in the news and see in the papers, one article or border disputes. Our model helps to ensure the

CSEIT195191 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 327-333 ]
327
Elakiya M., S. Radhika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 327-333

location of the fishing boats through a new Ultra low-power UWB-RFID system for precise
technology using radio waves. location-aware applications
Ultra low-power radio-frequency identification
(RFID) tag with preciseness localization is commonly
the sanctioning technology for location-aware
detector applications. The RFID tag could be a
transmitter comprising of a micro-controller board
and a UWB impulse radio board. Power saving and
precision localization is achieved by improvement of
the circuit style for immoderate short pulses further
as system design and operation.

Petrovic and R. Kanan, “Extremely low power indoor


localization system,” in IEEE 8th International
II. LITERATURE SURVEY Conference on Mobile Adhoc and Sensor Systems
(MASS), pp. 801–806, Oct. 2011.
“Lynceus2Market- An innovative individuals
localization system for safe evacuation of huge Extremely Low Power Indoor Localization System
traveler ships” In this paper style, implementation and testing of an
An innovative individual localization system for safe very low power and low price wireless detector
evacuation of huge traveler ships network for observing person or object position in an
interior atmosphere are mentioned. The localization
This paper is mainly based on the maritime disasters relies on the standard received signal strength
to control the cause, during emergency evacuation of intensity (RSSI) technique. The very low-power
large passenger ships. The Lynceus2Market project feature is achieved by employing a new Duty-Cycle
addresses this challenge through delivering a dominant (DCC) technique. The DCC technique
revolutionary operational system for safe evacuation relies on the frequency Identification (RFID)
supported innovative people localization principle to wake-up the mobile unit at the correct
technologies. time and save energy once it's not used.

This technology will be provided with SME –driven S. Nasrin and P. 1. Radcliffe, "Novel protocol allows
market segments of smoke alarm, fire detection DIY home automation," Telecommunication
systems, life saving equipment, emergency Networks and Applications Conference (ATNAC),
management decision support systems and assistive Novel Protocol Enables DIY Home Automation.
search and rescue equipment.
In this paper, we tend to describe a brand new design
Y. Zhou, C.L. Law and J. Xia, “Ultra low-power for a home automation system that is constructed on
UWB-RFID system for precise location-aware novel network protocols.
applications,” in Wireless Communications and
Networking Conference Workshops (WCNCW), pp. The new system permits a user or home-owner to
154–158, April 2012. buy off the shelf devices and management them by a

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


328
Elakiya M., S. Radhika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 327-333

computer or mobile device terribly} very easy Objective measures of response throughout these
manner. phases could facilitate determine kids in danger for
The system does not require expert configuration or a anxiety, we tend to propose an alternate technology,
central controller therefore the cost will be an inertial measurement unit (IMU), that permits
significantly reduced. The key enabling technology non-invasive activity of the movements related to
that makes this possible is an XML definition threat response.
Of the device capabilities, show needs, and
management signals. Internet of Things: a definition & taxonomy: Bruno
Dorsemaine, Jean-Philippe Gaulier, Jean-Philippe
Md Imdadul Hoque,1,* Abdullah Al Amin,1 Md Wary and Nizar Kheir
Jumshadur Rahaman,1 and Mohammad Hossam-E-
Haider1 1Electrical Electronic & Communication Internet of Things: a definition & taxonomy
Engineering Department Military Institute of Science
and Technology, Dhaka, Bangladesh The Internet of Things (IOT) has varied fields of
Designing and Performance analysis of GSM/GPS application including health care, resource
based mostly heavier-than-air craft following Device management, asset tracking, etc. depending on the
utilization case, varied technologies like RFID,
Low flying aircrafts are always powerful to trace Wireless sensing element Network (WSN) or smart
because of the radar horizon limitation. However an Objects is used. With each of these comes a specific
aircraft is often to be tracked for the security purpose vision of what the IoT and connected objects are and
.The combination of world Positioning System (GPS) – to our information – there's no global image of the
and international System for Mobile IoT.
Communications (GSM) also can perform the duty
satisfactorily. During this work, two GSM networks Jun Zheng David Simplot-Ryl Chatschik Bisdikian
were utilized to trace low flying aircrafts in several Hussein T. Mouftah
routes. The device style is projected basing on the
experimental results of chase. GUEST EDITORIAL – THE INTERNET OF THINGS

Ellen W. McGinnis*, Ryan S. McGinnis*, Member, The growth of the internet shows no signs of
IEEE, Maria Muzik, Jessica Hruschak, Nestor L. retardation down and it steady becomes the cause of
Lopez-Duran, Noel C. Perkins, Kate Fitzgerald, a brand new pervasive paradigm in computing and
Katherine L. Rosenblum communications. This new paradigm enhances the
standard internet into a smart net of Things (IoT)
Movements indicate threat response phases in created around intelligent interconnections of
children at-risk for anxiety numerous objects within the physical world, like
vehicles, cell phones, habitats, and habitat occupants.
Temporal phases of threat response similarly as It utilizes affordable information gathering and
Potential Threat (Anxiety), Acute Threat (Startle, dissemination devices, like sensors and RFID tags,
Fear), and Post-threat Response Modulation area unit that facilitate fast interactions among the objects
referred to as underlying markers of tension themselves as well because the objects and persons in
disorders. anyplace and at any time. IoT isn't a mere extension
of today’s net or an online of net systems.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


329
Elakiya M., S. Radhika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 327-333

Virtual Rehabilitation with Children. Challenges for


clinical adoption. By Stephanie Glegg.
Virtual Rehabilitation with Children III. COMPONENTS REQUIRED

The main draw of virtual rehabilitation is that the A. GPS


capability it allows clinicians to regulate the
The Global Positioning System (GPS) is a U.S. space-
characteristics of the virtual surroundings whereas
based global navigation satellite system.
modifying the degree of challenge to suit individual
patient desires. As an example, by reducing the
It provides reliable positioning, navigation, and
amount and speed of visual stimuli, removing visual
temporal order services to worldwide users on a
clutter and additive feedback, and directional stimuli
nonstop basis day and night, anyplace on or close to
to 1 aspect of the body only, therapists will modify a
the planet.
child with a brain injury to participate in virtual
reality in a way that optimizes success and recovery.
B. GSM MODEM
Augmented feedback (e.g., visual, auditory, and
haptic) and opportunities for motivating repetitive GSM electronic equipment may be wireless
practice produce this approach considerably electronic equipment that works with a GSM wireless
appealing for motor learning applications network.

ChildGuard: A Child-Safety Monitoring System A wireless modem behaves like a dial-up modem.
Zhigang Gao, Hongyi Guo, Yunfeng Xie, and Yanjun The main distinction between them is that dial-up
Luo Hangzhou Dianzi University Huijuan Lu and Ke electronic equipment sends and receives information
Yan China Jiliang University. through a set telephone line whereas wireless
electronic equipment sends and receives information
ChildGuard: A Child-Safety Monitoring System through radio waves.

The ChildGuard system exploits mobile devices to The operating of GSM electronic equipment relies on
observe the period of time movement of commands.
unsupervised children. It additionally sends
reminders and alarms to kids and notifies guardians C. LCD
of abnormities in a child’s routine. for instance, two
typical measures embody hiring manned guards or LCD (Liquid Crystal Display) screen is associate
using video surveillance, but guards aren’t practical electronic show module and notices a large vary of
or affordable when it involves monitoring kids, and applications.
video surveillance systems typically have blind zones.
to that end, ChildGuard may be a security A 16x2 {lcd liquid crystal display |LCD| digital display
methodology for observation kids that uses present |alphanumeric display} display is incredibly basic
computing devices, like smartphones or wearable module and is extremely unremarkably employed in
devices—which are growing in each quality and varied devices and circuits.
performance.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


330
Elakiya M., S. Radhika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 327-333

These modules area unit most popular over seven The sensors contain two in contact with an
phases and alternative multi segment LEDs. electrolyte.
The reasons being: LCDs are economical; easily The electrodes are usually invented by fixing a high
programmable. extent valuable on to the porous hydrophobic
membrane.

The perating conductor contacts each the solution


and also the close air to be monitored typically via a
porous membrane.

D. TEMPERATURE SENSOR - The LM35

The LM35 is associate degree computer circuit


detector which will be accustomed live temperature
with associate degree electrical output proportional
to the temperature (in oC)

The LM35 - An Integrated Circuit Temperature


Sensor
Why Use LM35s To Measure Temperature?
F. ARDUINO UNO
1.You can measure temperature more accurately than
a using a thermostat. Arduino/Genuino Uno is a microcontroller board
2.The sensor circuitry is sealed and not subject to supported the ATmega328P. It has fourteen digital
oxidation, etc. input/output pins (of that six will be used as PWM
outputs), six analog inputs, a sixteen MHz quartz, a
USB association, an influence jack, an ICSP header
and a reset button. It contains everything required to
support the microcontroller; merely connect it to a
laptop with a USB cable or power it with a AC-to-DC
adapter or battery to urge started.

G. ARDUINO IDE:

The ATmega328 on the Arduino/Genuino Uno comes


preprogrammed with a boot loader that allows you to
upload new code to it without the use of an external
hardware programmer. It communicates
victimization the first STK500 protocol (reference, C
header files).
E. GAS SENSOR

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


331
Elakiya M., S. Radhika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 327-333

radiation and SOS light with alarm buzzer for their


child's surroundings and the ability to locate their
H. EMBEDDED C child or alert someone in acting to rescue or comfort
the child. The smart child safety wearable can be
Embedded C may be a set of language extensions for
enhanced much more in the future by using highly
the C programming language by the C Standards
compact Arduino modules such as the LilyPad
committee to handle commonality problems that
Arduino which can be sewed into fabrics. Also a
exist between C extensions for various embedded
more power efficient model will have to be created
systems.
which will be capable of holding the battery for a
longer time.
Historically, embedded C programming needs
nonstandard extensions to the C language so as to
V. REFERENCES
support exotic options like fixed-point arithmetic,
multiple distinct memory banks, and basic I/O
[1]. "Lynceus2Market - An innovative people
operations.
localisation system for safe evacuation of large
passenger ships." Online]. Available:
I. Power Supply:
http://www.lynceus-project.eu/ .
This section describes how to generate +5V DC [2]. Y. Zhou, C.L. Law and J. Xia, "Ultra low-power
power supply UWB-RFID system for precise location-aware
applications," in Wireless Communications and
Networking Conference Workshops
(WCNCW), pp. 154–158, April 2012.
[3]. D. Petrovic and R. Kanan, "Extremely low
power indoor localisation system," in IEEE 8th
International Conference on Mobile Adhoc and
Sensor Systems (MASS), pp. 801–806, Oct.
2011.
[4]. N. Simon et al., "Indoor localization system for
emergency responders with ultra low-power
radio landmarks," in Instrumentation and
IV. CONCLUSION AND DISCUSSION Measurement Technology Conference (I2MTC)
,pp 309-314, may 2015
This paper briefly analyzes and reviews the related [5]. S. Nasrin and P.I Radcliffe, " Novel protocol
work carried out for child safety. When compared enables DIY home automation,"
with all the technologies and paper we finally prefer Telecommunication Networks and Applications
using SMS text enabled communication between the Conference (ATNAC) , 2014 Australasian,
child’s wearable and also the parent because the GSM Southbank, VIC, 2014, pp. 212-216.
mobile communication is nearly present everyplace. [6]. Internet of Things: a definition & taxonomy:
The usage of this can make us track the child’s Bruno Dorsemaine, Jean-Philippe Gaulier,
location easily. Only GSM/ GPS is used to track the Jean- Philippe Wary and Nizar Kheir.
child. It provides parents with the real-time
location, temperature around their child, UV

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


332
Elakiya M., S. Radhika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 327-333

[7]. EDITORIAL- THE INTERNET OF THINGS:


Jun Zheng David Simplot- Ry1, Chatschik
Bisdikian and Hussein T.Mouftah
[8]. Virtual Rehabilitation with children.
Challenges for clinical adoption. By Stephanie
Glegg
[9]. ChildGuard: A Child-Safety Monitoring System
Zhigang Gao, Hongyi Guo, Yunfeng Xie, and
Yanjun Luo Hangzhou Dianzi University
Huijuan Lu and Ke Yan China Jiliang
University.
[10]. Guest Editorial: Mobile wearable
communications Hassnaa Moustafa, Holger
Kenn, Kamran Sayrafian, William Scanlon and
Yan Zhang.
[11]. Ellen W. McGinnis*, Member, IEEE, Maria
Muzik, Jessica Hruschak, Nestor L. Lopez-
Duran, Noel C. perkins, kate Fitzgerald,
Katherine L.Rosenblum "Movements indicate
threat response phases in children at risk for
anxiety"
[12]. Md Imdadul Hoque,1,* Abdullah Al Amin,1 Md
Jumshadur Rahaman,1 and Mohammad
Hossam-E-Haider1 1Electrical Electronic &
Communication Engineering Department
Military Institute of Science and Technology,
Dhaka, Bangladesh "Designing and
Performance Evaluation of GSM/GPS Based
Helicopter Tracking Device"

Cite this article as :

Elakiya M., S. Radhika, "A Survey on Child Safety


Wearable Device to Prevent Child Trafficking Using
Arduino", International Journal of Scientific Research
in Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 327-333, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195191
Journal URL : http://ijsrcseit.com/CSEIT195191

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


333
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195192

Bus Notification System Using Crowdsourcing


B. Harshini Reddy, K. Jhansi Rani, M.Vaishnavi
Information Technology, SNIST, Ghatkesar, Telangana, India
ABSTRACT

Most of the daily commuters of RTC face problems to know the exact time of arrival of a bus to a particular stop.
The time of arrival of bus depends on many factors like traffic, condition of a bus, climate etc. So it is very
tough for a person waiting for a bus without knowing whether it has left or not from a particular stop. So here
we are providing a solution for this problem by using latest technologies like an Android Application. The
solution is that we want to develop an ANDROID APP[1] which gives the notification of a bus arrival in a
particular route for a particular stop. The backbone of our application is using CROWDSOURCING. It is
nothing but taking the help of some people to help other people. The passengers who have seen or boarded the
bus can report the bus location along with an image of bus with the help of our app. Then this location will be
sent as a notification to all other commuters in that particular route. The commuters who have missed the bus
from the stops through which the bus has already passed, can go with an alternative option and the other
commuters waiting at the stops to which bus has to arrive can wait accordingly. This helps in saving our
precious time which many of us wasting it in waiting for a bus. In this way our application helps the daily
commuters to know the location of a bus and saves their precious time.
Keywords : Crowd Sourcing, Androidapp, Notification, Location

I. INTRODUCTION ordinates of the bus and send information to the


server.
Most of the daily commuters of RTC are facing
problems to know at which time bus will arrive or it The solution is that we want to develop an
has gone from particular stop. And the time of arrival ANDROID APP which gives the notification of a bus
of bus depends on traffic, condition of a bus, climate arrival in a particular route for a particular stop. The
etc..so, it is tough for a person to wait for a bus. backbone of our application is using
CROWDSOURCING[3]. It is nothing but taking the
So here we are with a solution, to this problem by help of some people to help other people. The
using Bus notification System with an android passengers who have seen or boarded the bus can
application. Our main motto is that the person report the bus location along with an image of bus
receives the exact location of the bus with the with the help of our app. Then this location will be
estimated departure time of the requested bus[7]. To sent as a notification to all other commuters in that
get into this process the person needs to register particular route.
himself in the application. Once the person gets
himself registered he can receive the notifications The commuters who have missed the bus from the
with the help of VMD(Variable Message stops through which the bus has already passed, can
Display)[2].This VMD is used to track the co- go with an alternative option and the other

CSEIT195192 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 334-339 ]
334
B. Harshini Reddy et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 334-339

commuters waiting at the stops to which bus has to Initial review which is organized is for proving that
arrive can wait accordingly. This helps in saving our problem regarding buses being late to bus stops exists.
precious time which many of us wasting it in waiting At the same time, we assessed the level of significant
for a bus. In this way our application helps the daily of the problem. We observed the expected arrival
commuters to know the location of a bus and saves time to each bus stop for a period of time for two
their precious time. weeks. The result of the observation is that most of
the time the bus failed to arrive at the bus stop as
II. LITERATURE SURVEY they are scheduled to arrive. Furthermore, we had
got a shocking observation result that is sometime the
For any enterprise or task, the initial phase would be bus skipped its trip and causing the passengers to wait
literature survey concerning with formerly issued for the next bus that might arrive an hour later.
journals. Prior to decision making, a thorough Sometimes the bus arrives late at the first bus stop for
patrons review regarding journals is done in a more than fifty minutes. It is causing the passengers
sequence manner. Complete knowledge and in that entire route to be delayed and arrive at late to
interpretation concerning the enterprise would be their workplaces. Also, the problem even happens
the main principle of this literature review. It also during peak hours where most of the passengers will
helps us lay a firm base for upcoming tasks. be rushing to their workplaces. These observation
Research which is undergone serves as guideline and results successfully proved that our problem
citation to the enterprise. Once the literature analysis statement is valid and these problems have to be
is done, we should be in a position to write to such a solved urgently for the benefit of the city and the bus
degree that manifests our experience and observation passengers. The second study is via survey. The main
regarding that particular domain and realize principal aim of this survey is to help understand the current
subject which are pertinent for the project. It’s demand and acceptance for our proposed solution in
important to have an idea about the one’s that can be the market. In order to understand the culture of the
ignored and must be in a position to predict end passengers, the problems faced by the passengers, and
result. the need of our proposed system, and the present
By summing up wholly, undeniably the intention of condition of public transportation we conducted a
literature survey is to find out the oversight survey by distributing hundred survey forms to the
pertaining to a task. The SMS Based Forecasting passengers who travel frequently through these
System of Public Transport Arrival Time[5]. public transportation.
After issue being identified and then coming up with According to our survey results the majority of the
suitable solving methodology we organized some commuters of the public transportation are
research concerning bus passengers stating our issue workforce and students who use this for travelling to
explanation, also explaining our recommended their respected workplaces daily. Also the most
solution. Three main segments are concentrated in common and major problem that is that the buses
our analysis. They are: usually do not arrive to the bus stops according to
1. Demonstrate as well as appraise the importance their scheduled time. Due this problem there is an
regarding issue description . inconvenience caused to the passengers and they are
2. The ultimatum and necessity of suggested end forced to wait for unpredictable frame of time.
result must be recognized and understood. Majority of the passengers who took part in survey
3.Providing the issue description demonstrated is reported that a SMS system which gives updates
correct, extract and analyse it. about the whereabouts and the arrival timings of the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


335
B. Harshini Reddy et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 334-339

bus would be a better solution for their problems. off. The methods used in this system are
Based on the this survey results and the information onPause(),onResume(),onKeyDown() and
obtained we are very positive that our proposed onDestroy().The Post-configuration process is
solution would be accepted by the bus passengers. managed by onDestroy() and onKeyDown() which
The current status of the problems faced by the makes the usage convenient.
passengers is very high. Thus we derived some
hypotheses to help us to understand the factors The implemented modules in this system are:-
which are affecting the bus’s arrival time. These
hypotheses are extremely important for us to 1. Bus Unit
understand the parameters involved in this system. 2. Central Control Unit
Based on the result of the hypotheses, we are able to 3. Client-Side Application
understand the possible causes of bus delays and this
The bus unit consists of GPS which tracks the co-
information will provide great assistant.
ordinates such as latitude, longitude and other
attributes of that particular location and send s it to
This report is a survey which helps the passengers to
the server. The GPS is free to use i.e.., no subscription
overcome the difficulties they are facing in their
fee is required. The GPS has the capability to
daily transportation. This helps to conserve time
receive information about these locations from three
which in very important. This not only saves the
satellites and calculates the position based on this
time but also it helps to form an advance transport
information. There are three different types of
system.
transceivers in GPS such as Data Loggers, Data
pullers and Data Pushers .Based on the application,
There are many resources present which gives us the
the appropriate transceiver will be chosen to gather
information regarding the bus numbers and its details.
the information and track the position. The GPS
In this review we get advance information like
attached to sends the information to the servers at
vacant seats, location of the bus, estimated time of
regular intervals of time. This information is received
arrival and departure to a particular place which can
by the device from the servers.
be viewed by the travellers.Gradual advancement of
bus systems reaches demands pertaining to various
The received data is analysed by the servers .The GPS
areas. Main motto of the application is to have a
antenna is connected to the right jack and fixes it’s
comfortable, simple and accessible bus facility that
position to receive signals in the appropriate places
uplifts the usage of public transport system.
.There is a slot allocated for a SIM card, the GSM
towers helps in receiving messages from different
III. PROPOSED METHODOLOGY
users and respond to them. The positive and negative
wires are connected to 12V or 24V vehicle power
This model consists of four parts. Each and every
system [4]. The tracker device is turned on to receive
individual model is divided based on their
the signals from the satellite hence thus enabling the
functionalities. The main activity of our application
device to receive the latitude and longitude values of
perform major tasks which include Pre-configuration
the position of the bus. At any point of time the GPS
and Instantiation methods. It consists information
receiver gives us the location values. The coordinates
about the total distance between the two locations
of the bus along with the timestamp is present in the
and also about the previous and current location. The
bus unit, these values are compared with the previous
status of the GPS is also examined whether it's on or

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


336
B. Harshini Reddy et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 334-339

coordinates if there is any distinction then the new and the elements which are in correspondence with
co-ordinate values are updated and will be sent to the the stack are integrated in a tight manner to provide
server over the Internet[4]. the convenient and proper execution frame for
mobile devices.
The location details of the bus are saved in a
particular format with few attributes such as As per the advancement of the technology the
Latitudinal position, longitudinal position, ID and Android is developed for the touch screen devices
timestamp etc. Every bus has a unique ID which like mobile devices and tablets. This android is an
helps to identify different buses easily. Every bus has operating system developed by Google. The android
a different identification number called as ID. Server is based on the Linux kernel[6] whereas the
in this system takes an important role and performs Android's user interface is based on direct
major functions. Here the server stores and maintains manipulation with the help of human touches that
the complete information. The server also performs help to perform actions on the screen and also to give
actions like a gateway between the user and the bus input[5]. In extension to the devices Google added
module. The overall information about the routes, Android TV, Android Wear etc. with an exclusive
arrival time, department time are stored in the new user interfaces. The complication of the mobile
database. world has completely been reduced with the use of
Android. The unimaginable features are implemented
The final user can only view the application which is and high quality phones are made easily with the
an interactive web based module. Here the help of this Android OS.
application provides different services to remote
users. The user module requires two inputs that is the
current location point and the end location point. As
the user requests the server, it sends the list the buses
available in that location. Ones the bus is selected it
shows the real time location and time of that
particular bus on screen.

IV. RESULTS AND ANALYSIS

The GPS of the system is built in such a way that it


tracks the exact location and provides dual role of
communication. The communication is done
between the driver and user. This process takes a
vital role in the development of the application. All
the information is provided below. Here we will Figure 1. Screen 1 of bus notification system
come across an operating system, run-time
environment, and software stack comprising
applications, libraries, middleware and services
which are a part of the Android Architecture.
Visually this architecture can be represented in its
best manner. The layers of the stacks available in it

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


337
B. Harshini Reddy et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 334-339

Figure 2. Screen 2 of bus notification system Figure 4. Login for bus notification system

Figure 3. Registration for bus notification system Figure 5. Location accessing for bus notification
system

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


338
B. Harshini Reddy et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 334-339

method of tracking the location makes the journey


easy.

Currently we do not have sufficient data to enable us


to draw prediction result with weather factor in
consideration. Therefore, we are very positive that
the inclusion of weather factor will increase the
accuracy of PredictBus system’s prediction
significantly. Next, we would like to improve the
system database’s capacity.

Last but not least, we hope to see this application to


be able to merge a few bus routes and predict the
arrival time. This is an important prediction for users
that needs to switches buses in order for them to
arrive at their destination.
Figure 6. Map page for bus notification system
VI. ACKNOWLEDGEMENT
V. CONCLUSION

We would like to thank our Research Guide Dr.


The present application relates to a system and Subhani Shaik, Associate Professor in Department of
methods for notifying passengers of an approaching Information Technology, SNIST, Hyderabad for their
vehicle.Utilizing this application passenger can save continue support and valuable suggestions
their time and safe in controlled environment.And all throughout carried this work. Authors are also
these are done by just installing our application grateful to the reviewer for the renovation of
android app and we can find the location of a bus. manuscript. We would also like to thank the
Department of information Technology providing us
The project is completely done on Android based with the facility for carrying out the simulations.
platform.Various attributes have been implemented
so that the system is very convenient to use with the VII. REFERENCES
best and advance features.the domain use in the
system is SQL.Since the GPS is added it displays maps [1]. https://www.webopedia.com/TERM/A/android_
automatically.The tracking method is also been made app.html
easy with this GPS. [2]. https://en.wikipedia.org/wiki/Variable-
message_sign
As the application is done using cloud soucing,it is [3]. https://en.wikipedia.org/wiki/Crowdsourcing
easily accessible for all the android usrs.the [4]. www.irjet.net
application is useful for both driver-traveller and [5]. www.studymode.com
tourists.This application can also be modified and be [6]. Submitted to University Malaysia Perlis
used for tracking any other means of transport.This [7]. Submitted to Asia Pacific University College of
Technology and Innovation.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


339
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195151

An Extension to Traditional Qwerty Cryptographic Method


V. Vennila*1, S. Umamaheswari2
*1Assistant Professor, Department of Information Technology, Bon Secours College for Women,
Thanjavur, Tamil Nadu, India
2Assistant Professor, Department of Information Technology, Bon Secours College for Women,

Thanjavur, Tamil Nadu, India

ABSTRACT

Cryptography comes from the Greek word kryptos which means hidden. The aim of our research is to provide
secrecy for the data during transmission. In this we discuss about the qwerty cipher algorithm, its merits and
demerits. The existing qwerty algorithm is based on the use of a keyboard which is not much secured. We have
proposed an enhancement to the existing algorithm by using a diagonalized qwerty algorithm which is more
secure and highly difficult for the intruder to understand or decrypt the cipher text, even if it is decrypted the
result would be gibberish.
Keywords : Transmission, Intruder, Decrypt, Cipher, Gibberish.

I. INTRODUCTION
Traditional ciphers consist of substitution or
Cryptography is the strongest tool used for secret transposition techniques [2] [3]. In Substitution
writing which is controlling against the security ciphers, replace letters in the plain text with other
threats [6]. The study of encoding and decoding is letters or symbols keeping the order as same. It maps
called as cryptography [1]. In cryptography, a cipher plain text elements into cipher text elements [4].
is an algorithm consists of well-defined steps for
performing encryption or decryption. In encryption
ABCDEFGHIJKLMNOPQRSTUVWXY
process, the original plain text is coded into the Z
cipher text and in the decryption process; the plain
ZYXWVUTSRQPONMLKJIHGFEDCB
text is restored from the cipher text [2]. We need a A
strong encryption algorithm in order to encrypt the
plain text into cipher text [7]. The sender and Fig 2. Substitution Cipher
receiver must have obtained the secret key in a
secure fashion and must keep the key secure. The In Transposition ciphers, keep all of the original
process of encryption and decryption are shown letters intact, but mix up their order [5]. It transposes
below systematically the position of plain text elements.
Transposition cipher hides the message by
rearranging the order. Transposition cipher, just
reorder the letters [8]. A character in the first
Fig 1. Encryption and Decryption process position of plaintext may appear at 12th position in

CSEIT195151 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 340-342 ]
340
V. Vennila, S. Umamaheswari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 340-342

cipher text, and a character at 6th position in the Decryption Technique:


plaintext may appear at 18th position in cipher text. Cipher text: MGY DHMCM HY EPC CLME
Key: QWERTY
Example of the transposition cipher is given bellow, Plain text: SUN RISES IN THE EAST

IV. PROPOSED DIAGONALIZED QWERTY


CIPHER

The proposed diagonalized qwerty cipher is the


improvement of the qwerty cipher for making the
communication more secure and it is difficulty for
Fig 3. Transposition Cipher the hackers to hack the plain text. Diagonalized
qwerty cipher is much secured when compared to
II. EXISTING QWERTY CIPHER the qwerty cipher.
Step 1. The model of a qwerty keyboard is taken.
QWERTY is the most common modern day keyboard
layout. The name comes from the first six keys
appearing on the top left row of the keyboard and
read from left to right: Q-W-E-R-T-Y. In the existing
qwerty cipher the top line is the keys on a keyboard
from top left to bottom right, and the bottom line is
the alphabet. It uses a seemingly random letters. The Fig 5. Model of a QWERTY keyboard
model of qwerty cipher is given bellow in which the
plain text uses the qwerty keyboard to cipher the Step2. The model of diagonalized qwerty keyboard is
plain text. This cipher is easy to crack. framed.

In the qwerty cipher the keys in the keyboard are


taken in an diagonal manner for framing an
diagonalized qwerty keyboard. Which is more
secured form of communication while transmitting
message from a source to an destination. The
Fig 4. Top box represent keyboard keys and the
mapping of diagonals is shown below.
bottom represent alphabets.
Q W E R T Y U I O P
III. EXAMPLE FOR QWERTY CIPHER

Encryption Technique:
A S D F G H J K L
Plain text: SUN RISES IN THE EAST
Key: QWERTY
Cipher text: MGY DHMCM HY EPC CLME Z X C V B N M

Fig 6. Mapping of qwerty keyboard.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


341
V. Vennila, S. Umamaheswari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 340-342

The diagonalized qwerty keyboard is represented as VII. CONCLUSION


13 keys in each row for the purpose of
encrypting/decrypting a plain text in a secured way. Finally through this paper we have pointed the
traditional qwerty cipher merits and demerits. In
order to overcome the demerits, we have proposed an
extension to traditional qwerty cipher which can be
used more efficiently while encrypting the plain text.
However the used diagonalized qwerty algorithm can
be improved for encrypting large amount of data.

VI. REFERENCES
Fig 7. Representation of qwerty keyboard
[1]. ”Transformation of plain Text with two
V. EXAMPLE FOR DIAGONALIZED QWERTY ciphers”,(Rajani Bala,Swapnika saxena,Sonal
CIPHER beniwal), Volume 3, Issue 5, May 2013.
[2]. “Implementation of Ceasar Cipher with Rail
Encryption Algorithm: Fence for Enhancing Data Security”,(Ajit Singh,
Plain text: CRYPTOGRAPHY IS MY FAVOURITE Aarti Nandal, Swati Malik), Volume 2, Issue 12
SUBJECT , December 2012.
Key: DIAGONALIZED QWERTY [3]. “An Extension to Traditional Playfair
Cipher text: IKZTPFQKBTWZ CN DZ OBLFXKCPJ Cryptographic Method”( Ravindra Babu, Uday
NXAEJIP Kumar, Vinay Babu, Aditya, Komuraiah),
Volume 17, No.5, March 2011.
Decryption Algorithm: [4]. “An Improved Ceasar Cipher
Cipher text: IKZTPFQKBTWZ CN DZ OBLFXKCPJ Algorithm”,(Ochoche Abraham,
NXAEJIP GaniyO.Shefiu), Volume 2, Issue-5,1198-1202.
Key: DIAGONALIZED QWERTY [5]. "Alpha-Qwerty Cipher: An Extended Vigenere
Plain text: CRYPTOGRAPHY IS MY FAVOURITE Cipher”( Khalid Imam Rahmani,Neeta
SUBJECT Wadhwa, Vaibhv Malhotra), Volume 3, No.3,
May, 2012.
VI. ADVANTAGES [6]. William Stalling “Network Security Essentials
(Applications and Standards)”, Pearson
This diagonalized qwerty cipher has various Education, 2004 .
advantages over simple cipher. It is more difficult to [7]. Atul Kahate (2009), Cryptography and Network
cryptanalyze. The result cannot be easily Security, second edition, McGraw-Hill.
reconstructed. Brute force attack cannot crack the [8]. William Stallings, Cryptography and Network
code. Overcome all the limitations of the qwerty Security, 5th impression, 2008.
cipher.
Cite this Article : V. Vennila, S. Umamaheswari, "An Extension to Traditional Qwerty Cryptographic Method", International
Journal of Scientific Research in Computer Science, Engineering and Information Technology (IJSRCSEIT), ISSN : 2456-3307,
Volume 5 Issue 1, pp. 340-342, January-February 2019. Available at doi : https://doi.org/10.32628/CSEIT195151
Journal URL : http://ijsrcseit.com/CSEIT195151

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


342
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195172

A Review on Hadoop Eco System for Big Data


Anushree Raj1, Rio D’Souza2
1M.Sc. Big Data Analytics Department, St Agnes College (Autonomous) Mangalore, Karnataka, India
2Computer Science and Engineering Department, St Joseph Engineering College Mangalore, Karnataka, India

ABSTRACT

In this era of information age, a huge amount of data generates every moment through various sources. This
enormous data is beyond the processing capability of traditional data management system to manage and
analyse the data in a specified time span. This huge amount of data refers to Big Data. Big Data faces numerous
challenges in various operations on data such as capturing data, data analysis, data searching, data sharing, data
filtering etc. HADOOP has showed a big way of various enterprises for big data management. Big data hadoop
deals with the implementation of various industry use cases. To master the Apache Hadoop, we need to
understand the hadoop eco system and hadoop architecture. In this paper we brief on the Hadoop architecture
and hadoop eco system.
Keywords: Big Data, Hadoop architecture, Hadoop eco system components, HDFS, MapReduce.

I. INTRODUCTION The proposed work the second sections gives you a


detailed review on Big Data, third section speaks of
Hadoop is a open source tool. It provides an Hadoop Ecosystem and explains all of its components
efficient framework for running jobs on multiple briefly and the fourth section concludes the paper by
nodes of clusters. Cluster means a group of systems providing the main advantages of Hadoop.
connected via LAN. Apache Hadoop provides parallel
processing of data as it works on multiple machines II. BIG DATA
simultaneously [7]. Big Data and Hadoop efficiently
processes large volumes of data on a cluster of Bigdata is a term used to describe a collection of data
commodity hardware [8]. Hadoop is for processing that is huge in size and yet growing exponentially
huge volume of data. Apache Hadoop software with time [1]. Big Data generation includes stock
library is a framework that allows for the distributed exchanges, social media sites, jet engines, etc. Big
processing of large data sets across clusters of Data could be Structured, Unstructured or Semi-
computers using simple programming models. It is structured. Volume, Variety, Velocity, and
designed to scale up from single server to thousands Variability are few characteristics of Big data [2]. It
of machines, each offering local computation and plays a major in Improved customer service, better
storage. Rather than rely on hardware to deliver high operational efficiency, Better Decision Making.
availability, the library itself is designed to depict and
handle failures at the application layer, so delivering There were two major challenges with Big Data [9]:
a highly available service on the top of a cluster of
computers, each of which may be prone to failures.

CSEIT195172 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 343-348 ]
343
Anushree Raj, Rio D’Souza Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 343-348

• Big Data Storage: To store Big Data, in a HDFS:


flexible infrastructure that scales up in a cost HDFS is a distributed file system. It supports a single
effective manner, was critical. file system name space, which stores data in a
• Big Data Processing: Even if a part of Big traditional hierarchical format of directories and files.
Data is Stored, processing it would take Across an instance, data is divided into 64MB chunks
years. that are triple-mirrored across the cluster to provide
To solve the storage issue and processing issue, two resiliency. It is optimized for streaming access of
core components were created in Hadoop – large files that are in the 100s of MB upwards on
HDFS and YARN. HDFS solved the storage issue as it HDFS and access them through MapReduce to
stores the data in a distributed fashion and is easily process them in batch mode [4]. HDFS files are write
scalable [8]. YARN solved the processing issue by once files. There is no concept of random writes.
reducing the processing time drastically. HDFS doesn't do random reads.
HBase:
III. HADOOP ECHO SYSTEMS Apache HBase is a database that stores it's data in a
distributed filesystem. The filesystem of choice
Hadoop Eco System is aligned in four different layers typically is HDFS owing to the tight integration
such as Data Storage, Data processing, Data Access between HBase and HDFS. HBase provides low
and Data Management. Each layer comprises of latency access to small amounts of data from within a
different Hadoop eco system components [7]. The large data set. We can access single rows quickly
figure shows the Hadoop eco system with its from a billion row table. It provides flexible data
components. model to work with and data is indexed by the row
key. It can perform fast scans across tables [5]. It
scales in terms of writes as well as total volume of
data.

2. Data Processing Layer in Hadoop


The data processing framework is a tool used to work
with the data itself.
The data processing components of hadoop eco
system are MapReduce and YARN.
Figure Hadoop Eco Systems
MapReduce:
1. Data Storage Layer in Hadoop MapReduce is the tool that actually gets data
Hadoop aims to execute the distributed processing processed. MapReduce runs as a series of jobs, with
with the minimum latency possible. this is achieved each jobs essentially a separate java application that
by executing Map processing on the node that stores gives out into the date and starts pulling out
the data, a concept known as data locality. information as needed [6]. Hadoop is not really a
The data storage components of hadoop eco system database. It stores data and we can pull data out of it,
are the HDFS and HBASE. but there are no queries involved. Hadoop is more a
data warehousing system so it needs a system like
MapReduce to actually process the data.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


344
Anushree Raj, Rio D’Souza Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 343-348

YARN: Apache Pig is a platform for managing large sets of


YARN stands for “Yet Another Resource Negotiator”. data which consists of high-level programming to
Yarn runs a non-MapReduce jobs within the Hadoop analyze the data. Pig also consists of the
framework. YARN is a key element of the Hadoop infrastructure to evaluate the programs [7].
data processing architecture that provides different The advantages of Pig programming is that it can
data handling mechanisms, including interactive SQL easily handle parallel processes for managing very
and batch processing [3]. It improves the large amounts of data. The programming on this
performance of data processing in Hadoop by platform is basically done using the textual language
separating the resource management and scheduling Pig Latin.
capabilities of MapReduce from its data processing
component. YARN allows different data processing Mahout:
methods like graph processing, interactive processing, Apache Mahout is an open source library of scalable
stream processing as well as batch processing to run machine learning algorithms that focuses on
and process data stored in HDFS. Therefore YARN clustering, classification, and recommendations [3].
opens up Hadoop to other types of distributed The primitive features of Apache Mahout are: The
applications beyond MapReduce [5]. YARN algorithms of Mahout are written on top of Hadoop,
architecture consists of Resource management, Node so it works well in distributed environment. Mahout
management, Application master and container. uses the Apache Hadoop library to scale effectively
in the cloud. Mahout offers the coder a ready-to-use
3. Data Access Layer in Hadoop framework for doing data mining tasks on large
Hadoop can usually hold terabytes or petabytes of volumes of data. Mahout lets applications to analyze
data to process, hence Data Access is an extremely large sets of data effectively and in quick time. It
important aspect in any project or product, especially Includes several MapReduce enabled clustering
with hadoop. implementations such as k-means, fuzzy k-means,
The Data access components of hadoop eco system Canopy, Dirichlet, and Mean-Shift, Supports
are: Distributed Naive Bayes and Complementary Naive
Hive: Bayes classification implementations, Comes with
Apache Hive is a component of Hortonworks Data distributed fitness function capabilities for
Platform (HDP). Hive provides a SQL-like interface evolutionary programming, Includes matrix and
to data stored in HDP [10]. Hive gives a SQL- vector libraries.
like interface to query data stored in various
databases and file systems that integrate with Avro:
Hadoop. Hive provides the necessary SQL abstraction Avro is a remote procedure call and
to integrate SQL-like queries (HiveQL) into the data serialization framework developed within
underlying Java without the need to implement Apache's Hadoop project [10]. It is a schema-based
queries in the low-level Java API. Since most data serialization utility, which accepts schemas as input.
warehousing applications work with SQL-based In spite of various schemas being available, Avro
querying languages, Hive aids portability of SQL- follows its own standards of defining schemas like
based applications to Hadoop type of file, location of record, name of the record,
fields in the record with their corresponding data
Pig: types.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


345
Anushree Raj, Rio D’Souza Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 343-348

Sqoop: Features of Oozie are: It execute and monitor


Apache Sqoop is a tool that is extensively used to workflows in Hadoop, provides periodic scheduling
transfer large amounts of data from Hadoop to the of workflows, Trigger execution of data availability
relational database servers and vice-versa [7]. Sqoop and uses HTTP and command line interface and web
can be used to import the various types of data from console.
Oracle, MySQL and such other databases.
Sqoop needs a connector to connect the different Ambari:
relational databases. Almost all Database vendors Apache Ambari is an open source administration tool
make a JDBC connector available specific to that deployed on top of Hadoop cluster and responsible
Database, Sqoop needs a JDBC driver of the database for keeping track of running applications and their
for interaction. status. Apache Ambari can be referred to as an open
source web-based management tool that manages,
Drill: monitors and provisions the health of Hadoop
Apache Drill is a low latency distributed query clusters [3]. It provides a highly interactive dashboard
engine for large-scale datasets, including structured which allows the administrators to visualize the
and semi-structured/nested data [3]. Drill is capable progress and status of every application running over
of querying nested data in formats like JSON and the Hadoop cluster.
Parquet and performing dynamic schema discovery. Its flexible and scalable user-interface allows a range
Drill does not require a centralized metadata of tools such as Pig, MapReduce, Hive, etc. [11], to be
repository installed on the cluster and administers their
performances in a user-friendly fashion.
Thrift:
Thrift is a driver level interface that provides API for Flumes:
client implementation. It combines a software stack Apache Flume is a distributed, reliable, and available
with a code generation engine to build cross- system for efficiently collecting, aggregating and
platform services which can connect applications moving large amounts of log data from many
written in a variety of languages and frameworks [11]. different sources to a centralized data store. Suppose
there are 100 servers with unstructured log files [11].
4. Data Management Layer in Hadoop We have to load data in HDFS and analyze the same.
Hadoop data management software(entity) is for We can use flume for this purpose. It has three tiers
managing enormous amounts of scattered data. It is Agents, Collector ad storage. Flume can capture real
part of the ecosystem of modern data management time data from the server. It can sink data from logs
[11]. into HDFS. Flume is used to move the log data
The Data Management components of hadoop eco generated by application servers into HDFS at a
system are: higher speed.
Oozie:
Oozie is a workflow scheduler system to manage Zookeeper:
Apache Hadoop jobs. Oozie is integrated with the Apache Zookeeper is a coordination service for
rest of the Hadoop stack supporting several types of distributed application that enables synchronization
Hadoop jobs such as Java MapReduce, Streaming across a cluster [3]. Zookeeper in Hadoop can be
MapReduce, Pig, Hive and Sqoop [10]. Oozie is a viewed as centralized repository where distributed
scalable, reliable and extensible system. The main applications can put data and get data out of it. It is

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


346
Anushree Raj, Rio D’Souza Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 343-348

used to keep the distributed system functioning HCatalog works like a key component of Hive and it
together as a single unit, using its synchronization, enables the users to store their data in any format
serialization and coordination goals. For and any structure.
simplicity's sake Zookeeper can be thought of as a file HCatalog is a table storage management tool for
system where we have znodes that store data instead Hadoop that exposes the tabular data of Hive
of files or directories storing data. Zookeeper is metastore to other Hadoop applications. It enables
a Hadoop Admin tool used for managing the jobs in users with different data processing tools (Pig,
the cluster. MapReduce) to easily write data onto a grid [7].
HCatalog ensures that users don’t have to worry
ZooKeeper is a centralized service for maintaining about where or in what format their data is stored.
configuration information, naming, providing This is a small tutorial that explains just the basics of
distributed synchronization, and providing group HCatalog and how to use it.
services. ZooKeeper is simple, distributed, reliable,
and fast [11]. IV. CONCLUSION
• maintaining configuration information: It
maintains cluster configuration info which is Apache Hadoop is the most popular and powerful big
shared across all the nodes in the cluster. data tool, Hadoop provides world’s most reliable
• naming: Zookeeper can be used as naming storage layer – HDFS, a batch Processing engine –
service, so that one node in the cluster can find MapReduce and a Resource Management Layer –
another node in the large cluster ex: 1000 node YARN [11]. HADOOP is a highly scalable storage
cluster platform, because it can store and distribute very
• providing distributed synchronization: We can large data sets across hundreds of inexpensive servers
also use zookeeper for solving distributed that operate in parallel. Unlike traditional relational
synchronization problems in the cluster by using database systems (RDBMS) that can't scale to process
locks, Queues etc. large amounts of data, Hadoop enables businesses to
• providing group services: Zookeeper also helps in run applications on thousands of nodes involving
group service with the selection of a master in thousands of terabytes of data. HADOOP also offers a
the cluster (Leader election process). cost-effective storage solution for businesses'
exploding data sets. Hadoop enables businesses to
ZooKeeper can work in replicated mode as well as easily access new data sources and tap into different
standalone mode types of data (both structured and unstructured) to
generate value from that data [13]. This means
HCatalog: businesses can use Hadoop to derive valuable
HCatalog is a table storage management tool for business insights from data sources. A key advantage
Hadoop. It exposes the tabular data of Hive of using Hadoop is its fault tolerance. When data is
metastore to other Hadoop applications. It enables sent to an individual node, that data is also replicated
users with different data processing tools (Pig, to other nodes in the cluster, which means that in the
MapReduce) to easily write data onto a grid. It event of failure, there is another copy available for
ensures that users don’t have to worry about where use.
or in what format their data is stored [3].

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


347
Anushree Raj, Rio D’Souza Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 343-348

V. REFERENCES [9] PekkaPääkkönen, DanielPakkala1, “Reference


Architecture and Classification of Technologies,
[1] Natalia Miloslavskaya ,Alexander Tolstoy, “Big Products and Services for Big Data Systems,”
Data, Fast Data and Data Lake Concepts” 7th [10] https://intellipaat.com/tutorial/hadooptutorial/introductio n- hadoop/
Annual International Conference on [11] Apache Hadoop. http://hadoop.apache.org/
Biologically Inspired Cognitive Architectures, [12] Kala Karun. A , Chitharanjan. K,” A Review on
Volume 88, 2016, Pages 300–305 Hadoop – HDFS Infrastructure Extensions”,
[2] B. Saraladevi, N. Pazhaniraja, P. Victer Paul, Conference on Information and
M.S. Saleem Basha, P. Dhavachelvan, “Big Data Communication Technologies,2013,IEEE
and Hadoop-A Study in Security Perspective,” [13] Naveen Garg , Dr. Sanjay Singla, Dr. Surender
2nd International Symposium on Big Data and Jangra, “Challenges and Techniques for Testing
Cloud Computing (ISBCC’15), Procedia of Big Data,” Procedia Computer Science 85 (
Computer Science 50 ( 2015 ) 596 – 601 2016 ) 940 – 948
[3] DT Editorial Services, “Big Data(covers
Hadoop2, Map Reduce, Hive, Yarn, Pig, R and Cite this article as :
Data Visualization)” by Dreamtech Press
[4] “Hadoop, MapReduce and HDFS:A developer Anushree Raj, Rio D'Souza, "A Review on Hadoop
perspective,”(Procedia Computer Science, Eco System for Big Data", International Journal of
Volume 48, 2015,Pages 45-50) Scientific Research in Computer Science, Engineering
[5] A Novel and efficient de-duplication system for and Information Technology (IJSRCSEIT), ISSN :
HDFS (Procedia Computer Science,Volume 92, 2456-3307, Volume 5 Issue 1, pp. 343-348, January-
2016, Pages (498-505) February 2019. Available at doi :
[6] Tharso Ferreira, Antonio Espinosa, Juan Carlos https://doi.org/10.32628/CSEIT195172
Moure, Porfidio Hern´andez, “An Journal URL : http://ijsrcseit.com/CSEIT195172
Optimization for MapReduce Frameworks in
Multi-core,” International Conference on
Computational Science, ICCS 2013, Procedia
Computer Science 18 ( 2013 ) 2587 – 2590
[7] Can Uzunkaya, Tolga Ensari, Yusuf Kavurucu,
“Hadoop Ecosystem and Its Analysis on
Tweets,” World Conference on Technology,
Innovation and Entrepreneurship, Procedia -
Social and Behavioral Sciences 195 (2015 ) 1890
– 1897
[8] Sachin Bende, Rajashree Shedge, “Dealing with
Small Files Problem in Hadoop Distributed File
System,” 7th International Conference on
Communication, Computing and Virtualization
2016, Procedia Computer Science 79 ( 2016 )
1001 – 1012

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


348
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195186

A Novel Approach on Prevention of a Mobile Exploration in


Machine Using AI
1J.Antony Daniel Rex, 2S.Dinakaran & 3S.Muthukumaran
1,3 Assistant Professor, PG and Reseach Department of Computer Science, St.Joseph College of Arts & Science
(Autonomous) Cuddalore, Tamilnadu, India
2Ph. D. Scholar, Anna University, Salem, Tamilnadu, India

ABSTRACT

This paper deals with a low cost solution to problem avoidance for a mobile machine using just a single Artifical
Intelligennce. It allows the machine to navigate smoothly in an unknown environment, avoiding collisions,
without having to stop in front of problems. The problem avoidance process is made up of three distinct stages -
the mapping algorithm, the core problem avoidance algorithm, and the steering algorithm. The mapping
algorithm takes the raw Artifical Intelligennce readings and processes them to create higher resolution maps
from the wide-angle Artifical Intelligennce. The problem avoidance algorithm is based on the potential field
theory which considers the machine to be a test charge that is repelled by all the problems around it, and
which moves in the direction of the resultant of the forces acting on it. An algorithm which steers a mobile
machine based on the differential drive system is also discussed.
Keywords : Mobile Machine, Problem Avoidance, Autonomous Navigation, Ultrasonic Range Sensor.

I. INTRODUCTION consist of a combination of an infrared light emitting


device and an infrared light sensor. These sensors are
Problem avoidance is a primary requirement for any merely proximity detectors, they cannot determine
autonomous mobile machine. The machine acquires the range of the problem in front. The range of these
information about its surroundings through sensors sensors is also limited to a maximum of 80cm.
mounted on the machine. Various types of sensors
can be used for problem avoidance, as detailed below. Ultrasonic range sensors, which determine the range
of the object in front of it. They work by sending a
A. Types of Sensors: short burst of ultrasonic waves, and measuring the
Bump sensors, which are micro-switches activated time taken for the echo to be received. They have a
when the machine touches an problem. This is a wide beam angle, typically 30. These sensors have
simple, inexpensive method of problem detection, ranges of up to 6m.
but operates only on contact, which makes it useful
only for slow moving machines. Laser range finders, which work on the same
principle as ultrasonic range sensors, except that they
Infrared proximity detectors, which detect the use LASER instead of ultrasound. Laser range finders
presence of an object in front of the sensor. They

CSEIT195186 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 349-356 ]
349
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

have a range of upto 30m, and are very accurate, Beyond the range value, there may be problems
having an angular resolution of upto 0.25. somewhere along the arc, but it is not certain at
which positions the problems are present.
However, they are very expensive compared to other Figure 1 shows the next sensor reading, which is
sensors of these, the ultrasonic range sensor was obtained after rotating the sensor through a certain
found most suitable for our requirement because of angle, say 5. This reading, with a range value of 1.2m
its low cost and ranging capability. covers angles -10 to +20. After obtaining this reading,
we are certain that there is no problem between
II. MAPPING these angles for a range of 1.2m. Thus, in the
overlapping region, the larger distance value is
An ultrasonic range sensor detects the range of an chosen over the smaller one.
object in front of it by sending a short burst of
ultrasonic waves, and measuring the time taken for If a subsequent overlapped reading is smaller than the
the echo to be received. These sensors typically have existing value, then we have to consider the ‘age’ of
a wide beam angle of 15 to 30 As a result, the cone the existing information. If the existing large range
gets wider with increasing distance from the value is very recent, then the overlapped area is taken
transmitter. This increases the uncertainty in the to have the larger range, because it can be assumed
spatial dimensions of the problem. Most mobile that no object moved into the intersecting area
machines which use ultrasonic range finders for between the two scans. If the existing large range
problem avoidance have 12 or 24 sensors in a ring value is old, then the smaller value is taken, because
around the machine, to cover all directions. This is the machine may have moved within that time, or a
not very cost-effective, and moreover, all of them moving problem may have appeared.
cannot be triggered at once due to crosstalk and
interference problems. A maximum of 4 sensors Using this technique, the accuracy of the ultrasonic
which are at 90 to each other can be triggered at a range information can be greatly improved upto the
time. Our solution was to use a single sensor mounted step size of the stepper motor. However, there are
on a stepper motor to pan the sensor across 180. This some cases where the information is still not accurate.
can be expanded to 4 sensors mounted at 90 to each Consider the case gap between problems A and B can
other on the stepper motor for maximum scanning never be detected by the sensor at that distance.
speed and full coverage. Such a setup also provides Once the machine moves closer to these problems,
the flexibility to perform fine scanning of an area for the gap can be detected.
a more accurate map of the environment, with an
angular resolution of the minimum step size of the IV. PROBLEM AVOIDANCE
stepper motor, as opposed to the 30 resolution of the
sensor. All mobile machines feature some kind of problem
avoidance. The algorithm could be as simple as
III. MACHINE SCANNING ALGORITHM stopping if an problem is detected in front of it. This
is common in track-following machines, where the
A single range reading from the sensor implies that machine follows a pre-determined track. Another
the area in a 30 sector in front of it with a radius of solution is wall-following, in which the machine
the range value contains no problems. A reading of a moves alongside a wall at a fixed distance, following
range of 1m. This reading covers angles -15 to +15. the contour of the wall till it reaches its destination.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


350
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

In the case of an exploration machine, the


environment is unknown and problem avoidance h as
to be dynamic in nature. Our approach is an
adaptation of the potential field method or VFF
method in which the machine is considered as a test
charge which is repelled by problems which are
considered as like charges. Section II explains our
mapping methods and an algorithm to obtain
accurate data from the wide angle ultrasonic range
sensor. Our potential field method adaptation is
discussed steering control for the differential drive
Figure 1 : The second reading from the ultrasonic
machine based on the output of the potential field
range finder overlapping the first
algorithm is described.
The machine is then steered in the direction of the
resultant of all the forces added up. This results in
Information about the ranges of problems around the
motion of the machine similar to a charge moving in
machine obtained from the mapping technique is
a potential field - avoiding like charges and moving
stored in a table which is continuously updated. This
towards the unlike charge. For the case of an
range information is then used to calculate a
exploration machine without a target, the algorithm
destination direction which the machine must move
has to be slightly modified. A threshold limit is set
towards. This calculation is performed using our
beyond which the charge assigned to the range is
problem avoidance algorithm based on the potential
positive and attracts the machine. Thus, problems
field method.In the potential field method, the
which are close by to the machine will repel it, and
machine is considered as a test negative charge. All
the free spaces attract the machine.
the problems are considered to be negative charges
which repel the machine, and the target is assigned a
positive charge which attracts the machine. The
forces of all these charges on the machine are This is the basic equation representing the force
calculated in accordance with Coulomb’s law, i.e. the exerted on the machine by a single problem at a
force is proportional to the charge and inversely distance r and an angle from the machine. k is a
proportional to the square of the distance. numerical constant selected to bring the force into
the required range. The charge q is not present in this
equation because different angles do not have
different charge values. This equation is modified to
make near ranges repel the machine and far ranges
attract, as shown in the following equation.

Here, t is a threshold distance at which the problem


neither attracts nor repels the machine, and ranges
beyond the threshold distance are attractive. The

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


351
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

total force on the machine is obtained by vectorial


addition of all the individual forces.

This is the expression for the resultant of all the


Figure 2.1 : Unit circle in which all possible
forces acting on the machine. The machine is steered
force vectors lie
in this direction in order to avoid problems. The
magnitude of the vector can be used to control the
On this circle, a point at 90 indicates that the
speed of the machine. This results in slow movement
machine has to move straight, a point at 0 indicates
in a cluttered environment, and faster motion when
that the machine has to turn right in-place, and at all
there are fewer problems surrounding the machine.
points in between, it has to exhibit varying degrees of
turning.
V. STEERING CONTROL FOR A DIFFERENTIAL
DRIVE SYSTEM

A mobile machine which uses differential drive has


separate speed and direction control for the left and
right sets of wheels. Such a system is very flexible for
a mobile machine because it does not have a
minimum turning radius, it can execute a turn
around its own center. Arc turning is achieved by
driving the left and right wheels at different speeds.
Driving one set of wheels forward and the other set
backward results in in-place turning.
Figure 2.2 : The unit circle shrunk into an ellipse

An algorithm was designed to convert the force


This unit circle is shrunk into an ellipse as shown in
vector obtained from the problem avoidance
Figure 2.2, by dividing the x coordinate by a constant.
algorithm into the left and right wheel speeds
This ensures that the wheel speeds of the machine
required for the differential drive system. The
are lower while turning and faster when moving
magnitude and direction of the force vector is plotted
straight. Turning at high speeds can result in slipping
as a point on a graph, after being normalized so that
and inaccurate turns, this prevents it from happening.
the maximum possible magnitude is 1. Thus, all
possible force vectors will lie in a unit circle around
the origin, as shown in Figure 2.1.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


352
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

The mobile machine we designed and constructed,


and implemented the problem avoidance algorithms
on. It uses a differential drive system for mobility, but
uses tracks instead of wheels, like a tank. The drive
motors used are stepper motors. It has a single
ultrasonic range sensor mounted on a stepper motor
in front which can pan across an angle of 180 in steps
as small as 0.45. The machine is controlled by a
computer through the parallel port connected to the
required driver circuits on the machine itself.

RTLinux was chosen as the software platform because


Figure 2.3 : Rotation of the force vector
real time functionality was required, especially to
measure the sonar range timing, and to trigger the
Now the existing force vector is rotated clockwise by
stepper motors. The default interval at which
45, as indicated in Figure 2.3. The new coordinates of
ultrasonic range readings are taken is 15, which
the vector are then taken as the left and right wheel
provides 12 readings over 180. Each reading overlaps
speeds respectively (after normalization). This result
the previous by 15. This provides for fast scanning of
in the machine turning in the required direction
the environment -- an average of 500ms to scan 180,
based on the direction of the original force vector.
and thus the machine can move at high speeds
This dynamic steering algorithm ensures that the
without danger of collision. If an area needs to be
machine doesn't have to stop in front of an problem
mapped more accurately, or in case of a highly
to go around it, instead it navigates smoothly to the
cluttered environment, the step size is reduced and
required direction.
the maps obtained are of higher resolution, which
can help in navigating precisely through smaller
The force vector is updated every time new range
pathways.
information is obtained,which is approximately every
20 or 30 milliseconds, and the wheel speeds are also
VII. CONCLUSION
correspondingly modified every time. Modifications
in the wheel speeds will not be abrupt, because every
This traveler machine can explore in any unknown
reading changes the range associated with only a
environment smoothly avoiding problems. Thus it
narrow angle, resulting in a very small change in the
can be used even in exploring other planets using
resultant force vector. Since the wheel speeds vary
very precise sensors. This method of avoiding
smoothly according to the sensor information, and
problems is very useful in army, military applications.
the speeds are updated almost in real time, the
These kinds of machines hold the technology for the
machine follows a smooth course around the
coming future.
problems present in the environment, while still
being able to respond immediately to changes in the
VIII. REFERENCES
environment like dynamic problems.

[1] “Sensors Handbook”-Solaman, Sabrie.


VI. EXPERIMENTAL RESULTS

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


353
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

[2] “Sensory based motion planning”-IEEE Computer Networking, Wireless and Mobile
transactions on Machineics & Automation- Communications. Vol No – 3, Issue No - 1, pp.
Kamon Ishay. 245-256.
[3] A.S.Syed Navaz, C.Prabhadevi & [11] A.S.Syed Navaz, H.Iyyappa Narayanan &
V.Sangeetha”Data Grid Concepts for Data R.Vinoth.” Security Protocol Review Method
Security in Distributed Computing” January Analyzer (SPRMAN)”, August – 2013,
2013, International Journal of Computer International Journal of Advanced Studies in
Applications, Vol 61 – No 13, pp 6-11. Computers, Science and Engineering, Vol No –
[4] A.S.Syed Navaz, V.Sangeetha & C.Prabhadevi, 2, Issue No – 4, pp. 53-58.
“Entropy Based Anomaly Detection System to [12] A.S.Syed Navaz, G.M. Kadhar Nawaz &
Prevent DDoS Attacks in Cloud” January 2013, B.Karthick.” Probabilistic Approach to Locate a
International Journal of Computer Mobile Node in Wireless Domain”, April –
Applications, Vol 62 – No 15, pp 42-47. 2014, International Journal of Computer
[5] Y. Kim, H. Shin, and H. Cha, “Y-mac: An Engineering and Applications, Vol No – 6, Issue
energy-efficient multichannel mac protocol for No – 1, pp.41-49.
dense wireless sensor networks,” in IPSN ’08, [13] S. Gandham, Y. Zhang, and Q. Huang,
April 2008, pp. 53–63. Distributed time-optimal scheduling
[6] A.S.Syed Navaz, A.S.Syed Fiaz, C.Prabhadevi, forconvergecast in wireless sensor networks,”
V.Sangeetha & S.Gopalakrishnan “Human Computer Networks, vol. 52, no. 3, pp. 610–
Resource Management System” Jan – Feb 2013, 629, 2008.
International Organization of Scientific [14] A.S.Syed Navaz & G.M. Kadhar Nawaz, “Ultra
Research Journal of Computer Engineering, Vol Wideband on High Speed Wireless Personal
8, Issue 4, pp. 62-71. Area Networks” August – 2014, International
[7] A.S.Syed Navaz, S.Gopalakrishnan & R.Meena Journal of Science and Research, Vol No – 3,
“Anomaly Detections in Internet Using Issue No – 8, pp.1952-1955.
Empirical Measures” February 2013, [15] A.S.Syed Navaz & Dr.G.M. Kadhar Nawaz &
International Journal of Innovative Technology A.S.Syed Fiaz “Slot Assignment Using FSA and
and Exploring Engineering, Vol 2 – Issue 3. pp. DSA Algorithm in Wireless Sensor Network”
58-61. October – 2014, Australian Journal of Basic and
[8] “Problem detection based on qualitative Applied Sciences, Vol No –8, Issue No –16,
&quantitative 3d reconstruction-IEEE pp.11-17.
transaction on pattern analysis and [16] A.S.Syed Navaz & A.S.Syed Fiaz, “Load
intelligence”-Zhang-Zhonggei. Balancing in P2P Networks using Random
[9] A.S.Syed Navaz, M.Ravi & T.Prabhu, Walk Algorithm” March – 2015, International
“Preventing Disclosure of Sensitive Knowledge Journal of Science and Research, Vol No – 4,
by Hiding Inference” February 2013, Issue No – 3, pp.2062-2066.
International Journal of Computer [17] Qiao Li, Baihai Zhang, Lingguo Cui, Zhun Fan,
Applications, Vol 63 – No 1. pp. 32-38. V. Vasilakos Athanasios, Epidemics on small
[10] A.S.Syed Navaz , T.Dhevisri & Pratap worlds of tree-based wireless sensor networks,
Mazumder “Face Recognition Using Principal Journal of Systems Science and Complexity
Component Analysis and Neural Networks“ December 2014, Volume 27, Issue 6, pp 1095-
March -2013, International Journal of 1120.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


354
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

[18] A.S.Syed Navaz, J.Antony Daniel Rex, S.Jensy and Research, Vol No - 5, Issue No - 1, pp.1581-
Mary. “Cluster Based Secure Data Transmission 1584.
in WSN” July – 2015, International Journal of [26] A.S.Syed Fiaz, N.Asha, D.Sumathi & A.S.Syed
Scientific & Engineering Research, Vol No - 6, Navaz “Data Visualization: Enhancing Big Data
Issue No - 7, pp. 1776 – 1781. More Adaptable and Valuable” February –
[19] Ricardo Severino, Nuno Pereira, Eduardo 2016, International Journal of Applied
Tovar, Dynamic cluster scheduling for cluster- Engineering Research, Vol No - 11, Issue No -
tree WSNs, Springer, Vol.3, No.493, 2014. 4, pp.–2801-2804.
[20] A.S.Syed Navaz, J.Antony Daniel Rex, P.Anjala [27] W. Song, F. Yuan, and R. LaHusen, “Time-
Mary. “An Efficient Intrusion Detection optimum packet scheduling for many-to-one
Scheme for Mitigating Nodes Using Data routing in wireless sensor networks,” in MASS
Aggregation in Delay Tolerant Network” ’06, pp. 81–90.
September – 2015, International Journal of [28] A.S.Syed Navaz & Dr.G.M. Kadhar Nawaz
Scientific & Engineering Research, Vol No - 6, “Flow Based Layer Selection Algorithm for
Issue No - 9, pp. 421 – 428. Data Collection in Tree Structure Wireless
[21] A.S.Syed Navaz, P.Jayalakshmi, N.Asha. Sensor Networks” March – 2016, International
“Optimization of Real-Time Video Over 3G Journal of Applied Engineering Research, Vol
Wireless Networks” September – 2015, No - 11, Issue No - 5, pp.–3359-3363.
International Journal of Applied Engineering [29] A.S.Syed Navaz & Dr.G.M. Kadhar Nawaz
Research, Vol No - 10, Issue No - 18, pp. 39724 “Layer Orient Time Domain Density Estimation
– 39730. Technique Based Channel Assignment in Tree
[22] M.Ravi & A.S.Syed Navaz. “Development of Structure Wireless Sensor Networks for Fast
Mail Tracking System” October – 2015, Data Collection” June - 2016, International
International Journal of Innovative Research in Journal of Engineering and Technology, Vol No
Computer and Communication Engineering, - 8, Issue No - 3, pp.–1506-1512.
Vol No - 3, Issue No - 10, pp.9757-9762. [30] M.Ravi & A.S.Syed Navaz "Rough Set Based
[23] S.Jensy Mary, A.S Syed Navaz & J.Antony Grid Computing Service in Wireless Network"
Daniel Rex, “QA Generation Using Multimedia November - 2016, International Research
Based Harvesting Web Information” November Journal of Engineering and Technology, Vol No
– 2015, International Journal of Innovative - 3, Issue No - 11, pp.1122– 1126.
Research in Computer and Communication [31] A.S.Syed Navaz, N.Asha & D.Sumathi “Energy
Engineering, Vol No - 3, Issue No - 11, Efficient Consumption for Quality Based Sleep
pp.10381-10386. Scheduling in Wireless Sensor Networks”
[24] A.S Syed Navaz & A.S.Syed Fiaz “Network March - 2017, ARPN Journal of Engineering
Intelligent Agent for Collision Detection with and Applied Sciences, Vol No - 12, Issue No - 5,
Bandwidth Calculation” December – 2015, pp.–1494-1498.
MCAS Journal of Research, Vol No – 2, pp.88- [32] N. Lai, C.King, and C. Lin, “On Maximizing the
95, ISSN: 2454-115X. throughput of convergecast in wireless sensor
[25] A.S Syed Navaz & K.Durairaj “Signature networks,” in GPC ’08, pp. 396– 408.
Authentication Using Biometric Methods” [33] A.S.Syed Fiaz, K.S.Guruprakash & A.S.Syed
January – 2016, International Journal of Science Navaz “Prediction of Best Cloud Service
Provider using the QoS Ranking Framework”

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


355
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

January – 2018, International Journal of InternationalJournal of Computer Science


Engineering & Technology, Vol No - 7, Issue Engineering and Information Technology
No -1.1, pp.– 486-488. Research, Vol No: 3; Issue No: 1, pp. 257-266.
[34] A.S.Syed Navaz, Asha.N, Vanmathi [41] A.S.Syed Fiaz, N.Devi, S.Aarthi "Bug Tracking
Chandrasekaran & J.Jayashree “Resourceful and Reporting System", March 2013,
Investigate Encryption Method Using Data International Journal of Soft Computing and
Hunt in Mobile Cloud Service”, August – 2018, Engineering, Vol No: 3; Issue No: 1, pp. 257-
ARPN Journal of Engineering and Applied 266.
Sciences, Vol No - 13, Issue No - 15, pp.–4543- [42] M. Usha, J. Akilandeswari and A.S.Syed Fiaz
4549. “An efficient QoS framework for Cloud
[35] Asha.N, A.S.Syed Navaz, J.Jayashree & Brokerage Services”, Dec. 2012, International
J.Vijayashree “RFID Based Automated Gate Symposium on Cloud and Service Computing,
Security System”, November – 2018, ARPN pp: 76-79, 17-18, IEEE Xplore.
Journal of Engineering and Applied Sciences,
Vol No - 13 , Issue No - 22, pp.– 8901 – 8906.
[36] A Xiaojun Zhu, Xiaobing Wu, Guihai Chen, An
exact algorithm for maximum lifetime data
gathering tree without aggregation in wireless
sensor networks, Springer, Wireless Networks
2014.
[37] .S. Syed Navaz, N. Asha & Kuna
Venkateswararao “Face Recognition and
Detection with New Image Enhancement
Techniques” December – 2018, Journal of Cite this article as :
Advanced Research in Dynamical and Control
Systems, Vol No -10 , Issue No – 14 (Special J.Antony Daniel Rex, S.Dinakaran, S. Muthukumaran,
Issue), pp.– 760-764. "A Novel Approach on Prevention of a Mobile
[38] A.S.Syed Fiaz, I.Alsheba & R.Meena “Using Exploration in Machine Using AI ", International
Neural Networks to Create an Adaptive Journal of Scientific Research in Computer Science,
Character Recognition System”, Sep 2015, Engineering and Information Technology
Discovery - The International Daily journal, (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
Vol.37 (168), pp.53-58. 349-356, January-February 2019. Available at doi :
[39] A.S.Syed Fiaz, M. Usha and J. Akilandeswari “A https://doi.org/10.32628/CSEIT195186
Brokerage Service Model for QoS support in Journal URL : http://ijsrcseit.com/CSEIT195186
Inter-Cloud Environment“,March 2013,
International Journal of Information and
Computation Technology, Vol.3, No.3, pp 257-
260,
[40] A.S.Syed Fiaz, R.Pushpapriya, S.Kirubashini &
M.Sathya “Generation and allocation of
subscriber numbers for
telecommunication“,March2013,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


356
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT183857

Tuition Plus+ For Managing Tuition Class Data


Bharat Dhak*1, Prachi P. Paunikar2, Simran S. Khan3, Snehal S. Ramteke4, Yashkumar R. Jais5, Priya R. Gupta6
*1Lecturer, Computer Science and Engineering Department, Priyadarshini J. L. College of Engineering, Nagpur,
Maharashtra, India
2-6 Computer Science and Engineering Department, Priyadarshini J. L. College of Engineering, Nagpur,
Maharashtra, India

ABSTRACT

Tuition Plus+ is a website and a web-based application provides a simple interface for maintenance of a student
records and information. It is used by educational institution like tuitions or colleges which can manage
student’s records easily. Tuition Plus website deals with different modules like registration, organization,
Subjects, Course, Batches, Student, Account, Enquiry and Report Module. It tracks all the details of a student
from a day to day course. It will also track all the students fees structure like installments. These will be
available through a secure, online interface embedded in the Tuition Plus website. It will also have registration
details, batch execution details, students’ details in all aspects, the various academic notifications to the staff and
students updated by the tuition organisation
Keywords : Tuition Plus+, DFD, OTP

I. INTRODUCTION the Tuition administrator and ensures highest


possible level of security. The system features a
[1]The design and implementation of a complex logging system to track all users access and
comprehensive student information system and user ensure conformity to data access guidelines and is
interface is to replace the current paper records. expected to increase the efficiency of the Tuition
Tuition Staff are able to directly access all aspects of a record management thereby decreasing the work
student’s academic progress through a secure, online hours needed to access and deliver student records.
interface embedded in the Tuition Plus+ website. The
system utilizes user authentication, displaying only One of the major reasons of developing Tuition Plus+
information necessary for an individual’s duties. is to replace the existing system with computerized
Additionally, each sub-system has authentication system. The staff of the organisation can directly
allowing authorized users to create or update access the details of the student, regarding their fees
information in that sub-system. All data is structure and payment using online interface. The
thoroughly reviewed and validated on the server website consists of authentication system such as
before actual record alteration occurs. In addition to a student login and merchant login. Using merchant
staff user interface, the system plans for student user login, tuition administrator can access the
interface, allowing users to access information and information of student. The data will be securely
submit requests online thus reducing processing time. saved in the SQL server which is managed by the
All data is stored securely on SQL servers managed by tuition administrator. Only the head administrator

CSEIT195197 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 357-362 ]
357
Bharat Dhak et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 357-362

will have the authority to update or edit the easy and interactive manner which provides facilities
information of student as well as the information of like online registration and profile creation of
their organisation. Student login will help the student’s thus reducing paper work and automating
students to see their own data, fees structure, etc. the record generation process in an educational
organization, which will also save the time of a
[1]Previously, the Tuition relied heavily on paper student. This website also provides a common
records for this initiative. While paper records are a platform for registering their organization online that
traditional way of managing student data there are is why the need of making separate website by
several drawbacks to this method. Like when a investing money will reduce.
student's wants to register for a particular course then
it must have to visit the particular tuition class and III. PROPOSED SYSTEM
register themselves. It takes a very long time to
convey the information to the student. Paper records The purpose is to design a Tuition Plus+ website,
are difficult to manage and track. The physical which contains up to date information of an
exertion required to retrieve, alter, and re-file the educational organization that should improve
paper records are all non-value added activities. This efficiency of a tuition record management and about
system provides a simple interface for the a different courses that they will provide to the
maintenance of student information. It can be used student.
by educational institutes or colleges to maintain the
records of students easily. Achieving this objective is OBJECTIVES
difficult using a manual system as the information is
scattered, can be redundant and collecting relevant • Provide the online interface for a student’s
information may be very time consuming. All these information, courses, fees structure, etc.
problems are solved using Tuition Plus+. The paper • Increase the efficiency of tuition record
focuses on presenting information in an easy and management.
interact manner which provides facilities like online • Decrease time required to access and
registration and profile creation of student’s thus delivered records.
reducing paper work and automating the record • To make a system more secure.
generation process in an educational institution. • Decrease time spent on non-value added
tasks.
II. RELATED WORK

SYSTEM DESIGN
The existing system is web based student information
management is also good but that have certain
A. DATA FLOW DIAGRAM
drawback like these existing website is only for a
Data Flow Diagram shows the graphical
college but our system is for all educational
representation of the flow of Tuition Plus+. It shows
institution .Also the existing system does not provide
the visualization of processing of data. It shows how
any information about a fees structure but that is
users of the Tuition Plus+ will interact with the
included in our system .
system. DFD contains the different modules that can
be viewed as a function that is used to transform the
The purpose is to design a Tuition Plus+ website that,
these website focuses on presenting information in an

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


358
Bharat Dhak et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 357-362

given input into the desired output. Different


modules are as follows:

a. Registration Module
b. Organization Setting Module
c. Subject Module
d. Course Module
e. Batch Module
f. Student Module
g. Account Module
h. Enquiry Module
i. Report Module
There are two different login because both the users
have different level of authorization to the system.
Merchants have the authority to access the data of
the students in their organization whereas users have
the authority to access their own data.

A. Registration Module

Merchant User will fill this registration form on


submit it will send OTP to user mobile. Users have to
fill the registration form the email and password
entered in this form will be used further for login in
B. Detailed Structure
Tuition Plus+ software. Users have to enter OTP
which sent by Tuition Plus+ on entered mobile
The system has two logins i.e. User Login and
number in input given in above screen. Users can
Merchant Login. Merchant Login is the owner of the
create their own workspace it will generate their
company, which do overall setup of the organization
own unique URL, which they can use further to
and User Login, is the individual login of the student
access Tuition Plus+. Users have to enter their
to check their personal details.
workspace to login in Tuition Plus+ to access their
organization login.

B. Organization Setting Module

This organization setting is done by the merchants.


Organization must setup their information like they
fill their complete details. They have to put their
financial approach. The branch is specifies if it has
number of branches. After login user have to setup
their organization by putting all essential information
in form user can upload logo of their organization it
will appear in students receipt. User can also create

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


359
Bharat Dhak et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 357-362

branches of their organization by clicking manage batch for subjects one or more subjects all students in
branch tab. It will set company as default primary that batch will show in batch list.
branch user can change primary branch. User can allocate batch to one student from student’s
profile.
C. Subject Module
It has two options batch for course and batch for
Merchant Login have the authentication to add subjects.
subject that are taught in their organization. User can 1) Batch for course:
create subjects from add button in subject It will show all batches, which have all
module.Once subjects are created user can create subjects of that course.
course through course panel. User can edit or delete 2) Batch for subjects:
subjects. Add subject for particular branch, which are It will show batch for every subjects of that
already created from organization setting. User can course.
select more than one branch to assign same subject in
multiple branches. F. Student Module

The student is of centre focus, because in every


D. Course Module
organization student plays the very important role.
Once subjects are, created user can create course User Login is the authorization, which is given to the
through course panel. Add course for particular student. In this module student can view their own
branch that are already created from organization information but cannot edit it. The information,
setting. User can select more than one branch to which is shown in this module, is the personal detail
assign same course in multiple branches. of the students as well as their payment detail.

E. Batch Module G. Account Module

Batch module has two steps to create batch In account, module user can manage fees, duration,
1) Create batch: tax and discount for courses.
2) This tab includes basic information about In this student, can able to pay in three different
batch name, batch capacity, start date, end ways:
date, start time, end time. A. Onetime payment
3) Batch subjects B. Instalment payment
In batch subjects user have to assign one or C. Monthly payment
more subjects in batch with trainer name In Payment Management, tab user can manage
subject start and end time, start and end date. course payment methods, instalment amount and
User can edit batch or can view batch. schedule of instalment.
1) One Time :
View batch In Onetime, payment method complete fees
It will show all details fill during batch creation. User should be paid at a time.
can add more subjects after creating batch from 2) Installment:
assign subject’s button. In Batch student, tab user can In installment user can create any number of
assign multiple students to single batch for separate installments and can schedule duration foe
subjects or for all subjects. After assigning students to that installment

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


360
Bharat Dhak et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 357-362

3) Monthly: more things automatically in less time. Tuition Plus+


In monthly installment user can create allows you to manage staff data and student data in
installment for per month. an observatory format that ensures data completeness,
accurate reporting and quick access to information. It
Payment details will show on print and on submit
is the Better, Smarter, Efficient and Very Simple Way
registration students will get conformation through
to Manage the amount of data at one Place. Tuition
SMS and email with registration receipt attachment
Plus is designed to help you and your Student to
in email.
develop the way of working or the use of Tuition
Plus makes you more smart for managing the
H. Enquiry Module
thousand of data of students which none can do.
This module will help organization to maintain all
V. REFERENCES
enquiries in records so that no any enquiry will miss
by any user in organization so the enquiry module
[1]. Web based student information management
will help to connect with the different interested
system S.R.Bharamagoudar1, Geeta R.B.2,
students. After filling enquiry form user can save
S.G.Totad3 Assistant Professor, Dept. of
enquiry it will reflect in enquiry list.
Electronics & Communication Engg,
Basaveshwar Engg. College, Bagalkot,
After saving enquiry user can add next follow-up for
Karnataka1 Associate professor, Department of
that same enquiry and can add next follow-up for
IT, GMR Institute of Technology, RAJAM,
that Enquiry.
Andhra Pradesh2 Professor, Department of
Enquiry record generated by
Computer Science & Engineering, GMR
1. Website
Institute of Technology, RAJAM, Andhra
2. Newspaper
Pradesh 3
3. Social Media
[2]. G.Satyanarayana Reddy,Rallabandi
4. Friends
Srinivasu,Srikanth Reddy Rikkula,Vuda
Sreenivasa Rao," Management Information
I. Report Module
System To Help Managers For Providing
Report module gives the answer of following Decision Making In An Organization",
questions: International Journal of Reviews in Computing,
1. How much student came for enquiry? ISSN: 2076-3328, EISSN: 2076-3336
2. How much student is interested to take [3]. Bishnu Prasad Gautam, Shree Krishna
admission? Shrestha,"Effective Campus Management
3. From where mostly the crowd came. through Web Enabled Campus-SIA (Student
4. Mode of attraction. Information Application)", Proceeding Of the
International MultiConference of Engineers
IV. CONCLUSION and Computer Scientists 2012 Vol I, IMECS
2012, March 14-16, 2012, Hong Kong
We have developed this product to managing all data
[4]. Design and implementation of student's
with less paper work and to maintain all details easily.
information management system(SIM) based on
In Tuition Plus+ our main objective is that tuition
Active Server Page(ASP) College of Electrical
classes can see their graph and they can able to do
Engineering ,Northwest University for

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


361
Bharat Dhak et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 357-362

Nationalities, Lanzhou Jiatong University,


Chaina yangchenghui36@163.com
[5]. Unnati A. Patel, Dr. Swaminarayan Priya R. ,"
Development of a Student Attendance
anagement System Using RFID and Face
Recognition: A Review", International journal
Of Advance Research In Computer Science and
Management Studies,Volume2, ISSN: 232
7782,Issue 8,August 2014
[6]. Unnati A. Patel," Student Management System
based on RFID Technology", International
Journal of Emerging Trends & Technology in
Computer Science (IJETTCS), Volume 2, Issue
6, ISSN 2278-6856, November – December
2013
[7]. Mayuri Kamble et al, / (IJCSIT) International
Journal of Computer Science and Information
Technologies, Vol. 6 (2), 2015, 1341-1344

Cite this article as :

Bharat Dhak, Prachi P. Paunikar, Simran S. Khan,


Snehal S. Ramteke, Yashkumar R. Jais, Priya R. Gupta,
"Tuition Plus+ For Managing Tuition Class Data ",
International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 357-362, January-February 2019.
Journal URL : http://ijsrcseit.com/CSEIT195197

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


362
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195195

Heikin-Ashi Transformation and Vix Index data for Stock Market


Index Prediction and It’s Effects
Nishchal Sharma*1, Dr. Chaman Singh Chauhan2
*1Research Scholar, School of Engineering and Technology, Career Point University, Kota, Rajasthan, India
2Assistant Professor, Dept. of Computer Science & Application, Govt. College, Chamba, Himachal Pradesh,
India

ABSTRACT

This paper performs a comprehensive analysis of Vix Index data with Heikin Ashi Transformation of stock
index Neural Network Learning. It has been demonstrated that Heikin Ashi Transformation can improve the
learning effect of Neural Network and the effect can also be filter out if volume weights are also considered.
This paper introduces another improvement beside using volume-weighted data. Instead volatility index is used
as an input and its effect for neural network learning process is analyzed.
Keywords : Heikin Ashi Transformation, Vix Data, Neural Network, Stock Market Analysis, Time series
filtering

I. INTRODUCTION like human beings. If a large training data is available


then neural network can learn it properties and
Predicting uncertain events have been a curious make prediction based on the learning [1] . One of
subject for human beings. we were able to make the most desired fields of such research includes stock
better prediction framework with the computation market. Stock market is a very dynamic place with
power we have. The science of examining tons of information. Stock market data creates
information and its effects is extremely valuable. The patterns, that when harvested can provide better
information itself is of no use if it is raw and decision making opportunity. But stock market data
unprocessed. The process of locating hidden patterns is full of unwanted entropy. Such noise can be a short
or data inside can be very useful. High power term fluctuation in stock market data or may be long
machine can turn this data into recognizable patterns term disturbence. Such Noise is undesired within the
far easily now. Identification of patterns inside given data and may or may not be related to the underlying
data is a key variable in modern research. Since we stock data trend. We have already demonstrated that
have abundance of computational power available to such noise can interfere with learning of neural
us, we would formulate techniques that enable network and hence reduces the effectiveness of
machine to analyse data patterns and discover learning and its effectiveness. The Heikin-Ashi
underlying information easily. transformation has worked remarkably well reducing
noise and its effect can be further improved using Vix
Neural network are system that mimic human brain Index data.
and allow machine to learn underlying patterns just

CSEIT195195 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 363-365 ]
363
Nishchal Sharma, Dr. Chaman Singh Chauahn Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 363-365

Some Researchers [2] have demonstrated the HA-Close = (Ocurrent + Hcurrent +Lcurrent+Ccurrent)/ 4
importance of availability of favorable methods for HA-Open = (HA-OpenPrev + HA-ClosePrev) / 2
finance managers and incorporating many systems HA-High = Maximum of the Hcurrent , HA-Open or HA-Close
for portfolio management. Such connection of HA-Low = Minimum of the LCurrent , HA-Open or HA-Close
selection of technology and company is vital [3] for a
finance driven system. Artificial Neural network has Using Vix Data:
been a very important learning system [4] and it has
provided a definite edge for it’s user. The effect of Vix is a volatility index and tracks volatility measured
neural network and genetic algorithm has been as per the options of stock index. Since the value of
researched [5] and found to be very effective. Neural Vix may very between some fraction value, it is
network for trend determination [6] and optimized converted between [0,1.0] using inversion.
neural network [7] are a few key methods to work
with. Many researchers have incorporated hybrid A 2 hidden layer ANN is used with configuration
models [8] and had early success. Generating buy and 5,6,3,1. The choice of structure of ANN is quite
sell signal using deep neural network [9] as well as a subjective. This setup given relatively stable results.
self organizing rule based system [10] perform very Again test data is converted into 2 sets. One is for
well. Effect and result of ANN on NASDAQ learning and other for test results.
Composite [11] and results are inline with
expectations. III. RESULTS AND DISCUSSION

II. METHODS AND MATERIAL A Plot of Prediction by ANN and its original value
provides us with an absolute error. This absolute
Heikin-Ashi is widely used in candlestick charts. In error is plotted against Vix data and results are
Japanese, Heikin means “average” and “Ashi” means compared.
“pace”. Since it is based on candlestick patterns. We
give a brief intro to candlesticks. A candlestick Mean Average Error 11.87130668
Mean error when Vix is above 19 14.00708243
describes trade information of a given time period.
Mean Error when Vix is below 16 8.658732212
Every candlestick has 4 values Open, High, Low and
Close. These values represent the value at which
stock or index candle opened, highest value of stock,
lowest value of stock and closing value of the stock in
that time period. Heikin-Ashi transforms these 4
values using following formulas to create new
candlesticks. New Candlestick produces simpler and
smoother candlestick for analysis.

Transformation rules:
Let Ocurrent, Hcurrent , Lcurrent, Ccurrent represents current
open, high, low, close values.
Let OPrev, HPrev, LPrev, CPrev represents Previous
day/period open, high, low, close values. These results clearly show an increase in error rate
Then Heikin-Ashi values (HA) are calculated as:-

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


364
Nishchal Sharma, Dr. Chaman Singh Chauahn Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 363-365

when Vix is High. A vix above 19 causes a spike in [8]. "A hybrid stock trading framework integrating
error rate. Also a lower level of vix causes low error technical analysis with machine learning
rate. However variation in lower level of vix is not techniques - ScienceDirect."
linear with the error rate. This phenomenon can’t be [9]. "A Deep Neural-Network Based Stock Trading
explained and a further study might be needed. System Based on Evolutionary Optimized
Technical Analysis Parameters - ScienceDirect."
IV.CONCLUSION [10]. "Establishing Rules for Self-Organizing
Systems-of-Systems - ScienceDirect."
[11]. "Stock market index prediction using artificial
It is clear that Vix data may be an indication of
neural network - ScienceDirect."
volatility of underlying system and the noise may
increase during high period of volatility which affect
the learning capability of ANN but it is also clear that Cite this article as :
there is a still a disconnect between vix and stock
market movement and a further research might be Nishchal Sharma, Dr. Chaman Singh Chauhan,
needed. "Heikin-Ashi Transformation and Vix Index data for
Stock Market Index Prediction and It's Effects ",
V. REFERENCES International Journal of Scientific Research in
Computer Science, Engineering and Information
[1]. "Stock Market Volatility and Learning - ADAM Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 363-365, January-February 2019.
- 2016 - The Journal of Finance - Wiley Online
Available at doi :
Library."
https://doi.org/10.32628/CSEIT195195
[2]. "Data | Free Full-Text | Current Core Journal URL : http://ijsrcseit.com/CSEIT195195
Competencies Trend of Small to Medium
Enterprises (SMEs) in China—A Concurrent
Comprehensive Evaluation and Active Learning
Study of Newly Listed Chinese Stocks from
2015 through 2017."
[3]. "The contribution of technology selection to
core competencies - ScienceDirect."
[4]. "Using artificial neural network models in stock
market index prediction - ScienceDirect."
[5]. "Application of artificial neural network for the
prediction of stock market returns: The case of
the Japanese stock market - ScienceDirect."
[6]. "Predicting stock and stock price index
movement using Trend Deterministic Data
Preparation and machine learning techniques -
ScienceDirect."
[7]. "Predicting the Direction of Stock Market
Index Movement Using an Optimized Artificial
Neural Network Model."

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


365
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195185

Factors Affecting the Successful Implementation of Cloud and


IoT - A Study on IT and Business Leaders
Jipson George Thoomkuzhy*, Dr. Mohammed Nazeh
Department of Management Studies, Limkokwing University of Creative Technology, Cyberjaya, Selangor,
Malaysia
ABSTRACT

The Internet of Things (IoT) is turning into the following Internet-related transformation. It enables billions of
gadgets to be associated and speak with one another to share data that enhances the nature of our day-by-day
lives. Then again, Cloud Computing gives on-request, helpful and versatile system get to which makes it
conceivable to share computing assets, without a doubt, this, thus, empowers dynamic information integration
from different information sources. There are numerous issues hindering the fruitful implementation of both
Cloud and IoT. The integration of Cloud Computing with the IoT is the best path on which to defeat these
issues. The huge number of assets accessible on the Cloud can be greatly helpful for the IoT, while the Cloud
can acquire attention to enhance its impediments with true protests in a more powerful and dispersed way. This
paper gives to the various factors that affect the integration of the Cloud into the IoT by featuring the
integration advantages and implementation challenges. This paper mainly focus on a study which was
conducted on various IT and business leaders.
Keywords : Cloud, IoT, CIO, SLA, QoS

I. INTRODUCTION conditions of under-used IT framework. With its


particularly autonomic character, cloud-computing
Cloud empowers its dynamic accessibility takes out a colossal bit of the time largely required to
applications and framework, paying little respect to demand and plan IT assets. Cloud computing likewise
locale. Moreover, fast hierarchical advancement yields basic cost adventure sponsors in the land
works out as intended in perspective of the required for server utilization and besides power and
mechanical and psychological capacity to deal with cooling costs. As a result of virtualization and the
the capacity to make or arrange a game plan and fuse cloud's ability to tap assets (either through a private
enrolling power in its assistance and business benefits cloud or tapping direct open cloud assets), server
more quickly than would be conceivable with the estates can manage the chose weight to expand their
present rising organizations. Cloud computing in like physical impression.
way advances IT streamlining with the target those
IT assets is planned for most essential money saving II. LITERATURE REVIEW
favoured outlook. This is conceivable in light of the a) Reliability
way that cloud computing underpins gigantic Enduring quality is one of the key elements to be
adaptability to manage times of intrigue while considered in coursed computing condition. Decided
keeping up an imperative partition from extended quality is depicted as the likelihood that a given thing

CSEIT195185 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 366-396 ]
366
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

will play out its typical farthest point concerning a like way identified with fearless quality: an
given time distribution under a given strategy of organization that is on 24x7 yet goes continually
conditions. A cloud will be more dependable on the isolated is futile. As cloud organizations and
off chance that it is more denounce tolerant and applications wind up being even more amazing and
more adaptable to propelling conditions. It is difficult more dependent on the covered cloud mastermind, it
to have a cloud that is totally free from sorrows or twists up continuously hard rapidly perceiving and
dissatisfaction safe [Garrison, 2012]. Different kinds managing issues. Weights can ascend out of an
of dissatisfactions are interleaved in the passed on individual organization, and furthermore from the
computing condition, for example, flood bafflement, collaboration of various parts and mechanized
timeout dissatisfaction, asset missing dissatisfaction, structures over spread frameworks and server farms,
coordinate dissatisfaction, equipment dissatisfaction, acknowledging issues that set aside a long
programming frustration, and database opportunity to be settled.
disillusionment.
c. Scalability:
The boundless idea of cloud computing is Lee and Kim showed programming centred
uncommonly fundamental yet hard to isolate on the procedures for ensuring the high versatility of
grounds that cloud computing is comprised of the administrations in Cloud Computing. [Lee and Kim,
blend of different factors, for example, wide-zone 2006] High flexibility under high administration
frameworks, heterogeneous programming/adapt parts, loads seeks nothing out of pocket; it very well may be
and so on. There are different confused trades among ensured by grasping some versatility confirmation
the various parts of cloud computing. Thusly, the designs. The most broadly perceived conventional
standard quality models that are depicted for arrangement is to simply incorporate the required
unadulterated programming/outfit or regular resources. Nevertheless, as they propose in the paper,
frameworks cannot be connected with the surveying diverse plans can ensure high adaptability or
of cloud unwavering quality [Pocatine, 2010]. The adaptability.
downtime of the cloud server develop negatively
affects the cloud advantage enduring quality. There is always some cost drew in with running
Surveying and re-designing undeterred quality in versatility ensuring plans, for instance, the expense of
coursed computing is not so much direct errand. additional CPU and memory. The adaptability
Notwithstanding these, two or three contraptions are expanded through such plans should be with respect
utilized to overview and overhaul the cloud to the expense of applying the plans. That is, cost-
advantage consistency. reasonability should be considered in ensuring
flexibility. Administrations should give the level of
b) Availability the esteem decided in their SLAs. Administrations
High transparency is, at last, the colossal vessel of the with palatable flexibility should not to encounter the
cloud. It encapsulates the probability of wherever evil impacts of significant degradation of QoS.
and at whatever direct access toward organizations, Adaptability affirmation designs should ensure that
mechanical congregations, and information and is the administrations satisfy the confinements of meeting
connecting with the effect of dreams of a future with the immaterial edge of their QoS properties
organizations with no physical workplaces, generally [Collinchau, 2013]. From this, two convincing
speaking organizations with totally arranged and programming centreed plans can be resolved:
joined IT frameworks. Transparency or openness is in advantage replication and administration

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


367
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

development. Administration replication is a picked here is the number of dynamic sessions or


methodology for cloning administrations that are logon sessions in each web application.
successfully running on interchange centres to propel
the administration stack over the centres without In perspective of the moving typical of the scaling
impacting errands ahead of time. Rehashed pointer, a dynamic scaling computation is used to
administrations secure additional advantages gave by trigger a scaling event to the provisioning subsystem.
the new centre points for dealing with greater Dependent upon the revived bits of knowledge,
administration stack. Figuratively speaking, movement to scale up or down may be begun. Scaling
advantage replication enhances advantage flexibility up or down suggests that an event will be set off that
and diminishes the threat of QoS corruption by instructs the provisioning subsystem to start or close
dealing with greater advantage loads. To exhibit a down web-server virtual machine events in the cloud.
sufficiency of administration replication, they imitate Right when the web application is scaled up, the as of
the advantage replication plot for the seventeen late started virtual machine event will run the web
extraordinary volumes of administration stack application. After the web application events are
[Collinchau, 2013]. On each advantage stack, they readied, the front-end stack balancer configuration
take a gander at general administration system with record is then invigorated and resuscitated to place
advantage replication plot the extent that typical them into dynamic administrations. As determined
response time. heretofore, the scaling estimation is executed in the
Service Monitor sub-system and is used to control
Administration migration is an arrangement for and trigger the scale-up or down in the Provisioning
setting an administration on a choice centre when a sub-structure on the quantity of virtual machine
particular centre can't give high QoS in view of a cases in light of the estimations of the scaling marker
physical issue or programming issue. After [Frey, Reich, et al. 2013].
administration migration, the moved administration
has out a vague influence from it did on the insecure d) Security
centre, and the volatile centre point is ousted from Like PC security programs, cloud security
the once-over of administration centre points. consolidates a relative general concern: keeping up
Components, advantage stack and machine of the validity of data, guaranteeing access is limited to
adaptability ensuring plan, are same with advantage underwrite clients, and keeping up the availability of
replication re-enactment. In this re-enactment, they data and organizations (EDUCAUSE, 2010).With
have a circumstance that an administration is drawn cloud computing, the data and organizations are
to a closer centre from clients. To enable this, they outer to the grounds and in like way controlling
acknowledge that the response time is in respect to moreover, securing these points of interest changes
the partition. Appropriately, an administration is into an impressively more unpredictable and testing
moved to the speediest centre point to the extent that proposition. Data encryption, e-exposure, rehash and
the response time. They developed a web application dependability of data fortresses and recuperation of
for controlling the provisioning and de-provisioning data, the entire arrangement sensibility of the cloud
of web-server VM events, a dynamic scaling seller and laws concerning farthest point and access
estimation in perspective of the huge point of to data all breeze up fundamental issues. It is central
confinement or scaling marker of the web application that individual and controlled data are sincerely
[Chen and Kunz, 2016]. The scaling marker that is ensured and that they are appropriately overviewed.
Fratto (2009) called attention to that the security of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


368
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

data remains the commitment of the data a cloud setting, this test is befuddled by the circled
proprietor/outsourcer and scattered computing thought of fogs and the possible nonattendance of
professional organizations are not plainly subject for supporter care over where data is secured and who
data misfortune or fines or other lawful. Like PC has or can approach.
security programs, cloud security consolidates a
practically identical general concern. Data Framework Integrity: the cloud requires security
orchestrated in an open cloud and applications against deliberate subversion or mischief of the value
running in an open cloud may encounter specific of a cloud. Inside a cloud, there are accomplices:
security exposures than would be the situation in an endorsers, providers, and a combination of chiefs.
on region energized condition. The capacity to fragment get to rights to each one of
these social occasions, while monitoring harmful
Right when an association purchases into a cloud, attacks, is a key normal for keeping up cloud
each one of the data created and arranged will reliability. In a cloud setting, any nonappearance of
physically abide in premises had and worked by a porousness into a cloud's frameworks makes it more
provider. In this extraordinary situation, the troublesome for endorsers of checking the
fundamental issue is whether a supporter can obtain uprightness of cloud-encouraged applications. Cloud
an attestation that a provider is executing the computing administrations use the Internet as a
equivalent or proportionate controls as for what the transmission medium and change data development
endorser would have completed. The going with resources into administrations for end-customers,
issues develop when an endorser is endeavoring to including programming administrations, figuring
ensure scope for these controls: Compliance stage administrations, change organize
necessities, regarding data that an endorser is a administrations, and crucial structure leasing.
significance to move to a cloud, may call for specific
levels and granularities of audit logging, the time of Vaquero and Lindner prescribed that cloud
cautions, activity declaring, and data support. For computing could be described as the joining of virtual
encryption of data still, the quality of the encryption resources according to customer necessities,
figuring suite, the key administration plots a provider adaptably solidifying resources including hardware,
reinforces, and the number of keys for each datum change stages, and diverse applications to make
proprietor (individual or shared keys) should be administrations [Faisal, 2011]. The novel features of
known by the data proprietors. Data arranged in an cloud computing join the capacity of customer data in
open cloud and applications running in an open the cloud and the nonattendance of any prerequisite
cloud may experience particular security exposures for programming foundation on the client side. For
than would be the circumstance in an on-area whatever timeframe that the customer can interface
encouraged condition. with the Internet, most of the gear resources in the
cloud can be used as a client-side structure. When in
Data Privacy: Security watches out for the protection doubt, cloud-computing applications ask for driven,
of data for specific components, for instance, giving distinctive administrations as demonstrated by
endorsers or others whose data is taken care of in a customer requirements, and expert communities
structure. Assurance passes on authentic and chance charge by metered time, instances of usage, or
concerns and should be viewed as a particular test portrayed period.
and also a legal and good concern. Guaranteeing
insurance in any handling system is a specific test; in

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


369
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

The use of passwords as an affirmation strategy is in a remote server requires a costly expense of
even more outstanding to general customers, yet transmission and capacity. To alleviate the expense of
messages sent by the customer are weak against the transmission and capacity, it is required to partition
covert record by software engineers who might then IoT data into two sorts of data; one is put away in
have the capacity to use the data in the message to edge systems and the other is put away in cloud
sign into the administration as the customer. systems. The impact of Edge computing is uncovered
with the treatment of IoT data in edge systems.
Additionally, the One-Time Password (OTP)
approval system fluctuates from a great many people As of not long ago, most system hardware, for
gatherings' beginning of a secret word. By far, most example, switches, passages, and switches simply
grasp a mystery word to be a watchword picked by forward data conveyed from other system gadgets,
the customer to be critical and can be used repeatedly. not to peruse the substance or change them. In light
The complement of OTP, in any case, is the single- of end-to-end communication, data is recognized and
use nature of the mystery key. In the wake of getting continue at a last relating hub. This is a common use
affirmation from the customer, the structure side of cloud computing a customer service
must make an ensured transmission channel to communication. Be that as it may, in the IoT
exchange data with the customer. The Secure environment, some IoT data will be exchanged to a
Attachments Layer (SSL) is a run of the mill strategy cloud system and some IoT data will be conveyed to
for building secure channels, primarily using RSA an edge hub/haze hub. The primary purpose behind
encryption to transmit the secret keys required for this partition is to give constant preparing and
the opposite sides to encode and unravel data security upgrade. Despite the fact that there are
transmitted between them [Zheng, 2009]. Consistent numerous new advancements to diminish the
methodologies for guaranteeing customer data join postpone time and transmission time, it is difficult to
encryption going before a capacity, customer ensure ongoing handling. The commonplace utilize
approval strategies before capacity or recuperation, instance of this necessity is the Industrial Internet
and building secure channels for data transmission. and savvy processing plant. What's more, despite the
These security systems commonly require fact that, there are control capacities to give security,
cryptography figurings and propelled stamp the more fundamental decision is that not to uncover
techniques, as illuminated underneath. Ordinary data the private data to open systems. On the off chance
encryption systems fuse symmetric and Hilter kilter that we separate IoT data into private data and non-
cryptography computations. private data and keep private data inside an edge
organize, not to uncover them in an open system, it
e) Capacity will diminish numerous feeble purposes of security.
As huge IoT sensors, IoT actuators, and IoT gadgets
are associated with the Internet, IoT data volume In the event that we consider new difficulties of IoT
from these things are required to increment administrations, the enormous volume of IoT data, as
dangerously. Also, it is normal that a significant well as the gigantic number of IoT things, can be a
number of these high volumes of IoT data are created basic issue. Despite the fact that we recognize this
and additionally devoured inside edge systems, not to future issue, Internet design initially has the capacity
navigate through cloud systems. As of recently, of scalability and it will alleviate the scalability issue
primarily IoT data produced IoT things are moved in the IoT environment. In any case, we can't assess
and gathered in a remote server and to store IoT data the quantity of IoT things later on and we can't

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


370
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

ensure the Internet engineering still manage the interviewees avowed that execution issues are caused
scalability issue in the IoT environment. Edge generally by the low web association the adopters
computing will isolate the scalability area into edge have in there. In like way, little organizations should
systems and outside system (e.g., cloud systems) and ensure to have a snappy Web association which
this partition of scalability space can give a more would reduce inaction (Marston et al., 2011). As
effective approach to handle the enormous number of demonstrated by Thurman (2008), execution is one of
IoT things. Since Edge computing can deal with IoT the troubles of cloud advancement execution
data in an edge zone and store the IoT data in an contiguous data security, data affirmation, and
edge/haze hub, and continue IoT data on the off availability. There are challenges in grasping
chance that it is required, it can likewise isolate the conveyed computing advancement. Key execution
administration area into two sections. Edge obstacles are application quality (i.e. application
Computing can focus on the administration of IoT execution likewise, versatility), application faithful
things in an edge zone and collaborate with the quality (i.e. persevering administrations) and data
administration of other outside systems. security and insurance concern (i.e. illegitimate
access of organization data, security bursts) (Benlien
f) Throughput et al., 2009). In this manner, execution is a factor for
Throughput implies the execution of errands by a organizations who wish to get Cloud courses of
figuring administration or contraption over a specific action and use them capably.
period. It checks the proportion of completed clash
with time exhausted and may be used to evaluate the g) Lower costs
execution of a processor, memory and also In cloud computing, setting up an organization as a
framework trades. Throughput was considered to utility, and giving figuring as an organization to
evaluate the benefit of PC processors. This was, clients is a piece of the real work. One of the
generally, registered with respect to bunch beguiling and fundamental cost issues in cloud
occupations or assignments consistently and computing is the capacity to pay for the favorable
countless consistently. A couple of backups measure a position, in the viewpoint of necessities, keeping
structure's general throughput by evaluating the total away from the significant costs for PC frameworks
and unusualness of work, the number of synchronous purchases (Armbrust, 2009). Consequently, cost
customers and application/system responsiveness. sufficiency is another basic factor bearing
Additionally, for arranging trades, throughput is organizations should note while tolerating spread
evaluated by discovering the proportion of data computing. The organizations of today dependably
traded between zones in the midst of a foreordained filter for approaches to manage cut expense or
period, overall occurring as bits consistently (bps), additional cash. Organizations need to spend a critical
which has created to bytes consistently (Bps), piece of their wage on the IT foundation, while
kilobytes consistently (Kbps), megabytes consistently under 10% of their servers can be incredibly used,
(Mbps) and gigabytes consistently (Gbps). accomplishing an essential maltreatment of cash. Also,
these servers should be supplanted respectably at
The physical zone of the data can influence the standard intervals and ought to be kept up and
execution of the system if it is arranged in another bit administrated; expanding its aggregate cost practices
of the world. In this way, Cloud plans may significantly (Marston et al., 2011). Cost perfect
comprehend the postponement in the administration conditions are the most grounded driver influencing
(Kim et al., 2009). Concurring the result consider the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


371
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

IT chiefs' impression of SaaS openings (Benlian et. al, Organizations can in like manner immensely
2011). diminish their help costs (Miller, 2009). This implies
both gear and programming upkeep.
Above all, cloud computing enables organizations to
decrease their gear costs (Miller, 2009). While using According to Miller (2009 p. 26), organizations are
cloud administrations, organizations never again "never again compelled to what a singular work
require intense and high-evaluated PCs to run region PC can do" and now can "perform
applications inside the cloud. This starts with the supercomputing – like assignments utilizing the
lessened requirements for a dealing with power and vitality of countless and servers." Thus they can
storage space (Miller, 2009). Not in any way like perform more conspicuous errands and augmentation
standard programming, for running cloud has by and large their benefit and capability. On the
applications had PCs required less memory. They other hand, from a provider`s perspective, the basic
moreover can be with minimal hard plates in light of cash sparing favorable circumstances for the vendors
the way that there isn't foundation programming. are the foreseen employments and the more broad
Therefore, organizations can diminish costs by customer base (Goncalves and Ballon, 2011).
procuring lower-esteemed PCs. Since using cloud Disseminated computing cuts down legitimate
computing, organizations don't have to do high expenses and empowers an association to offload
interests in the IT system. This especially concerns three sorts of organization (Rosenthal et al., 2009).
the greater organizations (Miller, 2009). Instead of
putting a colossal proportion of trade out an extensive III. SURVEY AND TOOLS USED FOR THE SURVEY
number of extreme servers, the IT divisions of those
organizations can use the cloud` figuring vitality to The survey was conducted using online methods. The
supplant or improve the internal preparing resources. survey attempted to explore the major challenges
Plant administrator (2009) states, as Armbrust at al. faced by CIOs in cloud computing and IoT based
(2009), that organizations don't have to manage the organizations from various industries. This study was
zeniths and the nonpeak times in figuring power asks conducted on IT and business leaders/professionals
for any more drawn out by getting new equipment, from various countries across the globe. A sample size
the apex figuring demands are dealt with by cloud of 405 was used for this survey. Since the target
servers. Cloud computing furthermore prompts cut population is unknown the baseline was of 385 has to
down programming costs. be maintained. The sample size of 405 was used in
order to have an effective result after excluding the
Organizations never again need to buy disconnect errors. A website named e-mailmeform.com was used
programming packs for each PC (Miller, 2009). Or to create an online survey. It is a paid services based
maybe, a particular application is gotten to simply by portal which provides survey setup services. This has
the laborers using that application. Moreover, this in security features like one response from one
like manner infers saved the expense of presenting computer and one response from one IP address.
and keeping up that item on each PC. Another item These sort of security features would help to avoid
related cash sparing preferred standpoint is that duplicate submissions resulting in improved quality
organizations don't have to pay for an item refresh in responses. A total of 40 questions were asked to the
order to have the latest variations of the applications respondents and the results were captured and
(Miller, 2009). As all applications are in the cloud, correlated using the IBM SPSS statistical tool.
they are redesigned thusly by the provider.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


372
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

IV. SURVEY RESULTS & DISCUSSIONS


In all, the majority of the respondents are well above
Altogether, about 346 male and 59 female forty years of age while the least frequency of the
respondents were surveyed to explore the major respondents falls between ages 36 and 40. From the
challenges faced by CIOs in cloud computing and IoT pool of respondents, it’s a single participant that has
based organizations from various industries purely his highest form of education as at then being
from the healthcare, manufacturing and service line. PUC/XII, representing a minute 0.2% of the total
Thus the totality of the participants equals about 405 respondents. The number of graduate-respondent
in number. The feminine gender has the less valid elevated as high as 127 people in total representing
percent of participation as compared to the male 31.4% of the total respondents. A whooping 192
counterparts. participants claimed to be holders of a post-graduate
degree or qualification (47.4 % of the population)
We can therefore infer that there is likelihood of while, about thirty-five of them already has a PhD
more males functioning as IT business leaders than degree in relevant disciplines. Only forty-seven of
their female counterparts. This singular fact was the respondents studied a professional course
considered in the distribution of the survey accruing to about 11.6 percent of the total population.
questionnaires. Out of the 405 respondents to the The remaining 0.7% of the total respondents, that is
online questionnaire, only 55 people (both males and only three people, had a different educational
females) falls within the age range 21 to 25 years of background from the listed.
age; eighty-six participants similarly fell within the
age bracket of 26 to 30; whereas fifty-two
respondents that participated in the online survey
falls between age 31 to 35. The number of
participants between the age ranges of 36 to 40 is
forty-eight while the rest participant, about one-
hundred and sixty-four, are all over 40 years of age.

Fig 3 : Education

According to the population studies, about forty-one


team leaders responded to the online survey
representing a significant percentage of 10.1 percent.
Fig 1 : Gender
Only eighteen of the participants that responded
were junior managers in rank while thirty-one
participants were senior officers or managers within
their respective organizational domain. The number
of respondents that were vice-presidents or general-
managers amounted to thirty-five in all whereas a
whooping one hundred and eighty-two of the
participant were actually chief information officers,
Fig 2 : Age chief executive officers and/or chief functional

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


373
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

officers-which accounts for almost half percent of the


population. The remnant of the 405 respondents
occupied other organizational positions not reflected
in the listed designations.

Fig 5 : Overall Experience

Most of the volunteers who responded, about a


whooping population numbering about 197 people,
only have had between one to five years of
experience as a chief information officer (CIO) in
Fig 4 : Designation their present position which may range from junior
manager to senior managers, etc. Only a handful of
With respect to their overall experience, most of sixty-three persons who participated in the online
the respondents had less than ten years of experience survey have only had less than a year of experience as
and exposure concerning the job description of a related to the job competencies of a chief information
chief information officer; that is, about 169 officer in any position they may be holding. Then it
respondents had lesser that ten years’ experience of was calculated that about seventy-four of the total
the subject matter accounting for 41.7 percent of the 405 respondents had only served between five to ten
total population. years in the capacity of a chief information officer in
any capacity or position. The rest seventy-one
In the same vein, ninety-six of the participant had an respondents have had over ten years of experience in
overall experience and exposure of greater than or their present position with respect to the subject
equal to ten years, but lesser than 20 years at the matter.
same time. Also, ninety-nine of the participants had
already an overall experience worth more than
twenty years but below thirty years of age. The
remainder forty-one respondents already had
experiences of the subject matter over accumulated
years, typically above thirty years. Thus, the overall
percent of respondents with the highest frequency
suggests those of years below ten while those with
Fig 6 : Experience in Present Position
years above thirty had the lowest overall percentage.

The number of respondents within the


telecommunication, service and information
technology (IT) industries that participated in the
online survey was 226 in number representing a
whopping 55.8 percent of the total population.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


374
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

About twenty-five participants interviewed are


currently working in certain manufacturing
industries accounting for 25% of the totality of
respondents. Forty-seven of the respondents are
working in the healthcare sector of the economy
whereas thirty-three of the respondents are
domiciled in the finance, banking or the insurance
companies, with the rest seventy-four of them
practicing in other spheres of discipline accruing a
whole 18.3 of the percentage of the whole population.

Fig 9.

As showed in the data computed above, exactly 142


respondents out of the total number of people who
responded, which is 405 in number, strongly agreed
with the fact stating CIO should be able to answer
Fig 7 : Industry sector of your present organization how an IT solution is beneficial to the company.
Following this wise, a whopping number of 115
From the table provided above, majority of the CIO respondents, representing 28.4 percent of the total
respondents interviewed are currently working in respondents also did agree though not strongly with
either an IT/telecommunication and service firms the fact stated above. Meanwhile, very few
accounting for about 226 respondents in all. In respondents, about 32 in number and accounting for
addition, 25 of the remainder are from the 7.9 percent of the total respondents, disagreed with
manufacturing industries; 47 from healthcare settings that: Do you think that the changed role of CIOs has
while another 33 of the remaining respondents are resulted in successful implementation of Cloud
currently on a job with a finance/insurance computing and IoT based technologies? Only four of
companies. The others could not fit into any of the these respondents who disagreed, strongly expressed
afore-mentioned categories, thus representing 18.3 their own disagreement while about 112 of the
percent of the total number of respondents. remainder participants went neutral and
indeterminate on the matter.

a. Reliability and Availability

Fig 8 : Do you think that the changed role of CIOs


has resulted in successful implementation of Cloud
computing and IoT based technologies.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


375
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

FIG 10: Our cloud and IoT platforms have minimum


likelihood, collapse, or failure

Fig 13

As showed in the data computed above, exactly 66


Fig 11
respondents out of the total number of people who
responded, which is 405 in number, strongly agreed
As showed in the data computed above, exactly 57
with the fact stating CIO should be able to answer
respondents out of the total number of people who
how an IT solution is beneficial to the company.
responded, which is 405 in number, strongly agreed
Following this wise, a whopping number of 215
with the fact stating CIO should be able to answer
respondents, representing 53.1 percent of the total
how an IT solution is beneficial to the company.
respondents also did agree though not strongly with
Following this wise, a whopping number of 181
the fact stated above. Meanwhile, very few
respondents, representing 44.7 percent of the total
respondents, about 21 in number and accounting for
respondents also did agree though not strongly with
5.2 percent of the total respondents, disagreed that as
the fact stated above. Meanwhile, very few
a CIO, in case of cloud/ IoT breakdown, we have
respondents, about 32 in number and accounting for
adequate arrangements to recover. Only four of these
7.9 percent of the total respondents, disagreed that as
respondents who disagreed, strongly expressed their
a CIO, our cloud and IoT platforms have minimum
own disagreement while about 99 of the remainder
likelihood, collapse, or failure. Only seven of these
participants went neutral and indeterminate on the
respondents who disagreed, strongly expressed their
matter.
own disagreement while about 128 of the remainder
participants went neutral and indeterminate on the
matter.

Fig 14: We us quick and effective tools for addressing


Fig 12: In case of cloud/ IoT breakdown, we have issues in our cloud and IoT platforms
adequate arrangements to recover

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


376
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 17
Fig 15

As showed in the data computed above, exactly 74


As showed in the data computed above, exactly 69
respondents out of the total number of people who
respondents out of the total number of people who
responded, which is 405 in number, strongly agreed
responded, which is 405 in number, strongly agreed
with the fact stating CIO should be able to answer
with the fact stating CIO should be able to answer
how an IT solution is beneficial to the company.
how an IT solution is beneficial to the company.
Following this wise, a whopping number of 226
Following this wise, a whopping number of 225
respondents, representing 55.8 percent of the total
respondents, representing 55.6 percent of the total
respondents also did agree though not strongly with
respondents also did agree though not strongly with
the fact stated above. Meanwhile, very few
the fact stated above. Meanwhile, very few
respondents, about 9 in number and accounting for
respondents, about 15 in number and accounting for
2.2 percent of the total respondents, disagreed that as
3.7 percent of the total respondents, disagreed that as
a CIO, users can confidently use our cloud and IoT
a CIO, we possess quick and effective tools for
platforms continuously. Only three of these
addressing issues in our cloud and IoT platforms.
respondents who disagreed, strongly expressed their
Only four of these respondents who disagreed,
own disagreement while about 93 of the remainder
strongly expressed their own disagreement while
participants went neutral and indeterminate on the
about 93 of the remainder participants went neutral
matter.
and indeterminate on the matter.

Fig 18: We have taken enough measures to backup


and store data to be easily retrieved during
Fig 16: Users can confidently use our cloud and IoT emergencies
platforms continuously

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


377
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 21
Fig 19
As showed in the data computed above, exactly 79
As showed in the data computed above, exactly 96 respondents out of the total number of people who
respondents out of the total number of people who responded, which is 405 in number, strongly agreed
responded, which is 405 in number, strongly agreed with the fact stating CIO should be able to answer
with the fact stating CIO should be able to answer how an IT solution is beneficial to the company.
how an IT solution is beneficial to the company. Following this wise, a whopping number of 209
Following this wise, a whopping number of 201 respondents, representing 51.6 percent of the total
respondents, representing 51.9 percent of the total respondents also did agree though not strongly with
respondents also did agree though not strongly with the fact stated above. Meanwhile, very few
the fact stated above. Meanwhile, very few respondents, about 31 in number and accounting for
respondents, about 18 in number and accounting for 7.7 percent of the total respondents, disagreed that as
4.4 percent of the total respondents, disagreed that as a CIO, the cloud/ IoT platform is up and working at
a CIO, we have taken enough measures to backup all the times. Only five of these respondents who
and store data to be easily retrieved during disagreed, strongly expressed their own disagreement
emergencies. Only five of these respondents who while about 81 of the remainder participants went
disagreed, strongly expressed their own disagreement neutral and indeterminate on the matter.
while about 76 of the remainder participants went
neutral and indeterminate on the matter.

Fig 22: Our cloud platform experiences minimum


downtime

Fig 20: The cloud/ IoT platform is up and working at


all the times

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


378
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 23
As showed in the data computed above, exactly 98 Fig 25
respondents out of the total number of people who
responded, which is 405 in number, strongly agreed As showed in the data computed above, exactly 84
with the fact stating CIO should be able to answer respondents out of the total number of people who
how an IT solution is beneficial to the company. responded, which is 405 in number, strongly agreed
Following this wise, a whopping number of 192 with the fact stating CIO should be able to answer
respondents, representing 47.4 percent of the total how an IT solution is beneficial to the company.
respondents also did agree though not strongly with Following this wise, a whopping number of 213
the fact stated above. Meanwhile, very few respondents, representing 51.6 percent of the total
respondents, about19 in number and accounting for respondents also did agree though not strongly with
4.7 percent of the total respondents, disagreed that as the fact stated above. Meanwhile, very few
a CIO, Our cloud platform experiences minimum respondents, about 11 in number and accounting for
downtime. Only four of these respondents who 2.7 percent of the total respondents, disagreed that as
disagreed, strongly expressed their own disagreement a CIO, our cloud and IoT platforms are prepared to
while about 92 of the remainder participants went withstand server failures. Only two of these
neutral and indeterminate on the matter. respondents who disagreed, strongly expressed their
own disagreement while about 95 of the remainder
participants went neutral and indeterminate on the
matter.

Fig 24 : Our cloud and IoT platforms are prepared to


withstand server failures

Fig 26 : Our cloud and IoT platforms are resilient


enough to endure power outages, network failure and
environmental damage.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


379
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 29:

Fig 27: As showed in the data computed above, exactly 98


respondents out of the total number of people who
As showed in the data computed above, exactly 93 responded, which is 405 in number, strongly agreed
respondents out of the total number of people who with the fact stating CIO should be able to answer
responded, which is 405 in number, strongly agreed how an IT solution is beneficial to the company.
with the fact stating CIO should be able to answer Following this wise, a whopping number of 211
how an IT solution is beneficial to the company. respondents, representing 52.1 percent of the total
Following this wise, a whopping number of 210 respondents also did agree though not strongly with
respondents, representing 51.9 percent of the total the fact stated above. Meanwhile, very few
respondents also did agree though not strongly with respondents, about 9 in number and accounting for
the fact stated above. Meanwhile, very few 2.2 percent of the total respondents, disagreed that as
respondents, about 13 in number and accounting for a CIO, we have taken measures for automated
3.2 percent of the total respondents, disagreed that as backups in case the cloud fails. Three of these
a cio, OUR CLOUD AND IOT PLATFORMS ARE RESILIENT respondents who disagreed, strongly expressed their
ENOUGH TO ENDURE POWER OUTAGES, NETWORK own disagreement while about 93 of the remainder
FAILURE, ENVIRONMENTAL DAMAGE. only one of these participants went neutral and indeterminate on the
respondents who disagreed, strongly expressed their matter.
own disagreement while about 88 of the remainder
participants went neutral and indeterminate on the b. Scalability and Security
matter.

Fig 30 a : Our cloud and IoT platform can


Fig 28 : We have taken measures for automated accommodate a large number of users in its full
backups in case the cloud fails capacity

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


380
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 30b
Fig 32

In accordance with the computed data computed in


In accordance with the computed data computed in
Fig.30a and 30b, the number of people who
Fig.31 and 32, the number of people who responded
responded to the notion stating that as CIOs our
to the notion stating that as CIOs our cloud and IoT
cloud and IoT platform can accommodate a large
platform does not collapse or slow down while being
number of users in its full capacity, was 405. About
used by multiple users, was 405. About 98 of these
115 of these respondents representing 28.4 percent of
respondents representing 24.2 percent of the total
the total respondents strongly agreed with this notion,
respondents strongly agreed with this notion, while
while 211 respondents accounting for a whopping
187 respondents accounting for a whopping 46.2
52.1 percent of the total respondents did agree but
percent of the total respondents did agree but not
not strongly with this notion. However a few number
strongly with this notion. However a few number of
of respondents, not more than eleven in number did
respondents, not more than twenty-one in number
disagree with the notion with one more respondents
did disagree with the notion with four more
strongly disagreeing, with about 67 respondents
respondents strongly disagreeing, with about 95
representing 16.5 percent of the total respondents
respondents representing 23.5 percent of the total
were neutral as concerns this notion
respondents were neutral as concerns this notion

Fig 31: Our cloud and IoT platform does not collapse
Fig 33: We keep making improvements to the IoT
or slow down while being used by multiple users
and cloud platform to ensure higher scalability

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


381
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 34 Fig 36

In accordance with the computed data computed in In accordance with the computed data computed in
Fig.33 and 34, the number of people who responded Fig.35 and 36, the number of people who responded
to the notion stating that as CIOs We keep making to the notion stating that as CIOs our systems can
improvements to the IoT and cloud platform to handle multiple application requests arising at the
ensure higher scalability, was 405. About 112 of these same time, was 405. About 104 of these respondents
respondents representing 27.7 percent of the total representing 25.7 percent of the total respondents
respondents strongly agreed with this notion, while strongly agreed with this notion, while 213
203 respondents accounting for a whopping 50.1 respondents accounting for a whopping 52.6 percent
percent of the total respondents did agree but not of the total respondents did agree but not strongly
strongly with this notion. However, a few number of with this notion. However, a few number of
respondents, not more than eleven in number did respondents, not more than eleven in number did
disagree with the notion, with about 79 respondents disagree with the notion, with only two more
representing 19.5 percent of the total respondents respondent strongly disagreeing. About 75
were neutral as concerns this notion respondents representing 17.5 percent of the total
respondents were neutral as concerns this notion

Fig 35: Our systems can handle multiple application


requests arising at the same time
Fig 37: Our cloud and IoT platforms have been
designed to match up with the changing business
objectives

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


382
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 40
Fig 38 In accordance with the computed data computed in
Fig.39 and 40, the number of people who responded
In accordance with the computed data computed in to the notion stating that as CIOs We have adequate
Fig.37 and 38, the number of people who responded safeguards for different interconnected devices to
to the notion stating that as CIOs Our cloud and IoT avoid data theft or breach of privacy, was 405. About
platforms have been designed to match up with the 139 of these respondents representing 34.3 percent of
changing business objectives, was 405. About 102 of the total respondents strongly agreed with this notion,
these respondents representing 25.2 percent of the while 182 respondents accounting for a whopping
total respondents strongly agreed with this notion, 44.9 percent of the total respondents did agree but
while 202 respondents accounting for a whopping not strongly with this notion. However a few number
49.9 percent of the total respondents did agree but of respondents, not more than 9 in number did
not strongly with this notion. However, a few disagree with the notion, with only three more
number of respondents, not more than thirteen in respondent strongly disagreeing. About 72
number did disagree with the notion, with only four respondents representing 17.8 percent of the total
more respondent strongly disagreeing. About 84 respondents were neutral as concerns this notion
respondents representing 20.7 percent of the total
respondents were neutral as concerns this notion

Fig 41: We have adequate safeguards for different


Fig 39: We give utmost importance to network
interconnected devices to avoid data theft or breach
security and data security in our cloud and IoT
of privacy
platforms

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


383
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 42a Fig 43

In accord to the data computed in Fig.41 and 42a, the In accord to the data computed in Fig.42b and 43, the
number of people who responded to the notion number of people who responded to the notion
stating that as CIOs We have adequate safeguards for stating that as CIOs We constantly monitor our cloud
different interconnected devices to avoid data theft and IoT platforms to identify potential security
or breach of privacy, was 405. About 116 of these threats, was 405. About 121 of these respondents
respondents representing 28.6 percent of the total representing 29.9 percent of the total respondents
respondents strongly agreed with this notion, while strongly agreed with this notion, while 185
176 respondents accounting for a whopping 43.5 respondents accounting for a whopping 45.7 percent
percent of the total respondents did agree but not of the total respondents did agree but not strongly
strongly with this notion. However, a few number of with this notion. However, a few number of
respondents, not more than 10 in number did respondents, not more than 16 in number did
disagree with the notion, with only three more disagree with the notion, with only three more
respondent strongly disagreeing. About 100 respondent strongly disagreeing. About 80
respondents representing 24.7 percent of the total respondents representing 22.0 percent of the total
respondents were neutral as concerns this notion respondents were neutral as concerns this notion.

Fig 44 : We regularly update and improve our data


safety mechanisms to counter novel means of breach
Fig 42b : We constantly monitor our cloud and IoT to network security.
platforms to identify potential security threats

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


384
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 47
Fig 45
In accord to the data computed in Fig.46 and 47, the
In accord to the data computed in Fig.44 and 45, the number of people who responded to the notion
number of people who responded to the notion stating that as CIOs Our cloud and IoT platforms are
stating that as CIOs We regularly update and improve designed to make optimal use of existing resources,
our data safety mechanisms to counter novel means was 405. About 116 of these respondents representing
of breach to network security, was 405. About 114 of 28.6 percent of the total respondents strongly agreed
these respondents representing 28.1 percent of the with this notion, while 202 respondents accounting
total respondents strongly agreed with this notion, for a whopping 49.9 percent of the total respondents
while 190 respondents accounting for a whopping did agree but not strongly with this notion. However,
46.9 percent of the total respondents did agree but a few number of respondents, not more than 7 in
not strongly with this notion. However, a few number did disagree with the notion, with only one
number of respondents, not more than ten in number more respondent strongly disagreeing. About 79
did disagree with the notion, with only two more respondents representing 19.5 percent of the total
respondent strongly disagreeing. About 89 respondents were neutral as concerns this notion.
respondents representing 22.0 percent of the total
respondents were neutral as concerns this notion.

c. Capacity and Throughput

Fig 48: We constantly track the resources used up by


our cloud and IoT products and take steps to
minimize resource loss
Fig 46: Our cloud and IoT platforms are designed to
make optimal use of existing resources

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


385
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 4 Fig 51

In accord to the data computed in Fig.48 and 49, the In accord to the data computed in Fig.50 and 51, the
number of people who responded to the notion number of people who responded to the notion
stating that as CIOs We constantly track the stating that as CIOs We undertake capacity planning
resources used up by our cloud and IoT products and much before the implementation of cloud and IoT
take steps to minimize resource loss, was 405. About based technologies, was 405. About 105 of these
104 of these respondents representing 25.7 percent of respondents representing 25.9 percent of the total
the total respondents strongly agreed with this notion, respondents strongly agreed with this notion, while
while 203 respondents accounting for a whopping 203 respondents accounting for a whopping 50.1
50.1 percent of the total respondents did agree but percent of the total respondents did agree but not
not strongly with this notion. However, a few strongly with this notion. However, a few number of
number of respondents, not more than 12 in number respondents, not more than 10 in number did
did disagree with the notion, with only four more disagree with the notion, with only a single more
respondent strongly disagreeing. About 82 respondent strongly disagreeing. About 86
respondents representing 20.2 percent of the total respondents representing 21.2 percent of the total
respondents were neutral as concerns this notion. respondents were neutral as concerns this notion.

Fig 50: We undertake capacity planning much before Fig 52: We are constantly innovating means to reduce
the implementation of cloud and IoT based resource use and improve productivity
technologies

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


386
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 55

Fig 53 In accord to the data computed in Fig.54 and 55, the


number of people who responded to the notion
In accord to the data computed in Fig.52 and 53, the stating that as CIOs We try to eliminate errors, in
number of people who responded to the notion order to avoid resource loss, was 405. About 113 of
stating that as CIOs We try to eliminate errors, in these respondents representing 27.9 percent of the
order to avoid resource loss, was 405. About 118 of total respondents strongly agreed with this notion,
these respondents representing 29.1 percent of the while 187 respondents accounting for a whopping
total respondents strongly agreed with this notion, 46.2 percent of the total respondents did agree but
while 189 respondents accounting for a whopping not strongly with this notion. However, a few
46.7 percent of the total respondents did agree but number of respondents, not more than 14 in number
not strongly with this notion. However, a few did disagree with the notion, with only three more
number of respondents, not more than 14 in number respondent strongly disagreeing. About 88
did disagree with the notion, with only a single more respondents representing 21.7 percent of the total
respondent strongly disagreeing. About 83 respondents were neutral as concerns this notion.
respondents representing 20.5 percent of the total
respondents were neutral as concerns this notion.

Fig 56: We allocate only required resources to cloud


or IoT platforms to avoid losses
Fig 54: We try to eliminate errors, in order to avoid
resource loss

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


387
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 57 Fig 59

In congruence to the data computed in Fig.56 and 57, In congruence to the data computed in Fig.58 and 59,
the number of people who responded to the notion the number of people who responded to the notion
stating that as CIOs We allocate only required stating that as CIOs our cloud and IoT based
resources to cloud or IoT platforms to avoid losses, applications can process maximum tasks in a given
was 405. About 99 of these respondents representing unit of time, was 405. About 105 of these respondents
24.4 percent of the total respondents strongly agreed representing 25.9 percent of the total respondents
with this notion, while 205 respondents accounting strongly agreed with this notion, while 192
for a whopping 50.6 percent of the total respondents respondents accounting for a whopping 47.4 percent
did agree but not strongly with this notion. However, of the total respondents did agree but not strongly
a few number of respondents, not more than 13 in with this notion. However, a few number of
number did disagree with the notion, with only one respondents, not more than 13 in number did
more respondent strongly disagreeing. About 87 disagree with the notion, with only two more
respondents representing 21.5 percent of the total respondent strongly disagreeing. About 93
respondents were neutral as concerns this notion. respondents representing 23.0 percent of the total
respondents were neutral as concerns this notion.

Fig 58: Our cloud and IoT based applications can


process maximum tasks in a given unit of time Fig 60: We make sure we respond to the requests of
every user within minimum possible time

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


388
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 61 Fig 63

In congruence to the data computed in Fig.60 and 61, In congruence to the data computed in Fig.62 and 63,
the number of people who responded to the notion the number of people who responded to the notion
stating that as CIOs We make sure we respond to the stating that as CIOs We do our best to achieve
requests of every user within minimum possible time, maximum transactions per second, was 405. About
was 405. About 107 of these respondents representing 112 of these respondents representing 27.7 percent of
26.4 percent of the total respondents strongly agreed the total respondents strongly agreed with this notion,
with this notion, while 199 respondents accounting while 193 respondents accounting for a whopping
for a whopping 49.1 percent of the total respondents 47.7 percent of the total respondents did agree but
did agree but not strongly with this notion. However, not strongly with this notion. However, a few
a few number of respondents, not more than 11 in number of respondents, not more than 11 in number
number did disagree with the notion, with only five did disagree with the notion, with only two more
more respondent strongly disagreeing. About 83 respondent strongly disagreeing. About 87
respondents representing 20.5 percent of the total respondents representing 21.5 percent of the total
respondents were neutral as concerns this notion. respondents were neutral as concerns this notion.

Fig 62: We do our best to achieve maximum


transactions per second
Fig 64: Our throughput has been consistently
improving over the years

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


389
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 64b
Fig 66
In congruence to the data computed in Fig.64 and
64b, the number of people who responded to the In congruence to the data computed in Fig.65 and 66,
notion stating that as CIOs our throughput has been the number of people who responded to the notion
consistently improving over the years, was 405. stating that as CIOs We are constantly on the lookout
About 107 of these respondents representing 26.4 for new mechanisms to improve our throughput, was
percent of the total respondents strongly agreed with 405. About 109 of these respondents representing
this notion, while 197 respondents accounting for a 26.9 percent of the total respondents strongly agreed
whopping 48.6 percent of the total respondents did with this notion, while 198 respondents accounting
agree but not strongly with this notion. However, a for a whopping 48.9 percent of the total respondents
few number of respondents, not more than 13 in did agree but not strongly with this notion. However,
number did disagree with the notion, with only three a few number of respondents, not more than 12 in
more respondent strongly disagreeing. About 85 number did disagree with the notion, with only a
respondents representing 21.0 percent of the total single more respondent strongly disagreeing. About
respondents were neutral as concerns this notion. 85 respondents representing 21.0 percent of the total
respondents were neutral as concerns this notion.

d. Lower Costs

Fig 65: We are constantly on the lookout for new


mechanisms to improve our throughput

Fig 67: Our cloud and IoT platforms offer integrated


data storage solutions, hereby helping reduce data
storage costs

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


390
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 68 Fig 70

In congruence to the data computed in Fig.67 and 68, In congruence to the data computed in Fig.69 and 70,
the number of people who responded to the notion the number of people who responded to the notion
stating that as CIOs our cloud and IoT platforms offer stating that As cloud based programs can be accessed
integrated data storage solutions, hereby helping anywhere, it saves logistics and data transfer costs,
reduce data storage costs, was 405. About 99 of these was 405. About 109 of these respondents representing
respondents representing 24.4 percent of the total 26.9 percent of the total respondents strongly agreed
respondents strongly agreed with this notion, while with this notion, while 195 respondents accounting
204 respondents accounting for a whopping 50.4 for a whopping 48.1 percent of the total respondents
percent of the total respondents did agree but not did agree but not strongly with this notion. However,
strongly with this notion. However, a few number of a few number of respondents, not more than 12 in
respondents, not more than 14 in number did number did disagree with the notion, with about 5
disagree with the notion, with about 3 more more respondent strongly disagreeing. About 84
respondent strongly disagreeing. About 85 respondents representing 20.7 percent of the total
respondents representing 21.0 percent of the total respondents were neutral as concerns this notion.
respondents were neutral as concerns this notion.

Fig 71: As our cloud and IoT technologies can be


Fig 69: As cloud based programs can be accessed expanded for large number of users, the cost of
anywhere, it saves logistics and data transfer costs additional infrastructure is saved

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


391
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 72 Fig 74

In congruence to the data computed in Fig.71 and 72, In congruence to the data computed in Fig.73 and
the number of people who responded to the notion 764, the number of people who responded to the
stating As our cloud and IoT technologies can be notion stating that Cloud and IoT based technologies
expanded for large number of users, the cost of in reduction of server maintenance, power and other
additional infrastructure is saved, was 405. About 115 costs, was 405. About 106 of these respondents
of these respondents representing 28.4 percent of the representing 26.2 percent of the total respondents
total respondents strongly agreed with this notion, strongly agreed with this notion, while 189
while 191 respondents accounting for a whopping respondents accounting for a whopping 46.7 percent
47.2 percent of the total respondents did agree but of the total respondents did agree but not strongly
not strongly with this notion. However, a few with this notion. However, a few number of
number of respondents, not more than 19 in number respondents, not more than 16 in number did
did disagree with the notion, with about 4 more disagree with the notion, with about 6 more
respondent strongly disagreeing. About 76 respondent strongly disagreeing. About 88
respondents representing 18.8 percent of the total respondents representing 21.7 percent of the total
respondents were neutral as concerns this notion. respondents were neutral as concerns this notion.

Fig 73: Cloud and IoT based technologies in reduction Fig 75: Cloud based data backup saves the cost of
of server maintenance, power and other costs retrieving lost data during system crashes or natural
calamities

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


392
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

confront. Two similarity issues were perceived –


similarity with existing framework, what's more, a
similarity with other Cloud framework. The
examination set up excess as a factor. The discoveries
proposed that any expenses of the current foundation
must be considered before an official choice to
receive Cloud processing can be made in light of the
fact that some of the time utilizing the current
foundation could be the more productive choice.

The results shows that the new role of the CIO has
resulted in successful implementation of cloud and
IoT based technologies. The study also states that the
Fig 76
most of the companies cloud and IoT platforms has
less chance of failure and they are set up in such a
In congruence to the data computed in Fig.75 and 76,
way that they have adequate methods for recovery.
the number of people who responded to the notion
To have the best availability and reliability most of
stating that Cloud based data backup saves the cost of
the cloud infrastructures are using effective tools to
retrieving lost data during system crashes or natural
address the issues which can give higher confidence
calamities, was 405. About 117 of these respondents
rate to the cloud adopters. The confirmation from IT
representing 28.9 percent of the total respondents
and business leaders states that they have taken the
strongly agreed with this notion, while 188
necessary measure to backup and store data to be
respondents accounting for a whopping 46.4 percent
easily retrieved during emergencies. The results
of the total respondents did agree but not strongly
shows that the availability of the cloud is high with
with this notion. However, a few number of
minimal downtime and server failures. The
respondents, not more than 14 in number did
infrastructure is designed with enough redundancy so
disagree with the notion and about 83 respondents
that resiliency is achieved in case of failures. The
representing 20.5 percent of the total respondents
automated backup mechanisms are helping the
were neutral as concerns this notion.
organizations from any data loss.

V. CONCLUSION AND RECOMMENDATIONS


The analysis on the scalability and security aspects
states that the cloud and IoT platforms are capable to
Cloud services could give a business relative favorable scale and it is built with the required capacity so that
position over its rivals and cost advantage, through the performance doesn't get affected during peak
the advantages they give i.e. enhanced business operation period. The organizations are spending
agility and diminished capital use. The multifaceted considerable effort to bring in improvement to the
nature of the Cloud was observed to be a noteworthy cloud and IoT platforms to improve the scalability.
challenge in the adoption procedure, however, it This can also give an idea that there is huge amount
ought to be noticed that Cloud arrangements are of research or innovations are going on in this area.
getting less mind boggling with the improvement of The organizations are trying to explore options to
the innovation. Similarity was distinguished as improve the scalability by innovative methods. The
another test that a business is probably going to platforms are designed with proper business goals so

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


393
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

that any change in the business environment can be Moreover, the result analysis unveiled that the
effectively addressed by the cloud infrastructures. overall experience and/or exposure, and the industry
This also gives a learning to the companies sector of one’s present organization plus his/her
implementing and adopting cloud to make sure that background education status goes a long way in the
they are doing the required business forecasting and successful implementation of Cloud-IoT based
planning for designing the cloud and IoT technologies. Also, it became obviously necessary
infrastructures. A poor planning in this area can that cloud-based data back-up saves the cost of
negatively affect the future business of any retrieving lost data during system crashes or any
organization. So the organizations has to give utmost natural calamity, as well as helps in the reduction of
importance to match up with the environmental server maintenance, power and other costs. It is
changes in the business. The organizations are important that as cloud & IoT technologies are being
concerned about the security of their infrastructures expanded to be accessible by large number of users,
and they are doing the best possible measures to the cost of all additional infrastructures is being
safeguard their infrastructure. The organizations are saved. The results also reveals that as cloud-based
constantly monitoring their cloud and IoT platforms programs are being accessed anywhere, it saves
to identify any potential security threats that would logistics and data transfer costs much more
impact their business. From this study it gives a importantly. It proved that cloud & IoT platforms
reference that there is huge scope for research in offer integrated data storage solutions, thereby
scalability and security aspects of the cloud and IoT. helping to reduce data storage costs in the
Assessing the scalability and security of cloud-IoT atmosphere of its successful implementation.
technologies, including its capacity, throughput and
additional lower costs, it is concluded that the In the same vein, as business leaders willing to
targeted successful implementation of cloud-IoT successfully implement and integrate cloud
based technologies will be very much easier to technologies, it is necessary to achieve maximum
accomplish by improving cloud throughputs transaction per seconds and respond to the requests
frequently by looking out for newer operating of each user within the shortest possible timeframe,
mechanisms. thus enhancing good customer satisfaction and
service. It is also quite important to allocate only the
The analysis of the capacity and throughput required resources to cloud-IoT based platforms to
parameters states that the organizations are showing avoid critical losses. As business and IT leaders, it
more interest in optimizing the existing resources. becomes essential to constantly innovate, thus
The organizations implementing the cloud and IoT improving productivity while minimising the use of
has to prepare a roadmap of the capacity well in scarce resources and eliminating unnecessary errors.
advance before the implementation of cloud and IoT. It is also essential to indulge in capacity planning and
A proper planning would help the organizations to tracking of the available resources. Furthermore, it
optimise the resources when the volume of will do the company good to update data-safety
transactions increases in the infrastructure. An mechanisms frequently, and checking any breach in
architects and designers also need to consider the network security or security threats, as a way of
chances of error rates and do the required resource enhancing optimal usage of resources, i.e, network
planning to avoid any losses. The results of this study security and data security. In addition, cloud
states that there is huge scope for research in the area technologies can multi-task in terms of application
of digital optimization. requests without slowing down, and it is built to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


394
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

accommodate changing business objectives [3]. Pocatilu, P., Alecu, F., &Vetrici, M. (2010).
(flexibility). Measuring the efficiency of cloud computing
for e-learning systems. Wseas transactions on
The cloud-IoT platforms should be able to withstand
computers, 9(1), 42-51.
server failures, endure power outages, network
[4]. Deka Ganesh Chandra & Dutta Borah Malaya,
failures and any form of environmental damage one
(2012), Role of Cloud Computing in Education,
again. These and many more will surely enhance the
ICCEET, Pg: 832-836.
reliability, availability, scalability, security, capacity,
[5]. Dong, B., Zheng, Q., Qiao, M., Shu, J., & Yang,
throughput and cut operating costs of cloud-Iot based
J. (2009). BlueSky cloud framework: an e-
technologies in the quest of its successful
learning framework embracing cloud
implementation and integration. The execution was
computing. In Cloud Computing (pp. 577-582).
another issue that was recognized as a factor. The
Springer Berlin Heidelberg.
discoveries recommended it needs thought in light of
[6]. Faisal A. Alshuwaier, Abdullah A. Alshwaier,
the fact that receiving Cloud services could influence
Ali M. Areshey (2011), Applications of Cloud
the system execution and accessibility of the services
Computing in Education.
is outside of the business control. Likewise, security
[7]. Feng, D. G., Zhang, M., Zhang, Y., & Xu, Z.
was distinguished as the best mechanical test. The
(2011). Study on cloud computing security.
discoveries contradicted that by recommending that
Journal of Software, 22(1), 71-83.
substantial organizations have their very own
[8]. Fox, A., Griffith, R., Joseph, A., Katz, R.,
foundation that needs thought and it would be less
Konwinski, A., Lee, G., ... &Stoica, I. (2009).
demanding for little organizations to receive cloud
Above the clouds: A Berkeley view of cloud
services as a result of the moderately little set-up
computing. Dept. Electrical Eng. and Comput.
expense. Considering the constraints to the
Sciences, University of California, Berkeley,
exploration of this examination a proposal for further
Rep. UCB/EECS, 28, 13.
research can be made. With the end goal to address
[9]. Hailu, A. (2012). Factors influencing cloud-
these issues, a further research could be made
computing technology adoption in developing
utilizing a quantitative technique that could look at a
countries (Doctoral dissertation, Capella
bigger example of the populace.
University).
[10]. Chen, Y., & Kunz, T. (2016, April).
VI. REFERENCES
Performance evaluation of IoT protocols under
a constrained wireless access network. In
[1]. Iosup, A., Ostermann, S., Yigitbasi, M. N.,
Selected Topics in Mobile & Wireless
Prodan, R., Fahringer, T., &Epema, D. (2011).
Networking (MoWNeT), 2016 International
Performance analysis of cloud computing
Conference on (pp. 1-7). IEEE.
services for many-tasks scientific computing.
[11]. Collinchau. (2013, August 19). 7 performance
IEEE Transactions on Parallel and Distributed
metrics for measuring cloud ROI within HP
systems, 22(6), 931-945.
Cloud Management. Micro Focus. Retrieved 19
[2]. The cloud service industry’s 10 most critical
October 2018 from
metrics. Guiding Metrics. Retrieved 19
https://community.softwaregrp.com/t5/IT-
February 2018 from
Operations-Management-ITOM/7-
https://guidingmetrics.com/content/cloud-
performance-metrics-for-measuring-cloud-
services-industrys-10-most-critical-metrics

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


395
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

ROI-within-HP-Cloud/ba-
p/300699#.Wopi1INua01
[12]. Frey, S., Reich, C., &Lüthje, C. (2013,
September). Key performance indicators for
cloud computing SLAs. In The Fifth
International Conference on Emerging
Network Intelligence, EMERGING (pp. 60-64).

Cite this article as :

Jipson George Thoomkuzhy, Dr. Mohammed Nazeh,


"Factors Affecting the Successful Implementation of
Cloud and IoT - A Study on IT and Business Leaders ",
International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 366-396, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195185
Journal URL : http://ijsrcseit.com/CSEIT195185

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


396
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195177
Development of Remote Operated Agri Vehicle
Juno Isabel Susinthra M *1, K. Gayadri 2
*1Assistant Professor, Department of Computer Science, Bon Secours College for Women, Bharathidasan
University, Thanjavur, Tamil Nadu, India
2 M.Sc, Department of Computer Science, Bon Secours College for Women, Bharathidasan University,
Thanjavur, Tamil Nadu, India

ABSTRACT
An agricultural mechanism or agri mechanism could be a mechanism deployed for agricultural functions. The
most space of application of robots in agriculture is seeding robots square measure designed to exchange human
labour. The agricultural business is behind alternative complementary industries in victimization robots as a
result of the kind of jobs concerned in agriculture are not any and dig. This project strives to develop a
mechanism capable of activity operations like automatic seeding and dig. It additionally provides manual
management once needed and keeps tabs on the wetness with the assistance of wetness sensors. Then main part
here is that the microcontroller that supervises the whole method.
Keywords : Agri Vehicle, GPS based Remote Controlled Robot, Zigbee, Raspberry Pi, Global Positioning
System, UART, LM35
I. INTRODUCTION environmental factors which in turns helps to
monitor the system. Monitoring environmental
Agriculture is considered as the basis of life for the factors is not enough and complete solution to
human species as it is the main source of food grains improve the yield of the crops. There are number of
and other raw materials. It plays vital role in the other factors that affect the productivity to great
growth of country’s economy. It also provides large extent. These factors include attack of insects and
ample employment opportunities to the people. pests which can be controlled by spraying the crop
Growth in agricultural sector is necessary for the with proper insecticide and pesticides. Secondly,
development of economic condition of the country. attack of wild animals and birds when the crop grows
Unfortunately, many farmers still use the traditional up. There is also possibility of thefts when crop is at
methods of farming which results in low yielding of the stage of harvesting. Even after harvesting, farmers
crops and fruits. But wherever automation had been also face problems in storage of harvested crop. So, in
implemented and human beings had been replaced order to provide solutions to all such problems, it is
by automatic machineries, the yield has been necessary to develop integrated system which will
improved. Hence there is need to implement modern take care of all factors affecting the productivity in
science and technology in the agriculture sector for every stages like; cultivation, harvesting and post
increasing the yield. Most of the papers signifies the harvesting storage. This paper therefore proposes a
use of wireless sensor network which collects the system which is useful in monitoring the field data as
data from different types of sensors and then send it well as controlling the field operations which
to main server using wireless protocol. The collected provides the flexibility. The paper aims at making
data provides the information about different agriculture smart using automation and IoT

CSEIT195177 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 397-403 ]
397
Juno Isabel Susinthra M, K. Gayadri Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 397-403

technologies. The highlighting features of this paper sensing and control irrigation system using
includes smart GPS based remote controlled robot to distributed wireless sensor network aiming for
perform tasks like; weeding, spraying, moisture variable rate irrigation, real time in field sensing,
sensing, bird and animal scaring, keeping vigilance, controlling of a site-specific precision linear move
etc. Secondly, it includes smart irrigation with smart irrigation system to maximize the productivity with
control based on real time field data. Thirdly, smart minimal use of water was developed by Y. Kim. The
warehouse management which includes temperature system described details about the design and
maintenance, humidity maintenance and theft instrumentation of variable rate irrigation, wireless
detection in the warehouse. Controlling of all these sensor network and real time in field sensing and
operations will be through any remote smart device control by using appropriate software. The whole
or computer connected to Internet and the operations system was developed using five in field sensor
will be performed by interfacing sensors, Wi-Fi or stations which collects the data and send it to the
ZigBee modules, camera and actuators with micro- base station using global positioning system (GPS)
controller and raspberry pi. where necessary action was taken for controlling
irrigation according to the database available with the
II. METHODS AND MATERIAL system. The system provides a promising low cost
wireless solution as well as remote controlling for
The newer scenario of decreasing water tables, drying precision irrigation. In the studies related to wireless
up of rivers and tanks, unpredictable environment sensor network, researchers measured soil related
present an urgent need of proper utilization of water. parameters such as temperature and humidity.
To cope up with this use of temperature and moisture Sensors were placed below the soil which
sensor at suitable locations for monitoring of crops is communicates with relay nodes by the use of
implemented in. An algorithm developed with effective communication protocol providing very low
threshold values of temperature and soil moisture can duty cycle and hence increasing the life time of soil
be programmed into a microcontroller-based gateway monitoring system. The system was developed using
to control water quantity. The system can be microcontroller, universal asynchronous receiver
powered by photovoltaic panels and can have a transmitter (UART) interface and sensors while the
duplex communication link based on a cellular- transmission was done by hourly sampling and
Internet interface that allows data inspection and buffering the data, transmit it and then checking the
irrigation scheduling to be programmed through a status messages. The drawbacks of the system were its
web page. The technological development in cost and deployment of sensor under the soil which
Wireless Sensor Networks made it possible to use in causes attenuation of radio frequency (RF) signals.
monitoring and control of greenhouse parameter in
precision agriculture. After the research in the III. RESULTS AND DISCUSSION
agricultural field, researchers found that the yield of
agriculture is decreasing day by day. However, use of
technology in the field of agriculture plays important
role in increasing the production as well as in
reducing the extra man power efforts. Some of the
research attempts are done for betterment of farmers
which provides the systems that use technologies Fig 1. System Overview
helpful for increasing the agricultural yield. A remote

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


398
Juno Isabel Susinthra M, K. Gayadri Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 397-403

In the present system, every node is integration with with AVR microcontroller. Motion detector will
different sensors and devices and they are detect the motion in the room when security mode
interconnected to one central server via wireless will be ON and on detection of motion, it will send
communication modules. The server sends and the alert signal to user via Raspberry pi and thus
receives information from user end using internet providing theft detection.
connectivity. There are two modes of operation of
the system; auto mode and manual mode. In auto
mode system takes its own decisions and controls the
installed devices whereas in manual mode user can
control the operations of system using android app or
PC commands.

IV. ARCHITECTURE OF THE SYSTEM

Node 1:
Node1 is GPS based mobile robot which can be
controlled remotely using computer as well as it can
be programmed so as to navigate autonomously Fig 3.
within the boundary of field using the Co-ordinates Node 2
given by GPS module. Temperature sensor and Humidity sensor senses the
temperature and humidity respectively and if the
value crosses the threshold then room heater or
cooling fan will be switched ON/OFF automatically
providing temperature and humidity maintenance.
Node2 will also controls water pump depending upon
the soil moisture data sent by node3.
Node 3:
Node3 is a smart irrigation node with features like;
Smart control of water pump based on real time field
data i.e. automatically turning on/off the pump after
Fig 2. attaining the required soil moisture level in auto
Node 1 mode, switching water pump on/off remotely via
The Remote controlled robot have various sensors mobile or computer in manual mode, and continuous
and devices like camera, obstacle sensor, siren, cutter, monitoring of soil moisture.
sprayer and using them it will perform tasks like; Hardware used:
Keeping vigilance, Bird and animal scaring, Weeding, a) AVR Microcontroller At mega 16/32:
and Spraying The microcontroller used is, Low-power AVR® 8-bit
Node 2: Microcontroller, having 8K Bytes of In-System Self-
Node2 will be the warehouse. It consists of motion programmable Flash program memory,
detector, light sensor, humidity sensor, temperature Programmable Serial USART, 8-channel, 10-bit ADC,
sensor, room heater, cooling fan altogether interfaced 23 Programmable I/O Lines.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


399
Juno Isabel Susinthra M, K. Gayadri Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 397-403

b) ZigBee Module: The ultra-sonic sensor operates on the principle of


ZigBee is used for achieving wireless communication sound waves and their reflection property. It has two
between Node1 and Node2. The range for Zigbee is parts; ultra-sonic transmitter and ultra-sonic receiver.
roughly 50 meters and it can be increased using high Transmitter transmits the 40 KHz sound wave and
power modules or by using network of modules. It receiver receives the reflected 40 KHz wave and on
operates on 2.4 GHz frequency. Its power its reception, it sends the electrical signal to the
consumption is very low and it is less expensive as microcontroller. The speed of sound in air is already
compared to other wireless modules like Wi-Fi or known. Hence from time required to receive back the
Bluetooth. It is usually used to establish wireless local transmitted sound wave, the distance of obstacle is
area networks. calculated. Here, it is used for obstacle detection in
c) Temperature Sensor LM35: case of mobile robot and as a motion detector in ware
The LM35 is precision IC temperature sensor. Output house for preventing thefts. The ultra-sonic sensor
voltage of LM35 is directly proportional to the enables the robot to detect and avoid obstacles and
Centigrade/Celsius of temperature. The LM35 does also to measure the distance from the obstacle. The
not need external calibration or trimming to provide range of operation of ultra-sonic sensor is 10 cm to 30
accurate temperature range. It is very low cost sensor. cm.
It has low output impedance and linear output. The f) Raspberry Pi:
operating temperature range for LM35 is −55˚ to The Raspberry Pi is small pocket size computer used
+150˚C. With rise in temperature, the output voltage to do small computing and networking operations. It
of the sensor increases linearly and the value of is the main element in the field of internet of things.
voltage is given to the microcontroller which is It provides access to the internet and hence the
multiplied by the conversion factor in order to give connection of automation system with remote
the value of actual temperature. location controlling device becomes possible.
d) Moisture sensor: Raspberry Pi is available in various versions. Here,
Soil moisture sensor measures the water content in model Pi 2 model B is used and it has quad-core ARM
soil. It uses the property of the electrical resistance of Cortex-A53 CPU of 900 MHz, and RAM of 1GB. it
the soil. The relationship among the measured also has: 40 GPIO pins, Full HDMI port, 4 USB ports,
property and soil moisture is calibrated and it may Ethernet port, 3.5mm audio jack, video Camera
vary depending on environmental factors such as interface (CSI), the Display interface (DSI), and
temperature, soil type, or electric conductivity. Here, Micro SD card slot.
it is used to sense the moisture in field and transfer it Software used:
to microcontroller in order to take controlling action a) AVR Studio Version 4:
of switching water pump ON/OFF. Humidity sensor: It is used to write, build, compile and debug the
The DHT11 is a basic, low-cost digital temperature embedded c program codes which are needed to b
and humidity sensor. It gives out digital value and
hence there is no need to use conversion algorithm at
ADC of the microcontroller and hence we can give
its output directly to data pin instead of ADC. It has a
capacitive sensor for measuring humidity. The only
real shortcoming of this sensor is that one can only
get new data from it only after every 2 seconds.
e) Obstacle sensor (Ultra-Sonic):

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


400
Juno Isabel Susinthra M, K. Gayadri Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 397-403

d) SinaProg:
SinaProg is a Hex downloader application with AVR
Dude and Fuse Bit Calculator. This is used to
download code/program and to set fuse bits of all
AVR based microcontrollers.
e) Raspbian Operating System:
Raspbian operating system is the free and open
source operating system which Debian based and
optimized for Raspberry Pi. It provides the basic set
of programs and utilities for operating Raspberry Pi.
Fig 4. It comes with around 35,000 packages which are pre-
In node3, moisture sensor transmits the data using compiled software that are bundled in a nice format
HT12E Encoder IC and a RF transmitter. The for hustle free installation on Raspberry Pi.It has
transmitted data is received by node2 and there it is good community of developers which runs the
processed by microcontroller in order to control the discussion forms and provides solutions to many
operation of water pump. relevant problems. However, Raspbian OS is still
under consistent development with an main focus on
Burned in the microcontroller in order to perform improving the performance and the stability of as
desired operations. This software directly provides many Debian packages as possible.
.hex file which can be easily burned into the
microcontroller. V. EXPERIMENTATION AND RESULTS

b) Proteus 8 Simulator:
Proteus 8 is one of the best simulation software for
various circuit designs of microcontroller. It has
almost all microcontrollers and electronic
components readily available in it and hence it is
widely used simulator. It can be used to test programs
and embedded designs for electronics before actual
hardware testing. The simulation of programming of
microcontroller can also be done in Proteus.
Simulation avoids the risk of damaging hardware due
to wrong design.
c) Dip Trace: Fig 5. Experimental setup for Node1
Dip race is EDA/CAD software for creating schematic
As shown in figure 5, experimental setup for node1
diagrams and printed circuit boards. The developers
consists of mobile robot with central server, GPS
provide multi-lingual interface and tutorials
module, camera and other sensors. All sensors are
(currently available in English and 21 other
successfully interfaced with microcontroller and the
languages). Dip Trace has 4 modules: Schematic
microcontroller is interfaced with the raspberry
Capture Editor, PCB Layout Editor with built-in
pi.GPS and cameras are also connected to raspberry
shape-based auto router and 3D Preview & Export,
pi. Test results shows that the robot can be controlled
Component Editor, and Pattern Editor.
remotely using wireless transmission of PC

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


401
Juno Isabel Susinthra M, K. Gayadri Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 397-403

commands to R-Pi.R-Pi forwards the commands to As shown in above figure, node3 consists of a
microcontroller and microcontroller gives signals to moisture sensor connected to HT12E. Moisture
motor driver in order to drive the Robot.GPS module sensor transmits the data using HT12E Encoder IC
provides the co-ordinates for the location of the robot and a RF transmitter to the Node2 where it is
processed by microcontroller and accordingly water
pump is switched ON/OFF.

VI. CONCLUSION

The sensors and microcontrollers of all three Nodes


are successfully interfaced with raspberry pi and
wireless communication is achieved between various
Nodes. All observations and experimental tests proves
Fig 6. Experimental setup for Node2 that project is a complete solution to field activities,
irrigation problems, and storage problems using
As shown in above figure, node2 consists of motion remote controlled robot, smart irrigation system and
detector, temperature sensor, humidity senor, cooling a smart warehouse management system respectively.
fan, water pump, etc. connected to the Implementation of such a system in the field can
microcontroller board. The sensors gives input to the definitely help to improve the yield of the crops and
controller and according to that microcontroller overall production.
controls the devices in auto mode and also sends the
value of sensors to R-Pi and R-Pi forwards it to user’s VII. ACKNOWLEDGMENT
smart device using internet.Test results shows that
when temperature level increases above preset I am sincerely thankful to all my teachers for their
threshold level then cooling fan is started guidance for my seminar. Without their help it was
automatically in auto mode. The water pump also tough job for me to accomplish this task. I am
gets turned ON if moisture level goes below fixed especially very thankful to my guide Dr. R.S. Kawaka
threshold value. In manual mode, microcontroller for his consistent guidance, encouragement and
receives the controlling signals from R-Pi through motivation throughout the period of this work. I also
ZigBee and accordingly takes the control action. want to thank our Head of the Department (E&TC)
Dr. M. B. Mali for providing me all necessary
facilities.

VIII. REFERENCES

[1] S. R. Nandurkar, V. R. Thool, R. C. Thool,


“Design and Development of Precision
Agriculture System Using Wireless Sensor
Network”, IEEE International Conference on
Automation, Control, Energy and Systems
(ACES), 2014
Fig 7. Experimental setup for Node3

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


402
Juno Isabel Susinthra M, K. Gayadri Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 397-403

[2] JoaquínGutiérrez, Juan Francisco Villa-Medina, Cite this article as :


Alejandra Nieto-Garibay, and Miguel Ángel
Porta-Gándara, “Automated Irrigation System Juno Isabel Susinthra M, K. Gayadri, "Development
Using a Wireless Sensor Network and GPRS of Remote Operated Agri Vehicle", International
Module”,IEEE TRANSACTIONS ON Journal of Scientific Research in Computer Science,
INSTRUMENTATION AND MEASUREMENT, Engineering and Information Technology
0018-9456,2013 (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
[3] Dr. V .Vidya Devi,G. Meena Kumari, “Real- 397-403, January-February 2019. Available at doi :
Time Automation and Monitoring System for https://doi.org/10.32628/CSEIT195177
Modernized Agriculture” ,International Journal Journal URL : http://ijsrcseit.com/CSEIT195177
of Review and Research in Applied Sciences and
Engineering (IJRRASE) Vol3 No.1. PP 7-12,
2013
[4] Y. Kim, R. Evans and W. Iversen, “Remote
Sensing and Control of an Irrigation System
Using a Distributed Wireless Sensor Network”,
IEEE Transactions on Instrumentation and
Measurement, pp. 1379–1387, 2008.
[5] Q. Wang, A. Terzis and A. Szalay, “A Novel Soil
Measuring Wireless Sensor Network”, IEEE
Transactions on Instrumentation and
Measurement, pp. 412–415, 2010
[6] Yoo, S.; Kim, J.; Kim, T.; Ahn, S.; Sung, J.; Kim,
D. A2S: Automated agriculture system based on
WSN. In ISCE 2007. IEEE International
Symposium on Consumer Electronics, 2007,
Irving,TX, USA, 2007
[7] Arampatzis, T.; Lygeros, J.; Manesis, S. A survey
of applications of wireless sensors and Wireless
Sensor Networks. In 2005 IEEE International
Symposium on Intelligent Control & 13th
Mediterranean Conference on Control and
Automation. Limassol, Cyprus, 2005, 1-2, 719-
724

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


403
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195178
Adblock Usage in Web Advertisement
V. Santhi *1, S. Abirami 2
*1Assistant Professor, Department of Computer Applications, Bon Secours College for Women, Bharathidasan
University, Thanjavur, Tamil Nadu, India
2 M.Sc, Department of Computer Science, Bon Secours College for Women, Bharathidasan University,
Thanjavur, Tamil Nadu, India
ABSTRACT

Web users are increasingly turning to ad blockers to avoid ads, which are often perceived as annoying or an
invasion of privacy. While there has been significant research into the factors driving ad blocker adoption and
the detrimental effect to ad publishers on the Web, the resulting effects of ad blocker usage on Web users’
browsing experience is not well understood. To approach this problem, we conduct a retrospective natural field
experiment using Firefox browser usage data, with the goal of estimating the effect of ad blocking on user
engagement with the Web. We focus on new users who installed an ad blocker after a baseline observation
period, to avoid comparing different populations. Their subsequent browser activity is compared against that of
a control group, whose members do not use ad blockers, over a corresponding observation period, controlling
for prior baseline usage. In order to estimate causal effects, we employ propensity score matching on a number
of other features recorded during the baseline period. In the group that installed an ad blocker, we find
significant increases in both active time spent in the browser (+28% over control) and the number of pages
viewed (+15% over control), while seeing no change in the number of searches. Additionally, by reapplying the
same methodology to other popular Firefox browser extensions, we show that these effects are specific to ad
blockers. We conclude that ad blocking has a positive impact on user engagement with the Web, suggesting
that any costs of using ad blockers to users’ browsing experience are largely drowned out by the utility that
they offer.
Keywords: AdBlockPlus, Unblock Origin, Trivial Algorithm.

I. INTRODUCTION turn leads to reduced effectiveness from the point of


view of ad publishers. As a result, Web users are
For an average user, a typical day on the Web increasingly turning to ad blocking to mitigate the
involves exposure to ads. Indeed, advertising has negative effects of online ads. A recent study
become the primary revenue model for many popular estimates that over 600M devices worldwide were
websites, most notably search engines, media outlets using ad blocking by the end of 2016, of which over
and streaming services. There has been substantial half were mobile; this represents a 30% increase since
research into user perception of online ads and the 2015. During the early years of development of the
steps taken to avoid them. Ads are often seen as internet, advertisements displayed on websites were
annoying, or lead to a negative Web browsing not considered as invasive. Usually, they were
experience, and the prevalence of behavioral or presented in the form of static or dynamic banners
retargeted ads raises concerns about privacy. This in which included graphic designs. As technology

CSEIT195178 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 404-408 ]
404
V. Santhi, S. Abirami Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 404-408

developed and IT solutions became more available, Adaptive filtering of advertisements on web pages
the manner in which advertisements were shown on We presented a method to dynamically create custom
websites also developed. filters to avoid downloading unwanted URLs with
minimal interaction with the user, using a weighted-
II. LITERATURE SURVEY majority type algorithm. Both the standalone
extension (AdBlockLearner) and the test driver
Native Advertising and its Effects on Online (AdBlockLearnerTest) are available in [2], including
Advertising Native Advertising and its Effects on source code and documentation. They are compatible
Online Advertising According to Financial Times with most versions of Mozilla Firefox.
Limited (2013), the digital revolution has made many
business models obsolete and advertising companies Adblocking and Counter-Blocking: A Slice of the
see this as a welcome change. Online social Arms Race Adblocking tools like Adblock Plus
networking and internal structural changes in the continue to rise in popularity, potentially threatening
media market have changed how online advertising the dynamics of advertising revenue streams. In
is done. This change has many concurrent forms and response, a number of publishers have ramped up
implications, and the most apparent of them might be efforts to develop and deploy mechanisms for
the change in the various strategies and tactics detecting and/or counter-blocking adblockers (which
companies are adapting in response to said changes; we refer to as anti-adblockers), effectively escalating
mainly advertising online in the form of native the online advertising arms race. In this paper, we
advertising; while one company is inhibiting media develop a scalable approach for identifying third-
profitability in its current form, at the same time party services shared across multiple websites and use
furthering the its corporate sponsors' agenda to it to provide a first characterization of antiadblocking
promote and implement native advertising-or "non- across the Alexa Top-5K websites. We map websites
intrusive" advertising-as the norm of online that perform anti-adblocking as well as the entities
advertising. Abstract According to Financial Times that provide anti-adblocking scripts. We study the
Limited (2013), the digital revolution has made many modus operandi of these scripts and their impact on
business models obsolete and advertising companies popular adblockers. We find that at least 6.7% of
see this as a welcome change. Online social websites in the Alexa Top-5K use anti-adblocking
networking and internal structural changes in the scripts, acquired from 12 distinct entities – some of
media market have changed how online advertising which have a direct interest in nourishing the online
is done. This change has many concurrent forms and advertising industry.
implications, and the most apparent of them might be
the change in the various strategies and tactics Green Mining : Energy Consumption of
companies are adapting in response to said changes; Advertisement Blocking Methods
mainly advertising online in the form of native Extending battery life on mobile devices has become
advertising; while one company is inhibiting media an important topic recently due to the increasing
profitability in its current form, at the same time frequency of smartphone adoption. A primary
furthering the its corporate sponsors' agenda to component of smart phone energy consumption is
promote and implement native advertising-or "non- the apps that run on these devices. Many apps have
intrusive" advertising-as the norm of online embedded advertising and web browser apps will
advertising. show ads that are embedded on webpages. Other
researchers have found that advertising libraries and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


405
V. Santhi, S. Abirami Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 404-408

advertisements tend to increase power usage. But is determinants concerning users blocking
the converse true? If we use advertisement-blocking advertisements and IT tools. The paper refers to this
software will we consume less energy, or will the and systematizes knowledge in the scope of types of
overhead of ad-blocking consume more energy? This online advertisements and methods for blocking
study seeks to determine the effects of advertisements them using an adblock, and it identifies reasons and
on energy consumption, and the effects of attempts to main categories of reasons for users blocking
block the advertisements. We compared different advertisements.
methods of blocking advertisements on an Android
mobile phone platform and compared the power PROPOSED METHODOLOGY
efficiency of these methods. We found many cases
where ad-blocking software or methods resulted in The research presented in the paper was confronted
increased power use. with results of an analysis of application of ad blocks.
The obtained results will facilitate conducting further,
Blocking online advertising-A state of the art. more thorough research. Considerations included in
the paper can constitute a set of recommendations for
Compared with online advertising industry, there is publishers displaying advertisements on websites and
an even faster increase of ad blocker usage, which they can be useful for drawing conclusions and
influence badly on publishers’ and advertisers’ preparing guidelines for projects supporting
business. Thus more and more companies initialize sustainable development in the scope of online
their counter-ad blocking strategies, in which advertising.
customers choose to either disable their ad blockers
or leave without seeing the content. There are also
companies which abandon their counter-ad blocking
strategies after conducting them for a while due to
insufficient understanding of users’ ad blocking
behavior. In this study, we employed a quasi-
experiment framework and collected a large-size data
with the cooperation with Forbes Media. We aim to
identify factors influencing ad blocker usage.
Furthermore, we will model the interaction effects
among user profile, online behavior patterns, device
features on ad blocker usage propensity. Our study
contributes the literature of understanding ad blocker Connecting Network and Data Exchanging: -
usage by evaluating those principles using big amount
of real-world data. Computer network that interconnects computers or
systems within a large-scale network among internet
III. METHODS AND MATERIAL networks, it connects individual computer terminals,
mobile devices, and computer networks to the
EXISTING PROCESS Internet, enabling users to access Internet services,
Research concerning users blocking advertisements such as email and the software application. Network
constitutes a new research area both in the scope of computer devices that originate, route and terminate
analysis of collected data regarding that topic, the data are called network nodes. Nodes can include

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


406
V. Santhi, S. Abirami Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 404-408

hosts such as personal computers, phones, servers as Nodes Protected from Ad Block:-
well as networking hardware. Two such devices are Using metropolis sampler makes it more security for
said to be networked together when one device is Internet activity to be traced back to the user during
able to exchange data information with the other the visits of Web sites, online posts, and other
device, whether or not they have a direct connection communication forms. It intended to protect the
to each other. personal privacy of users, as well as their freedom
and ability to conduct confidential communication
Ad Block occurred (i.e. Ad Block Software):- by keeping their Internet activities from being
monitored. Assuming the networks is fully
Malicious software, is any software used to disrupt trustworthy, the system node must provide the
computer operation, gather sensitive information, or means to properly authenticate to connect other
gain access to private computer systems. Malware is nodes. Other nodes may impersonate trusted
defined by its malicious intent, acting against the computers, thus requiring system authentication. The
requirements of the computer user, and does not nodes and user may be trusted but within an
include software that causes unintentional harm due untrustworthy environment.
to some deficiency. The term badware is sometimes
used, and applied to both true (malicious) malware IV. RESULTS AND DISCUSSION
and unintentionally harmful software. During the
data transferring or downloading Computer users CONCLUSIONS
often download malware (i.e. malicious software) to The paper shows the reasons for using adblock
their computer by unknowingly visiting a malicious programs and presents the current most popular
webpage hosting a drive-by download attack, programs of that type as well as their characteristics.
clicking on a malicious link included in email, Apart from aforementioned, a survey study was
opening an attachment which includes an exploit, conducted involving a group of Polish internet users,
and destroy the data. which shows that they actively and consciously block
the display of advertisements in their web browsers.
Detecting, Blocking using Check Sum and Metropolis In this study we show how to adblock software
Sampler:- technologies are applied in Polish environment. We
discover a strong desire to block advertisements and
Anti-malware software is computer software used to well known ways to do this. Also in Poland there are
prevent, detect and remove the virus software. But many adapted solutions to improve adblock accuracy
the malware (i.e. malicious software) occurred during and coverage.
download the software or clicking the malware link
in spite of clicking the e-mail unknowingly it affects FUTURE ENHANCEMENT
entire system. Check sum algorithm was originally Future directions of research will concern the
developed to detect malware based on the priority of problem of intrusiveness of advertisements and ways
other kinds of malware during the data processing. for decreasing it. As it was mentioned in the paper,
And encounter based algorithm is used to block the users are not against advertisements being
malware (i.e. malicious software). Metropolis Sampler displayed—they oppose the way in which they
algorithm is to block and encounter the malware. To appear on their screen, e.g. by suddenly covering the
prevent the nodes and against to filtered the thread. viewed content. Further research will be also aimed
And it is recover the data in a large scale systems. at a quantitative analysis regarding elements blocked

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


407
V. Santhi, S. Abirami Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 404-408

on websites, such as unwanted advertisements,


scripts tracking user's activity and malicious activities
(malware).

V. REFERENCES

[1]. Adblock Plus, https://adblockplus.org, last


accessed 2017/07/07
[2]. AdBlock, https://getadblock.com, last accessed
2017/07/07
[3]. Ajdari, D., Hoofnagle, C., Stocksdale, T., Good,
N.: Web Privacy Tools and Their Effect on
Tracking and User Experience on the Internet
(2013)
[4]. AlleBlock, https://tarmas.pl/alleblock, last
accessed 2017/07/07
[5]. Allowing acceptable ads in Adblock Plus,
https://adblockplus.org/en/acceptable-ads, last
accessed 2017/07/07
[6]. CertyficateIT, https://www.certyficate.it, last
accessed 2017/07/07
[7]. EasyList, https://easylist.to, last accessed
2017/07/07
[8]. Eyeo GmbH, https://eyeo.com/en/press#stats,
last accessed 2017/07/07
[9]. Google, Microsoft and Amazon pay to get
around ad blocking tool,
https://www.ft.com/content /80a8ce54-a61d-
11e4-9bd3-00144feab7de, last accessed
2017/07/07

Cite this article as :

V. Santhi, S. Abirami, "Adblock Usage in Web


Advertisement", International Journal of Scientific Research
in Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume 5
Issue 1, pp. 404-408, January-February 2019.
Available at doi : https://doi.org/10.32628/CSEIT195178
Journal URL : http://ijsrcseit.com/CSEIT195178

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


408
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195190
Smart-Home Automation Using IoT-Based Sensing and
Monitoring Platform
R. Sugantha Lakshmi1, P. Karthika2, A. Rajalakshmi3, M. Sathya4
1 Assistant Professor, Department of Computer Science and Engineering, Kings College of Engineering,
Thanjavur, Tamil Nadu, India
2Department of Computer Science and Engineering, Kings College of Engineering, Thanjavur, Tamil Nadu,
India
ABSTRACT

To comfort the living conditions within of a human his home, she can use home monitoring and automation
are utilized. The standards of human’s comfort in homes can be categorized into several types. Among these
categories, the most significant ones are the thermal comfort, which is related to temperature and humidity,
followed by the visual comfort, related to colors and light, and hygienic comfort, associated with air quality.We
are proposing a system to monitoring the paremeters range which will increase the comfortness of the
human.Additionally,we incorporate the smart home technique which controls the home appliances by
intelligent automatic execution of commands after analyzing the collected data. Automation can be
accomplished by using the Internet of Things (IoT). This gives the inhabitant accesses to certain data in the
house and the ability to control some parameters remotely. We propose the complete design of an IoT based
sensing and monitoring system for smart home automation. That uses the EmonCMS platform for collecting
and Visualizing monitored data and remote controlling of home appliances and devices. The selected platform
is very flexible and user-friendly. The sensing of different variables inside the house is conducted using the
NodeMCU-ESP8266 microcontroller board, which allows real-time data sensing, processing and
uploading/downloading to/from the EmonCMS cloud server.
Keywords : IOT, CMS, Humidity, Air Quality, Temperature

I. INTRODUCTION load, but smart home technology is combination of


both and much more. Smart home technology is
Smart Home is collaboration of technology and currently being implemented for entire house in
services through a network for better quality living. particularly kitchen and living room. Basically, smart
A smart home allows the entire home to be home facilitates users with security, comfortable
automated and therefore provide ease and living and energy management features as well as
convenience to everyday activities in the home. This added benefits for disabled individuals.
technology is used to make all electronic devices to
act ‘smart’. In the near future almost all the electronic This technology might sound new but it just uses the
devices will take advantage of this technology existing technologies. A smart device is a common
through home networks and the internet. Many appliance with a much more complex computer
people think this technology as pure networking. installed to give it more functionality. These
Others think this technology will reduce their work functions are the ones which makes it so different.

CSEIT195190 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 409-413 ]
409
R. Sugantha Lakshmi et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 409-413

Cable broadband, DSL, Bluetooth and wireless ZigBee modules due to their low power and cost
technologies provide a way to have a home efficiency.
networked for devices to communicate with each
other as well as internet. These technologies either D. GSM Based Home Automation System
wired or wireless provide foundation on which smart A smart home automation system implemented by
home will operate. using Global System for Mobile communication
(GSM). The hardware architecture of the system
II. LITERATURE SURVEY consists of GSM modem, PIC16F887 microcontroller
and smartphone. The system used a GSM modem to
A. Bluetooth based home automation system control electric appliances through SMS request.
Home automation systems using smartphone, PIC16F887 microcontroller interfaced with a GSM
Arduino board and Bluetooth technology are secured modem and it is used to read and decode the received
and low cost. A Bluetooth based home automation SMS to execute the specific command.
system proposed by R.Piyare and M.Tazil. The
hardware architecture of this system consists of the F. EnOcean based home automation system
Arduino BT board and a cell phone, the The EnOcean is newly developing energy harvesting
communication between Arduino BT board and cell technology used in transportation, building and home
phone is wirelessly using Bluetooth technology. The automation systems. EnOcean’s technology is
Arduino BT board has a range of 10 to 100 meters, 3 productively in logistics as well as in the industry due
Mbps data rate and 2.4 GHz bandwidth. to energy efficiency and easily installing device
anywhere for users ease which significantly save the
B. Voice recognition based home automation installation cost up to 40%. Moreover EnOcean’s
Voice recognition based home automation system devices utilize 315 MHz band and it provides
proposed and implemented by a researcher. The convenient ways for home automation system.
hardware architecture of this system consists of
Arduino UNO and smartphone. The wireless III. EXISTING PROCESS
communication between the smartphone and the
Arduino UNO is done through Bluetooth technology. Smart home technology has the potential of bringing
Android OS has a built-in voice recognizing feature benefits to modern households and their inhabitants.
which is used to develop a smartphone application Yet, ever since its early development, it has been
which has ability to control the home appliances struggling to reach mass consumer adoption. Privacy
from user voice command. and security, trust issues, reliability, and price are just
some of the challenges smart home technology is
C. ZigBee Based Wireless Home Automation System facing. In addition, literature suggests that there is an
ZigBee based wireless home automation system has evident gap between the functionalities offered by
also been studied, it consists of three main modules, smart devices and users’ needs. Investigating these
handheld microphone module, central controller potential adoption challenges in some more detail, we
module and appliance controller module. Handheld conducted an interview study with existing smart
microphone module use ZigBee protocol and central home technology users. Results show that privacy
controller module are based on PC. In this system, and security are still the most prominent hindering
Microsoft speech API is used as a voice recognition factors, and that the often insufficient
application, wireless network is established using RF interoperability of devices becomes an ever-growing

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


410
R. Sugantha Lakshmi et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 409-413

concern. Also, smart home devices are consistently humidity, luminosity, and air quality (i.e. CO2 and
perceived as complex and expensive, and lack dust levels). Usually, the term “Big Data” is often
perceived value and trustworthiness. associated with IoT and refers to the large number
data collected from a large number of sensors and
IV. PROPOSED METHODOLOGY devices to be processed, which is the case here. In
fact, depending on the number of sensors and the
As mentioned earlier, the monitored parameters in measurement-sampling rate, a large number of data
homes are selected to ensure a smart, efficient and might be generated and transmitted through the
comfortable living conditions. Therefore, one hub is internet to the cloud server, which should be able to
made to measure temperature, humidity, light handle such large amount of data communication and
intensity, proximity, and CO2 levels. This hub can be archiving.
placed in each room of the house, and preferably at a
high level above the ground and below the ceiling. B. Data Transmission
Collected data can be processed and used to control Transmitting data between devices and controllers is
smart devices such as smart LEDs and A/Cs. The done using one or more communication technologies.
schematics of the proposed design. It comprises These include Bluetooth, ZigBee, Wi-fi, Ethernet or
mainly sensors, processor, user interface, and data GSM. Bluetooth and ZigBee are commonly used for
communication and transmission links. in-home data transmission and control and provide
the link between the sensors and the central
processing unit. These data communication
technologies are popular because of their low power
consumption, and simplicity to implement.
Furthermore, IoT requires the use of either Wi-fi or
Ethernet to connect to the internet. While Ethernet
is by far faster than Wi-fi connection, the high data
rate is not essential for home automation applications.
Additionally, Wi-fi has the advantage of mobility,
making it more widespread in most systems. Wi-fi
technology consumes more energy than alternatives
such as Bluetooth or ZigBee. The power consumption
can be reduced by lowering the frequency of data
V. IMPLEMENTATION RESULTS uploading.

This section reviews the typical configuration of C. Data Processing (Microcontroller)


home automation system using an IoT platform. The data collected in a home automation system is
Illustrates such a platform which consists mainly of usually processed and managed by a microcontroller
data sensing and acquisition, processing, transmission, such as Arduino, Raspberry Pi, and NodeMCU.
and display. Raspberry Pi is a small single microcontroller
computer. With a higher RAM of 256MB or 512Mb,
A. Data Sensing and Acquisition depending on the model, it can handle more complex
To ensure a good comfort level, the typically tasks than other controllers and used mostly as a
monitored parameters in houses are temperature, central processing unit for multiple devices. Most

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


411
R. Sugantha Lakshmi et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 409-413

new models of Raspberry Pi have USB and Ethernet VI. CONCLUSION


ports, making it easy to upload data to the internet.
On the other hand, Arduino is a single-board The proposed design of the smart solar home is very
microcontroller that can be simply programmed to flexible and can be easily expanded and applied to
execute commands. Arduino comes in a variety of larger buildings by increasing the number of sensors,
models with onboard flash memory ranging from measured parameters, and control devices. More
32kB to 512kB, and a typically a RAM of 2kB. functionality and smartness could be also added to
Evidently, this controller is less powerful than the the existing system for making the house automation
Raspberry Pi. However, most Arduino models are system grow, adapt, and evolve by itself using
cheaper, easier to handle, and powerful enough to advanced artificial intelligence.
deal with home automation tasks. Another option is
to use the Node MCU. It is an Arduino based VII. FUTURE ENHANCEMENT
microcontroller but with the addition of the ESP8266
Wi-Fi chipset. This microcontroller has a memory of There are a variety of enhancements that could be
128kB and 4MB storage. It is mostly used for a single made to this system to achieve greater accuracy in
IoT application, or to eliminate the need for a central sensing and detection.
processing unit. Since each part of the system can a) There are a lot of other sensors that can be used to
upload data to the server individually, this also increase the security and control of the home like
lowers the complexity of the coding and the pressure sensor that can be put outside the home to
connection chain. The biggest advantage of the Node detect that someone will enter the home.
MCU over the other alternatives is the significantly b) Changing the way of the automated notifications
low price for a controller that can connect to the by using the GSM module to make this system more
internet directly using Wi-fi, without the need for professional.
any additional peripherals of modules. An issue is c) A smart garage that can measure the length of the
that the Node MCU board has only one analog input, car and choose which block to put the car into it and
which limits its applications to as single data it will navigate the car through the garage to make
monitoring system. However, this drawback can be the parking easy for the homeowner in his garage.
compensated for by using the ASD115, which is an
analog to digital converter that has four analog input VIII. REFERENCES
pins and has a higher conversion resolution of 16-bits.
[1]. Hayet Lamine and Hafedh Abid, "Remote
D. Data Display & User Interface control of domestic equipment from an
The interaction between the user and the system can Android application based on Raspberry Pi
be done in several ways. One option is to use an card", IEEE transaction 15th international
application. There are many simple means to create a conference on Sciences and Techniques of
mobile or a web based app to display data, even with Automatic control & computer engineering -
a limited knowledge of programming. Another STA'2014, Hammamet, Tunisia, December 21-
control option is through mobile GSM where the user 23, 2014.
can send commands in codes by SMS to the [2]. Vaishnavi S. Gunge and Pratibha S. Yalagi,
microcontroller. This control method requires a "Smart Home Automation: A Literature
special GSM module added to the circuit. This Review", National Seminar on Recent Trends in
method can also be applied using emails. Data Mining- RTDM 2016.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


412
R. Sugantha Lakshmi et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 409-413

[3]. M. B. a. W.Fincher, "Standards of Human International Conference on Renewable


Comfort: Relative and Absolute," The Energies and Vehicular Technology, 2012.
University of Texas at Austin, Austin, 2009.
[4]. Archana N. Shewale and Jyoti P. Bari,
"Renewable Energy Based Home Automation Cite this article as :
System Using ZigBee", IJCTEE Volume 5, Issue
R. Sugantha Lakshmi, P. Karthika, A. Rajalakshmi, M.
3, June 2015.
Sathya, "Smart-Home Automation Using IoT-Based
[5]. Ali Mohammed Al-Kuwari, Cesar Ortega-
Sensing and Monitoring Platform", International
Sanchez, Atif Sharif and Vidyasagar Potdar,
Journal of Scientific Research in Computer Science,
"User-Friendly Smart Home Infrastructure:
Engineering and Information Technology
BeeHouse", IEEE 5th 5th International
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
Conference on Digital Ecosystems and
409-413, January-February 2019. Available at doi :
Technologies, May31-June3 2011, Daejeon,
https://doi.org/10.32628/CSEIT195190
Korea.
Journal URL : http://ijsrcseit.com/CSEIT195190
[6]. Lamir Shkurti, Xhevahir Bajrami, Ercan
Canhasi, Besim Limani, Samedin Krrabaj and
Astrit Hulaj, "Development of Ambient
Environmental Monitoring System Through
Wireless Sensor Network Using NodeMCU and
WSN Monitoring", 6th MEDITERRANEAN
CONFERENCE ON EMBEDDED
COMPUTING (MECO), JUNE 11-15 2017,
BAR, MONTENEGRO.
[7]. International Electrotechnical Commission,
Internet of Things: Wireless Sensor Networks,
Geneva, Switzerland: International
Electrotechnical Commission, 2014. [8]
Grégory Nain, Francois Fouquet, Brice Morin,
Olivier Barais, JeanMarc Jézéquel, "Integrating
IoT and IoS with a Component-Based
approach", 36th EUROMICRO Conference on
Software Engineering and Advanced
Applications, 2010
[8]. S. Jaffe, "DESIGN OF INEXPENSIVE AND
EASY TO USE DIY INTERNET OF THINGS
PLATFORM," Faculty of California Polytechnic
State University, 2016.
[9]. F. Touati, M. Al-Hitmi, and H. Bouchech,
"Towards understanding the effects of climatic
and environmental factors on solar PV
performance in arid desert regions (Qatar) for
various PV technologies," 2012 First

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


413
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195198
Anticipated Security Model for Session Transfer and Services
Using OTP
J. Jose Merlin1, A. Prathipa2, G. Ramyadevi3, P. Radhika4
Department of Computer Science and Engineering, St. Joseph’s College of Engineering and Technology,
1-3

Thanjavur, Tamil Nadu, India


4 Assistant Professor, Department of Computer Science and Engineering, St. Joseph’s College of Engineering and
Technology, Thanjavur, Tamil Nadu, India
ABSTRACT

Internet security is a branch of computer sciences often involving browser security, network security,
applications and operating systems to keep the internet as a secure channel to exchange information by
reducing the risk and attacks. There are a number of studies that have been conducted in this field resulting in
the development of various security models to achieve internet security. However, periodic security reports and
previous studies prove that the most secure systems are not immune from risk and much effort is needed to
improve internet security. This paper proposed a simple security model to improve internet applications
security and services protections, specified access control, cryptographic, cookies and session managements,
defense programing practices, care for security from early stage on development life cycle, use hardware
authentication techniques in access control, then propose cryptographic approach by mix MD5 with Based64,
consider session and cookies types and ways to keep it secure. Additionally, these practices discussed the most
important web security vulnerability and access control weakness and how to overcome such weaknesses,
proposed an approach to measure, analyze and evaluate security project according to software quality standard
ISO 25010 by using Liker scale, finally ended by case study. The effort of this paper represents a set of
techniques and tips that should be applied within each web application development process to maintain its
security.
Keywords : Combinationarial Optimization Identification, Public Key Infrastructure, Partial Forward Secrecy.

I. INTRODUCTION as efficiency, simplicity, and cost effectiveness. On


the other hand, these enterprises face a new
Web applications emerged as one of the main challenge on how to keep these applications and
technologies and progressed in the last decade. They services secure. They are spending many resources to
play a main role in our daily lives on areas such as manage and handle data storage in plain text in malt
health-care, online services, businesses (public and locations throughout the enterprise. Recently, the
private sectors), E-banking, and e-commerce. Web security of web applications and services derived
applications are expected to increase the trading attention from both the industry field and research
volume on the market exceeding $ 1 trillion in the community. There is a strong need for standard web
next several years. Many public or private enterprises security models for several reasons including: the
developed their applications and services based on integration of the heterogeneous and distributed
the internet to take advantage of its features as well systems, availability of high volumes of sensitive

CSEIT195198 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 414-418 ]
414
J. Jose Merlin et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 414-418

information and data maintained by corporation's application development process to maintain its
servers and government agencies, safety from easily security.
distributed malicious software, and computer crimes. Security Issues for Web-Based Applications: Issues
On the other hand, software security has become a And Solutions For The Safe Transfer Of Clinical
quality attribute used to measure the quality of Trials Data Over The Internet When building any
software according to standards such as ISO software web-based application, the issue of security is always
quality standard 25010. a concern. When the application is designed to allow
the sharing of clinical trials data across the Internet,
II. LITERATURE SURVEY security becomes one of the major topics. This paper
describes what approach iBiomatics, a SAS company,
Proposed Security Model For Web Based has taken in the construction of their biometrics
Applications And Services Portal. The paper presents a basic overview of
Internet security is a branch of computer sciences security for an Internet based application, describes
often involving browser security, network security, what security features were built into the application,
applications and operating systems to keep the what software tools were used to implement various
internet as a secure channel to exchange information aspects of the security model, and how the actual
by reducing the risk and attacks. There are a number hardware configuration can be a critical piece of the
of studies that have been conducted in this field overall security of a web based application.
resulting in the development of various security Securing Frame Communication in Browsers
models to achieve internet security. However, Many web sites embed third-party content in frames,
periodic security reports and previous studies prove relying on the browser’s security policy to protect
that the most secure systems are not immune from them from malicious content. Frames, however, are
risk and much effort is needed to improve internet often insufficient isolation primitives because most
security. This paper proposed a simple security model browsers let framed content manipulate other frames
to improve internet applications security and services through navigation. We evaluate existing frame
protections, specified access control, cryptographic, navigation policies and advocate a stricter policy,
cookies and session managements, defense which we deploy in the open source browsers. In
programming practices, care for security from early addition to preventing undesirable interactions, the
stage on development life cycle, use hardware browser’s strict isolation policy also hinders
authentication techniques in access control, then communication between cooperating frames. We
propose cryptographic approach by mix MD5 with analyze two techniques for inter-frame
Based64, consider session and cookies types and ways communication. The first method, fragment
to keep it secure. Additionally, these practices identifier messaging, provides confidentiality without
discussed the most important web security authentication, which we repair using concepts from
vulnerability and access control weakness and how to a well-known network protocol. The second method,
overcome such weaknesses, proposed an approach to post Message, provides authentication, but we
measure, analyze and evaluate security project discover an attack that breaches confidentiality. We
according to software quality standard ISO 25010 by modify the post Message API to provide
using Likert scale, finally ended by case study. The confidentiality and see our modifications
effort of this paper represents a set of techniques and standardized and adopted in browser
tips that should be applied within each web implementations.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


415
J. Jose Merlin et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 414-418

Protecting Browsers From Dns Rebinding Attacks the secret key or password shared between two users.
DNS rebinding attacks subvert the same-origin policy The main goal of password-based authenticated key
of browsers, converting them into open network exchange protocols is to restrict the adversary to this
proxies. Using DNS rebinding, an attacker can case only. The existing system consider that
circumvent organizational and personal firewalls, password-based authenticated key exchange in the
send spam email, and defraud pay-per-click three-party scenario, in which the users trying to
advertisers. We evaluate the cost effectiveness of establish a secret do not share a password between
mounting DNS rebinding attacks, finding that an themselves but only with a trusted server. In this
attacker requires less than $100 to hijack 100,000 IP existing system the key exchange between three
addresses. We analyze defenses to DNS rebinding authenticated users so, that the privacy of the system
attacks, including improvements to the classic “DNS is failed to overcome. Since we want to trust as little
pinning,” and recommend changes to browser plug- as possible the third party, we develop a new notion
ins, firewalls, and Web servers. Our defenses have called key privacy which roughly means that, even
been adopted by plug-in vendors and by a number of though the server’s help is required to establish a
open-source firewall implementations. session key between two users in the system, the
server should not be able to gain any information on
Assessment of Access Control Systems the value of that session key.
Adequate security of information and information
systems is a fundamental management responsibility. IV. PROPOSED METHODOLOGY
Nearly all applications that deal with financial,
privacy, safety, or defense include some form of A secure 2-party password-based key exchange is a
access control. Access control is concerned with 2PAKE protocol where the parties use their password
determining the allowed activities of legitimate users, in order to derive a common session key that will be
mediating every attempt by a user to access a used to build secure channels. Several practical
resource in the system. In some systems, complete schemes can be used in the instantiation of the 2-
access is granted after successful authentication of the party password-based key exchange of our generic
user, but most systems require more sophisticated construction. In these proposed system, instead of
and complex control. In addition to the exchange the password the owner going to send the
authentication mechanism (such as a password), link to get that particular data and in user side by
access control is concerned with how authorizations using that link they can retrieve that data easily.
are structured. In some cases, authorization may Once the user make logout means they cannot again
mirror the structure of the organization, while in enter into the system. The link can be used only once.
others it may be based on the sensitivity level of
various documents and the clearance level of the user V. IMPLEMENTATION RESULTS
accessing those documents. This publication explains
some of the commonly used access control services Search engine recognized
available in information technology systems. Search engine indexing collects, parses, and
stores data to facilitate fast and accurate information
III. EXISTING PROCESS retrieval. Index design incorporates interdisciplinary
concepts from linguistics, cognitive psychology,
Password-based authenticated key exchange is mathematics, informatics, and computer science. An
protocols which are designed to be secure even when alternate name for the process in the context

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


416
J. Jose Merlin et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 414-418

of search engines designed to find web pages on the industry-specific vertical search engines. SEO differs
Internet is web indexing. from local search engine optimization in that the
Popular engines focus on the full-text indexing of latter is focused on optimizing a business' online
online, natural language documents. Media presence so that its web pages will be displayed by
types such as video and audio and graphics are also search engines when a user enters a local search for
searchable. its products or services. The former instead is more
focused on national or international searches.
Meta search engines reuse the indices of other
Authentication
services and do not store a local index, whereas
An authentication module is a plug-in that collects
cache-based search engines permanently store the
user information such as a user ID and password, and
index along with the corpus. Unlike full-text indices,
compares the information against entries in a
partial-text services restrict the depth indexed to
database. If a user provides information that meets
reduce index size. Larger services typically perform
the authentication criteria, the user is validated and,
indexing at a predetermined time interval due to the
assuming the appropriate policy configuration,
required time and processing costs, while agent-
granted access to the requested resource. If the user
based search engines index in real time.
provides information that does not meet the
authentication criteria, the user is not validated and
denied access to the requested resource. OpenSSO
Enterprise is deployed with a number of
authentication modules.

Session Handling
A web session is a sequence of network HTTP request
and response transactions associated to the same user.
Modern and complex web applications require the
retaining of information or status about each user for
the duration of multiple requests.

User preference
Optimized search results
Understanding user preference settings. Completing
Search engine optimization (SEO) is the process of
the user preferences pages enables the system
affecting the online visibility of a website or a web
to use the field values that you specify as the default
page in a web search engine's unpaid results—often data that appears on various inquiry pages. ... This
referred to as "natural", "organic", or "earned" results.
enables default values to be set by user ID; the
In general, the earlier (or higher ranked on the
user can override the defaults when running an
search results page), and more frequently a website
inquiry.
appears in the search results list, the more visitors it
will receive from the search engine's users; these Temporary User Restriction
visitors can then be converted into To ensure a consistent user experience, we monitor
customers.[1] SEO may target different kinds of the level of requests to the my hr toolkit system from
search, including image search, video user accounts. If an abnormally high frequency of
search, academic search,[2] news search, and requests is detected, an escalating blocking process is
deployed

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


417
J. Jose Merlin et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 414-418

VI. CONCLUSION offers improvements in quality and video


Thus the process a secure 2-party password-based key conferencing applications in the near future.
exchange is a 2PAKE protocol where the parties use
their password in order to derive a common session VIII. REFERENCES
key that will be used to build secure channels. [1]. Michel Abdalla, Pierre-Alain Fouque, and David
Several practical schemes can be used in the Pointcheval, "Password-Based Authenticated Key
instantiation of the 2-party password-based key Exchange in the Three-Party Setting", IEEE
Proceedings on InformationSecurity, Volume 153,
exchange of our generic construction. In these
number 1, March 2006 , pp. 27-39.
proposed system, instead of exchange the password
[2]. Whitefield Diffie, Martin E. Hellman, "New directions
the owner going to send the link to get that particular
in cryptography", IEEE Transactions on Information
data and in user side by using that link they can Theory, pp. 644-654,1976.
retrieve that data easily. Once the user makes logout [3]. Behrouz A. Forouzan, Cryptography and network
means they cannot again enter into the system. The security, Tata McGraw-Hill, 2007.
link can be used only once. [4]. Hyun-SeokKim ,Jin-Young Choi, "Enhanced
passwordbased simple three-party key exchange
VII. FUTURE ENHANCEMENT protocol", Elsevier, Computers and Electrical
The introduction of VoIP has enabled many Engineering, 35, pp.107–114,2009.
[5]. Yuanhui Lin, MengboHou, Qiuliang Xu, "Strongly
organizations to supplement or replace existing
password based three party authenticated key
circuit switched telephone networks, providing a
exchange protocol", Ninth International conference on
new value added services. Numerous technological
Computational Intelligence and security, IEEE, pp.
problems are faced on the issue of quality of service. 555- 558,2013.
The packet loss is a major problem. This problem [6]. Rongxing Lu, Zhenfu Cao, "Simple three-party key
cannot be avoided but it can be minimized. The exchange protocol", Elsevier, computers & security 26,
packet repetition with static playout algorithm pp. 94-97,2007.
performs moderately than other replacement [7]. Chao Lv , MaodeMab, Hui Li, JianfengMaa, Yaoyu
techniques. Based on time delay, loss rate and MOS, Zhang, "An novel three-party authenticated key
the better results are obtained. Hence, the receiver exchange protocol using one-time key", Elsevier,
Journal of Network and Computer Applications (36),
based repetition provides high quality VoIP signal in
pp. 498–503,2013.
the streaming audio. The processing of damaged
[8]. Alfred Menezes, BerkantUstaoglu, "On reusing
packets has been established as a challengeable topic
ephemeral keys in Diffie Hellman key agreement
for further research. This work can be extended by protocol",International Journal of Applied
incorporating other statistical based methods. Voice Cryptography,ACM,pp. 154-158, 2010.
over IP provides unique and inexpensive services to
the internet people. Internet is a single network Cite this article as : J. Jose Merlin, A. Prathipa, G.
shared by the world. Audio conferencing is one of Ramyadevi, P. Radhika, "Anticipated Security Model for
the applications used nowadays in the internet Session Transfer and Services Using OTP", International
community. This is purely based on multiplexing and Journal of Scientific Research in Computer Science,
multicasting scheme. It promises to deliver cost Engineering and Information Technology (IJSRCSEIT),
ISSN : 2456-3307, Volume 5 Issue 1, pp. 414-418, January-
savings to users and service providers. Hence the
February 2019. Available at doi :
Voice M-M based interleaving provides high quality
https://doi.org/10.32628/CSEIT195198
VoIP signal in the streaming audio. This work can be
Journal URL : http://ijsrcseit.com/CSEIT195198
extended by incorporating video based methods. It

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


418
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951106
Managing university defined Study material by using Online
Portal
Neha Bitodkar #, Prof. Aniket Bhoyar*, Suraj Chavhan#, Priyanka Wahane#
#
Department of Information Technology, Rashtrasant Tukadoji Maharaj Nagpur University,
Nagpur, Maharashtra, India
*J D College of Engineering and Management, Nagpur, Maharashtra, India
ABSTRACT

E-learning system is very much popular in today`s digital era due to which one can easily get access to the
required information or knowledge. Due to online learning there are several solutions to a single problem is
available as the internet has grown in the largest manner. Even the online learning is changing the load of
traditional classrooms. E-learning provides an integrated environment where access to particular information
can be found at one place in various forms. The proposed technique is for creating an E-learning system with an
integrated environment for students of higher education. So that, they are able to get access to an appropriate
solutions for the particular question asked in an university examination. The system will provide free access to
the study material for students simply by logging in. This will be creating an integrated library for each subject
and create model answer section for questions related to the subject. This will help student to learn subject
easily, the question and their answer that can be asked in the university examination.
Keywords: Raspberry pi, USB camera, motor driver, motor, LCD display, road side speed sign controlling

I. INTRODUCTION question papers, model answer paper, video lecture,


E-books etc.
In today`s digital era online education is not a big
deal. There are several online courses started by govt This system is designed using technologies such as
for the online learning of students. Due to internet a Bootstrap, CSS and PHP. The database used is Mysql
large no. of information can be available of particular to store the login ID of the users and their data. This
subject. Where as integrated environment if designed technologies will make an integrated environment
there are less chances of confusion about the subject. that will contain information in several formats.
According to the syllabus designed by the university
examination, and the question paper set, model II. PROBLEM STATEMENT
answers section is designed which will make it easy
for the students to get extraordinary knowledge of To create a database that will manage study material,
the syllabus. the system is created of sections such as question
paper, model answer, E-books, semester-wise subject
This system will make students available with differentiation. Basically, this will provide the
complete information about the subject. The syllabus students all the study material of higher education at
will be covered completely in the question answer one place. So that it will be easier for them to study
section. The study material includes syllabus,

CSEIT1951106 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 419-421 ]
419
Neha Bitodkar et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 419-421

without any confusion as there are several solutions Author : Yogesh Deshpande , Samit Bhattacharya,
are available online for a single solution. Pradeep Yammiyavar
Description: This paper focuses on analyzing the
III. LITERATURE REVIEW ability of children`s for adopting e-learning
environment in comparison with traditional
Title 1 : Implementation of E-learning on full stack classrooms. Online learning can be an additional tool
concept for the school environment to get more knowledge
Author: Nitish Kumar, Chaitali Alhat, Rupali about the particular subject.
Bachhav
Description : The designed system provides materials IV. PROPOSED SYSTEM
in different formats. It can be in the form of videos,
pdf etc. Their purpose is to provide an open free The proposed system will be a beneficiary product for
course which provides access to the material after the the students of university and will get a complete
registration process is completed. This system focuses access to all the material related to their syllabus
on interactive learning between the students and the designed by university.
proffesors.
Title 2: Open educational resource : opportunities The system is designed using Bootstrap, CSS and PHP.
and challenges for Indian higher education. Mysql database is used for storing the login ID as well
Author : Dr. Indrajeet Dutta as the data of the user. The system will allow students
Description : This paper gives information about the for complete study of subjects due to the question
opportunities and challenges for Indian higher answer section is available. Question papers will be
education related with open educational resources. available on the subjects along with their model
This paper gives the detailed information about the answer.
several online courses started by the govt. such as
NMECIT, NKN, SHAKSHAT, NPTEL, EKLAVYA, The designed system contains information in
OSCAR E-GRID etc. This papers says that the different formats. It will also contain information in
opportunities can be increased as students can hear the form of video lectures and e-books. The system is
the lectures from international proffesors. branch wise and semester wise separated according to
Title 3 : Role of consortia for effective use of e- the university.
resources in Higher education : A practical approach Data Flow Diagram:
in Indian liabraries.
Author: Sunil Goria
Description : This paper says that in the last decade
the liabraries of higher education and the research
institutes are using the e-resources. In consortium
mode, one can be able to access the journals in Indian
libraries using e-resources. This paper focuses on
increasing use of e-resources.
Title 4: A behavioural approach to modelling Indian
children`s ability of adopting to e-learning
environment.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


420
Neha Bitodkar et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 419-421

V. CONCLUSION Cite this article as :

In this work, after login the study material will be Neha Bitodkar, Prof. Aniket Bhoyar, Suraj Chavhan,
available for the students. This is a beneficiary Priyanka Wahane, "Managing university defined
product for the students. University syllabus will be Study material by using Online Portal", International
available on this website along with related material Journal of Scientific Research in Computer Science,
such as question answer set, video lectures, model Engineering and Information Technology
answer, e-books, notes.It will help the students for (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
complete learning of subjects designed by university. 419-421, January-February 2019.
Due to E-learning one can access this information Journal URL : http://ijsrcseit.com/CSEIT1951106
form anywhere at anytime.

VI. REFERENCES

[1]. Nitish Kumar, Chaitali Alhat, Rupali Bachhav


"Implementation of E-learning system on full
stack concept", Mat Journals year 2018
[2]. Dr. Indrajeet DUTTA," Open educational
resources (oer): opportunities and challenges
for indian higher education", Turkish Online
Journal of Distance Education -TOJDE April
2016
[3]. Sunil Goria, "Role of consortia for effective use
of e-resources in higher education: A practical
approach in Indian libraries."International
Journal of information Dissemiation &
Technology, 2012
[4]. Yogesh Deshpande , Samit Bhattacharya,
Pradeep Yammiyavar, "A behavioral approach
to modeling Indian children’s ability of
adopting to e-learning environment" 2012 4th
International Conference on Intelligent Human
Computer Interaction(IHCI)

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


421
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951107

A Review on Car Collision Avoidance System based on


Interpretation of Vehicle Event Data Recorder
Prof. Tejal Irkhede, 2Rohit Kashayp, 2Nikhil Kumar, 2Vivek Ranjan, 2Manish Kumar
1

1Assistant Professor, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,


Maharashtra, India
2BE Students, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,

Maharashtra, India

ABSTRACT

According to the World Health Organization (WHO), in excess of a million people on the planet pass on
consistently by virtue of vehicle accidents. Notwithstanding care cause, this issue is up 'til now growing due to
rider's poor practices, for instance, tank driving, speed driving, riding without sufficient rest, riding with no top
protection, etc. The Event Data Recorder (EDR) or vehicle 'black box' is a standout amongst the most valuable
instruments accessible to crash Reconstructionist. For vehicles outfitted with this innovation, and when the
recorders fill in as planned, the EDR gives crash agents significant data that in numerous examples could not be
gathered. In this paper, we present the issues related to accident detections and some existing systems for
prevention of accidents.
Keywords : Car Accidents, Road Safety, Public Health, Black Box, Event Data Recorder

I. INTRODUCTION intensifying circumstance. In the event that the


pattern proceeds with, the absolute number of street
Fatalities and wounds coming about because of traffic passings in India would increment by 100%
streetcar crashes are a noteworthy and developing somewhere in the range of 2013 and 2027 [1].
general medical issue in India. Consistently about Without expanded endeavors and new activities, the
2,650 individuals get executed and 9,000 get harmed all-out number of street traffic passings in India is
because of car crashes. In 2013, the most recent year probably going to cross the sign of 250,000 by 2025.
for which data is accessible, 137,423 individuals The primary point of this investigation is to examine
passed on and 469,900 individuals got harmed the streetcar crashes in India at the national, state,
because of street accidents in India [1]. Car crashes and metropolitan city level. The center is
have now earned India a questionable qualification; distinguished the significant street wellbeing issues
with about 140,000 passings every year, the nation and examine countermeasures that would possibly
has overwhelmed China to top the world in street address the particular street security issues. The
fatalities. India is the main nation on the planet essential wellspring of data for the examination is
which faces in excess of 15 fatalities and 53 wounds Accidental Deaths and Suicides in India, 1970 to 2013
each hour as a result of street crashes. While in many distributed by the National Crime Records Bureau,
created and creating nations including China, the Ministry of Home Affairs, Government of India, New
circumstance is commonly improving, India faces an Delhi.

CSEIT1951107 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 422-428 ]
422
Prof. Tejal Irkhede et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 422-428

The investigation demonstrates that amid the most The mishap examination includes the investigation of
recent ten years, street unplanned fatalities in India accident scenes and the documentation of all
have expanded at the rate of 5% every year while the fundamental and accessible data of every segment
number of inhabitants in the nation has expanded (for example a human, vehicle, and street condition).
just at the rate of 1.4% every year. Because of this, Mishap reproduction is characterized by Baker and
casualty chance, street inadvertent passings per Fricke as[11]"… the endeavors to decide from
100,000 individuals, has expanded from 7.9 in 2003 to whatever data is accessible, how the mishap
11.2 in 2013. Casualty hazard in India isn't just happened". Mishap reproduction approach works in
fourfold than that in a portion of the created nations, reverse from the proof of the accident examination
for example, the United Kingdom and Sweden yet in and the remaining parts of the collide with
addition as yet expanding quickly. It is likewise investigating the situation of previously (pre-crash),
discovered that the conveyance of street coincidental amid (crash) and after the accident (post-crash). The
passings and wounds shifts as per age, sexual successive investigation of final products to the
orientation, month and time. Among individuals of underlying state of the events can build up "how" and
all age gatherings, individuals of monetarily dynamic "why" a specific sort of accident happens. Arithmetic
age gathering of 30-59 years is the most powerless. and Newtonian material science are connected in this
Nonetheless, on the off chance that we analyze sexual investigation. It tends to be expressed that crash
orientation insightful fatalities and accidents, we recreation returns to explore the contributory
found that the guys represented 85.2% all things elements and additionally causes behind the accident
considered and 82.1% of all wounds in 2013. event dependent on major and minor physical pieces
of information abandoned at the accident scene.
Besides, street accidents are generally higher in
May-June and December-January, which The strategies of accident reproduction, direction,
demonstrates that extraordinary climate, impacts the and harm based investigation by utilizing material
event of street accidents. Accidents remain generally science streamline the assurance of numerous vital
steady and high amid 9 AM - 9 PM and variable yet parameters of accident events. In addition, to acquire
low amid mid-night and early hours of the day. In a dependable end, itemized data incorporating the
any case, this does not infer that daytime driving is framework segments should be completely explored.
more hazardous than evening time driving. The The data important for reproduction begins with the
investigation additionally attempts to discover the accident scene [12]. The responses to the inquiries of
reason shrewd conveyance of street accidents. 'why', 'what', 'when' and 'how' should lead the
recreation procedure to develop the genuine situation
There are a few variables in charge of accidents yet of the pre-crash, crash, and post-crash. Capturing of
the drivers' blame is observed to be the most essential imperative pieces of information and recording of the
one; drivers' blame represented 78% of all-out accident scene assumes an essential job in the
accidents in 2013. The recognizable proof of elements reproduction. Damage data from tenant restorative
influencing street crashes got from the accident reports can be checked with the direction of the
examination and remaking has not been directed by inhabitants found inside the included vehicles at the
and by in the Asian nations. The objective of this scene. In this way, a "receptive outlook" insightful
examination was to start this street wellbeing practice frame of mind is extremely urgent to scan for all the
in Thailand by tending to the auspicious requirement point-by-point data from the scene.
for a top to bottom investigation of street accidents.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


423
Prof. Tejal Irkhede et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 422-428

II. LITERATURE REVIEW occurrence recognition calculation dictated by


nonparametric relapse in [3].
Presently a-days, there is intense interest for cars,
as a result of this traffic control ends up boisterous The structure in [4] spare the message just as report
and it prompts street accidents. If there should arise with convergence utilizing metadata library. An
an occurrence of mishap, long reaction time to go to incident location strategy in street [5], make
the injured individual may prompts increment utilization of CCTV which regularly observe
number of death. In writing a few paper examine development of auto follow. Anyway this system
mishap recognition and vehicle following. In [2], an screens the primary focused on guests move that are
inventive remote disclosure using MEMS fitted with a great deal of circle openings inside it is
accelerometer and GPS following system is utilized execution. Because of the escape clauses inside going
for identification of accident.The structure can before is viable a couple of shiny new computerized
recognize kind of setback (immediate and nonlinear episode identification just as credit detailing
fall) from accelerometer sign using limit figuring, approaches wound up advanced. A large portion of
remove in the wake of crushing of GPS ground these gadgets is generally significantly centered
speedand cruiser. After incident is recognized, short proposed for some wheeler vehicles. This particular
ready data (alert back rub and position of setback) is computerized recognition just as credit announcing
sent by methods for GSM framework. The system is approach began working with GPS UNIT just as GSM
attempted in genuine 850 applications using bicycles. innovative expertise. Anyway these sort of wound up
The test results exhibit that it can recognize direct specifically built expected for cars. The real
fall, non-straight fall and ostensible ride with no mechanized episode location approach which
bogus alarm, remote discovery utilizing MEMS frequently utilizes air case sensor just as
accelerometer and GPS following for unplanned accelerometer to help analyze a mishap wound up set
checking of vehicle. The confinement of structure forward by M. Chuan-Zhietal [6]. Later gadgets that
proposed in [2] is mishap recognition for just bikes, have been advanced from this begun working with
and detailing just utilizing GPS with no black box. GPS unit innovation expected for occurrence
recognition alongside planned for deciding the
The paper [3] titled "The 3-arrange AcuTrac, setting in regards to episode.
Motorcycle Tracking System",Elite security supplies
outline workis constrained to bike on the grounds The framework arranged by Jerath just as Jung Lee
that a bikes setbacks could have a few cases which utilizes GPS unit innovation to help watch out for
can't absolutely use for four wheeler mishap speed with the auto just as microcontroller to help
discovery. The constraints of the above structure inspect speed with the auto planned for sequential
ismaking utilization of capacity of the GPS insignificant seconds [7][8]. Speedy braking will
framework just as a region sensor to influence an maybe result in a bogus helpful while speed
extremely decent spread technique to identify enormous contrast could be the basically angle
incidents. Various gadgets are accessible for viewed as expected for identifying episode. The
occurrence recognition by researchers. The real greater part of these strategies wound up made
episode location strategies wound up by deciding explicitly expected for a few wheelers [9].
realtime focused on vehicles and examination to help
expectation focused on vehicle. This particular The greater part of these gadgets isn't utilized in a
discovery approach is known while Traffic- few wheelers while they is likely not effective. In

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


424
Prof. Tejal Irkhede et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 422-428

this manner gadgets wound up intended for both the were included to ensure airbags deployed when they
cars.This is a stage for crisis salvage which will work were supposed to.
ideally so as to lessen the brilliant time of entry of
rescuers if there should arise an occurrence of street Over the years, as electronics got cheaper, smaller
accidents, when each microsecond checks. Our paper and smarter, event data recorders became capable of
plans to introduce an innovation consequently doing more than simply monitoring airbags.
identifying the mishap and an equipment GPS beacon Automakers realized the devices could be used to
dependent on GSM/GPS innovation illuminating at provide information about the seriousness of an
the event of mishap with adequate subtleties like accident, and if a car was being operated properly
accurate area and time at which mishap occurred and when a crash occurred. Based on a separate NHTSA
alongside that the voice was recorded for two or regulation passed in 2012, if a vehicle today does
three minutes when the mishap occurs.This paper have an event data recorder, it must track 15 specific
build up a correspondence between the control data points, including speed, steering, braking,
station and the unit introduced in vehicles. Vehicles acceleration, seatbelt use, and, in the event of a crash,
will have GPS/GSM empowered following modules force of impact and whether airbags deployed.
and will be followed progressively utilizing cell
systems. The product installed in the microcontroller Depending on the automaker and car model, an event
will control the different activities of the gadget by data recorder may capture many more functions,
observing waveform from the vibration sensor. In the though car companies are not required to disclose
event of mishap the gadget will send an alarm exactly what those are. The language many use to
message alongside area data from GPS module to explain black boxes in owner's manuals also is
control station utilizing GSM arrange. It is a purposely general to cover technology updates and to
thorough and viable answer for the poor salvage save space.
reaction in the event of mishap. The mishap detailing
can naturally discover an auto collision; scan for the Put everything the devices do in an owner's guide
spot and after that send the fundamental data to the and "instead of one paragraph, you'd have potentially
salvage office covering topographical directions and another 20 or 30 pages. That really wouldn't be
the time and conditions in which a car crash realistic," says Richard Ruth, a black box equipment
occurred. At the server end, a control capacity will trainer, expert witness and consultant who worked at
remove important data and store it in a database, to Ford Motor Co. for 33 years, including a stint
which mishap data from models will be surveyed evaluating event data recorders and other safety
progressively. Our framework joins propelled equipment. "It's not going to change whether or not
equipment plan and refined control innovation into a you're going to buy the car."
conservative, dependable bundle.
Most event data recorders are programmed to record
III. DECODING CAR'S BLACK BOX data in a continuous loop, writing over information
again and again until a vehicle is in a front-end
A. How Black Boxes Work collision or other crash. When an accident occurs,
Event data recorders are not actually black boxes but the device automatically saves up to 5 seconds of data
tiny microcomputer chip sets. In most vehicles, they from immediately before, during and after an
are part of the airbag control module, and originally incident.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


425
Prof. Tejal Irkhede et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 422-428

Today, practically every major automaker selling cars police and other law enforcement agencies that are
in the United States builds event data recorders into investigating an accident, insurance companies
new vehicles. The exceptions are Volkswagen (which looking into a claim, lawyers representing parties in
auto industry watchers say is preparing for the car-crash lawsuits and accident reconstruction
NHTSA regulation to kick in), Ferrari and Maserati. consultants working for any of the above.
Traffic accident analysis consultant Harris Technical
Services maintains a list of car makes and models In states with no black box laws on the books, "state
from 1994-2014 with event data recorders. troopers could get the data without a subpoena if
there was a fatality," says Tom Kowalick, a self-taught
The NHTSA rule, which the agency has been black box expert who chairs an event data recorder
working on for years, was supposed to take effect standards working group that's part of the Institute of
September 1, 2014. However, auto industry insiders Electrical and Electronics Engineers. Kowalick also
say the agency is still reviewing more than 1,000 wrote some of the black box information on the
comments it received about the proposed regulation, NHTSA Web site. "If they want to grab it, there's
making that implementation date unrealistic. A nobody saying they can't."
NHTSA representative declined to comment on the
delay. To rectify that situation, 15 states have passed EDR
B. Getting Black Box Data regulation over the past decade. Under the theory
Black box data is difficult and expensive to get to, and that car owners have privacy rights, many of the state
interpreting it takes special training. Extracting the laws require automakers to notify new-car buyers
data after an accident involves using a data-retrieval that vehicles contain black boxes, such as in the
tool kit that consists of hardware, software and a owner's manual. State laws also spell out the
cable that plugs into a car's on board diagnostics port. conditions under which police or other parties can
That is the same port mechanics use to identify obtain EDR information without an owner's consent,
engine problems and insurance companies tap as the such as with a court order; for dispatching emergency
basis for use-based insurance policies. Crash data personnel; diagnosing, servicing or repairing the
retrieval tool kits are not cheap, running $2,000- vehicle; or probable cause in an accident. The
$10,000 and up, not including training costs. National Council of State Legislatures maintains an
updated list of state EDR laws.
It follows that since drivers own their cars or trucks,
they own data the vehicles generate, including black Black boxes have become a battleground in states
box data. But because it's so difficult and costly to such as California, where earlier this year, insurance
extract, it's virtually impossible for average car companies and automakers lined up on opposite sides
owners to do it on their own — assuming that they of a black box data protection bill that would have
even want to. required automakers to let car owners block or opt
out of recording vehicle information. The bill didn't
Who else can access the information is a point of make it out of the state Senate Transportation
contention. Automakers would like the right to Committee after heavyweights including the Alliance
access the information for numerous reasons of Automobile Manufacturers opposed it.
including safety, to make sure systems work the way
they should and to check for defects. Other parties Earlier in 2014, two U.S. senators introduced a
that want a black box's car crash data can include bipartisan bill that would provide some of the same

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


426
Prof. Tejal Irkhede et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 422-428

protections on a national level. The Driver Privacy turned off. No one can hack into something without
Act explicitly states that a black box's data can't be Internet access." Even if someone could break in,
retrieved by anyone other than vehicle owners specifically in an effort to get data from an event data
without their consent and protects any personally recorder, the devices have access codes that need
identifiable information. By April 2014, the bill had authorized commands to work, he says.
collected 23 co-sponsors and been approved by the
Senate Commerce Committee. As of July 2014, In addition, car companies use threat modeling and
however, no further action had been taken. simulated attacks to test security and to help design
controls that protect data, says Wade Newton,
C. Black Boxes, Privacy and Security communications director for the Alliance of
Meanwhile, electronic privacy advocates worry about Automobile Manufacturers, an auto industry trade
a related car-data security issue: that a car's diagnostic group that represents 12 large car companies. "From
port, through which black box data streams, isn't bumper to bumper, automakers use proven security
secure enough to withstand hacking, and therefore techniques to help prevent unauthorized access to
poses a danger. Security experts and "white-hat" software," he says.
hackers already are testing how to break into the
ports to show how vulnerable they are. They are IV. CONCLUSION
publicly sharing the results, as in a video called "How
to Hack a Car." The analysis shows that the distribution of road
accidental deaths and injuries in India varies
Kowalick, a longtime black box data privacy advocate, according to age, gender, month and time. It is found
started a company to sell a diagnostic port lock that that the economically active age group is the most
he invented. The $30 AutoCyb lock, which he vulnerable population group. In general, males face
markets on his company Web site, is inserted into the higher fatality and accident risk than their female
diagnostics port to turn off access and prevent counterparts. Moreover, road accidents are relatively
unwelcome parties from getting to the data or higher in May-June and December-January which
interfering with car systems. "Every car in America shows that extreme weather influences the
can be hacked," he says. "The diagnostic link occurrence of road accidents. Accidents are relatively
connector is unsecure. All you have to do is set up constant and high during 9 AM to 9 PM and variable
access to the vehicle and have the right tool." but low during mid-night and early hours of the day.
There are several factors responsible for accidents but
However, Ruth, the EDR consultant and former Ford drivers’ fault is the most important factor; drivers’
executive, maintains a physical lock couldn't stop a fault accounted for 78% of total accidents, 76.5% of
black box expert or mechanic from bypassing the total injuries and 73.7% of total fatalities in 2013.
diagnostics port and obtaining the data another way. Automatic vehicle accident detection and reporting
He also dismisses the notion that hackers would be System using black box is is the need of the hour. The
interested in information. framework should be placed in moving vehicle to
detect accident and report to In Case of Emergency
"What would the incentive be?" he asks. "I think it's (ICE). The black box will records the voice of victim
an over-reaction. For all practical purposes, the after the accident occurs which will be used for
owner controls physical access. There's no Internet further investigation.
port on the car that's live, especially when the car's

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


427
Prof. Tejal Irkhede et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 422-428

Lee Conference on Communications (APCC),


V. REFERENCES 2012 18th Asia-Pacific pp.no: 978-1-4673-4726-6.
[10] Jog, S.R., Sutaone, M.S,Badawe, V.V.,
[1] Accidental Deaths & Suicides in India, 1970 to "Ruggedisation methodologies for GPS based
2013 published by the National Crime Records Vehicle Tracking System," (ICECT), 2011 3rd
Bureau, Ministry of Home Affairs, Government International Conference on , vol.2, no.,
of India, New Delhi. pp.214,218, 8-10 April 2011.
[2] Wireless Black Box using MEMS Accelerometer [11] Baker, J., and Fricke, L., Process of Traffic
and GPS Tracking for Accidental Monitoring of Accident Reconstruction. Traffic Accident
Vehicles, 978-1-4577-2177- 9/12$25 (c) 2012 Reconstruction (L. Fricke, ed.), Evanston, IL,
IEEE. Northwestern University Traffic Institute. (2010).
[3] “The 3-stage acuTrac Motorcycle Tracking [12] Van Kirk, D. J. Vehicular Accident
System”,Elite security supplies, Investigation and Reconstruction. CRC press.
http://www.gpsfast.com. USA. pp.15-19. (2001).
[4] "Traffic-incident detection-algorithm basedon
nonparametric regression”,S. M. Tang and H. J. Cite this article as :
Gao, IEEE Transactions on Intelligent
Transportation Systems, vol. 6, 2005, pp. 38-42. Prof. Tejal Irkhede, Rohit Kashayp, Nikhil Kumar, Vivek

[5] Yong-Kul Ki ,Jin-Woo Kim and Doo-Kwon Baik," Ranjan, Manish Kumar, "A Review on Car Collision
Avoidance System based on Interpretation of Vehicle
A Traffic Accident Detection Model using
Event Data Recorder ", International Journal of Scientific
Metadata Registry ",Conference on Software
Research in Computer Science, Engineering and
Engineering Research, Management and
Information Technology (IJSRCSEIT), ISSN : 2456-3307,
Applications, 2006 ppno: 0-7695-2656-X. Volume 5 Issue 1, pp. 422-428, January-February 2019.
[6] “Providing Accident Detection in Vehicular Journal URL : http://ijsrcseit.com/CSEIT1951107
Networks Through OBD-II Devices and Android-
based Smartphones”Juan Carlos Cano,Jorge
Zaldivar, Carlos T. Calafate, Pietro Manzoni in
5th IEEE Workshop On User Mobility and
Vehicular Networks 2011 pp no : 978-1-61284-
928-7.
[7] "Method of Freeway Incident Detection Using
wireless Positioning,"H. Ru-fu,L. Chuan-zhi, Y.E.
Hong-wu, in Proceedings of the IEEE
International Conference on Automation and
Logistics, 2008, pp. 2801 -2804.
[8] "GPS-free terrain-based vehicle tracking on road
networks," Jerath, K.; Brennan, S.N., American
Control Conference (ACC), 2012, pp.307, 311,
27-29 June 2012
[9] "An accident detection system on highway
through CCTV with calogero-mosersystem,Jung

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


428
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951108
A Review on Online Medical Drug Tracking
Swati Hedau1, Ankita Shegaonkar1, Amruta Aknurwar1, Devyani Vaidya1, Mehul Sahu1, Prof. Jogi John2
1 BE Students, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,
Maharashtra, India
2Assistant Professor, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,
Maharashtra, India

ABSTRACT

Recently, there is an enormous increment in online looking for changed wares like fabric, adornments and the
closeout of medicines have been expanded by means of the online/web. The online following of medicine is one
of the innovation headways that is going to make a colossal interest in the up and coming days. Despite the fact
that it is helpful to get them, there is a high danger of self-medicine and abuse of drugs, particularly that go
under calendar H and X. In this article, we have examined in insight concerning online pharmacy, its focal
points, and drawbacks contrast among online and neighborhood pharmacy, and certain measures to keep away
from the abuse of e-pharmacy.
Keywords : Drug Tracking, Online Pharmacies, Laws of Pharmacy, Prescription

I. INTRODUCTION history and afterward compose a remedy. This kind


of online pharmacy is all the more speaking to
Web or online drug stores move pharmaceuticals, shoppers
including solution and non-physician, recommended
meds. The online closeout of drugs began in the late since it offers both doctor administrations and drug
1990s and has broadened so extensively that the US specialist administrations to shoppers [3]. At last, the
Food and Drug Administration (FDA) has executed a third sort of online pharmacy is delegated a
whole segment on its site devoted to "BeSafeRx: "maverick" pharmacy or online drug shops which
Know Your Online Pharmacy" [1]. enables shoppers to buy remedy drugs without a
doctor's counsel and a substantial solution 3. Such
The online pharmacy industry can be arranged into pharmacy is viewed as unlawful by most law
three noteworthy sorts. The first and most genuine implementation organizations.
sort is the conventional online pharmacy. These
pharmacies just apportion medicine to a purchaser The World Health Organization faces this issue
who has first gotten a solution from a specialist and with regards to fake medicines, which are
after that submits it to the online pharmacy [2]. The characterized as 'a worldwide general wellbeing
second kind of online pharmacy is commonly known emergency', expressing that 'medicines bought over
as online meeting drug stores, remote counseling the Internet from destinations that cover their
drug stores, or recommending drug stores where physical location are fake in over half of cases [4].
doctors survey a patient's self-announced therapeutic

CSEIT1951108 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 429-432 ]
429
Swati Hedau et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 429-432

There are numerous obvious dangers of purchasing (1) Legitimate Internet pharmacy sites giving
medicine over the Internet. These incorporate simple fantastic pharmacy administrations as indicated by
access to unlawful or unapproved drugs, medicine confirmation principles, and
offered without a solution, and drugs promoted with
phony wellbeing claims [5]. Additionally, it reroutes (2) Illegitimate online drug stores that are not
the conventional protections i.e., noteworthy confirmed and may not consent to national or global
cooperation with medicinal services experts (doctor expert benchmarks and directions.
as well as drug specialist) and subsequently putting
shoppers in danger [6]. Additionally, the greatest II. LITERATURE REVIEW
hazard that online drug stores present is to the
patient's wellbeing through fake and unacceptable A. Laws of Pharmacy
prescriptions [7]. According to Supreme Court statement, except in
emergency cases, doctor cannot consult through
Despite various potential dangers of buying meds phone. As far as India is concerned, there are no
online, there are a few evident advantages. Getting dedicated e-health and telemedicine laws in India.
medicines through the Internet is basic and There are no well-defined dedicated laws for online
advantageous as going forward and backward from a pharmacies. The laws applicable are:
traditional pharmacy had turned out to be extremely • Drug and Cosmetics Act 1940,
troublesome. Drugs can be requested whenever of • Drugs and Cosmetic Rules 1945,
day or on the other hand night and medicine could • Pharmacy Act 1948,
be dispatched by an online pharmacy medium-term • Indian Medical Act 1956 and
to his front entryway, thus sparing time and stress. • Laws related to e-commerce are defined under
What's more, utilizing online drug stores manages the Information Technology Act, 2000.
customers with more protection than the customary
specialist or pharmacy visit. Online drug stores are
especially engaging for those with ailments they wish B. Role Of Prescription
to keep private, as they don't require any immediate Online pharmacies will promote drug abuse, drug
human cooperation. misuse, self-medication, etc. Self-medication is a
rampant practice in India, and online sale of drugs
In past, the best-specialized technique for doctors would only encourage it.
in choices identifying with the drug for drug • Any medication taken without the supervision
specialists to disseminate medicines was written by of doctors may be dangerous and even
hand remedy. Presently it is being supplanted by potentially life-threatening.
electronic solution organize usually known as e- • A prescription issued by a doctor cannot be re-
pharmacy/online pharmacy. Online drug stores are used randomly. There is a danger that
organizations that move pharmaceutical scheduled drugs can be re-ordered and misused
arrangements that incorporate solution just drugs, on by the consumer.
the Internet by means of online requesting and mail • The D&C Act 1940, and the D&C Rules 1945
conveyance. Online drug stores can be arranged has guidelines on the sale of Schedule H and
fundamentally as: Schedule X drugs. These can be sold only on
prescription, and there are specific rules,
including for labeling and bar coding.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


430
Swati Hedau et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 429-432

Schedule X drugs include narcotics and pain relievers. Patients can take steps to ensure that
psychotropic substances. Chances of drug abuse and they use prescription medications appropriately by:
addiction are higher with the Schedule X category. • Never change a dosing regimen without first
Schedule X drugs require meticulous storage and discussing it with the doctor
dispensing records. The prescription has to be in • Never using another person’s prescription,
duplicate and one copy of which is to be retained by and never giving his or her prescription
the licensed pharmacist for 2 years. medications to others.
• Storing prescription stimulants, sedatives, and
C. Comparison Of Local Pharmacy And Online opioids safely patients should properly
Pharmacy discard unused or expired medications
Price: The prices for medications from an online Schedule H and schedule X drugs cannot be sold
pharmacy are less because there is no particular without prescription because it is easy for local
production unit, there is no physical store for the pharmacy stores to verify the authencity of a
medications storage and dispensing, and the company prescription, as it contains details like registration
has a large network of retailers to draw from, or a number of medical practitioner and name of the
combination of factors. hospital and there is no such checking by online
Convenience: Prescription can be submitted at any stores.
time, there is no need to rush to the pharmacy during • According to the report of high-level
business hours and do not have to wait in any long government the physicians should digitally
lines. However, there are also disadvantages. If it is sign committee since doctor’s prescription is
not the same website every time, one may not be mandatory for purchasing drugs, online
aware of any potential drug interactions, it can be pharmacies will need to mandatorily display
difficult to find someone to answer any specific the portal link on their homepage for
questions about medications. authencity verification by patients or
Local pharmacies are small independently owned customers and the prescriptions.
business rather than a pharmacy within one of the • In certain countries like United States,
large chains and they have staff during regular England, Sweden, Denmark and Finland have
business hours, and many also have an online accurate, regular, and systematic plans to use
counterpart as well. The large chains do use locals for electronic prescription system, and health
a large portion of their workforce. ministry of these countries was responsible
Pharmacist is responsible for dispensing the right for coordinating and leading the electronic
medicines and even counseling a patient about side- health.
effects and dosage. In the online space, where the • It is suggested to use experiences and
medicines are delivered at the patient’s home, there is programs of the leading countries to design
no possibility of an interaction between a patient and and develop the electronic prescription
a pharmacist. systems.
D. Measures To Be Taken To Avoid The Misuse Of
Online Pharmacy III. CONCLUSION
Before preventing the misuse of online pharmacy,
patients should stop the nonmedical use of Despite the fact that online pharmacy is more
prescription drugs like stimulants, sedatives or opiod subject than nearby pharmacy particularly in remote
regions, it ought to be planned to build up standards

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


431
Swati Hedau et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 429-432

and direction like transfer of filtered solutions. <http://www.fda.gov/NewsEvents/Newsroom/P


Patients ought to know on utilization of doctor ressAnnouncements/2004/ucm108232.htm>.
prescribed prescriptions. It is great to embrace the
plans and projects of driving nations to structure and Cite this article as :
build up the e-pharmacy framework.
Swati Hedau, Ankita Shegaonkar, Amruta Aknurwar,
IV. REFERENCES Devyani Vaidya, Mehul Sahu, Prof. Jogi John, "A
Review on Online Medical Drug Tracking",
[1] FDA. Food and Drugs Administration. Be- International Journal of Scientific Research in
SafeRx: Know Your Online Pharmacy. Computer Science, Engineering and Information
Available at Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
<http://www.fda.gov/Drugs/ResourcesForYou/ 5 Issue 1, pp. 429-432, January-February 2019.
Consumers/BuyingUsingMedicineSafely/Buyin Journal URL : http://ijsrcseit.com/CSEIT1951108
gMedicinesOvertheInternet/BeSafeRx-
KnowYourOnlinePharmacy/default.htm>.
[2] Fung, C.H., H.E. Woo & S.M. Asch (2014)
Mayo Clin Proc. 79: 188-94.
[3] Castronova, J.R. (2006) J Leg Med. 27: 207-24.
[4] The World Health Organization (2006)
Counterfeit medicines. Fact sheet revised 14
November 2006. Available at
<ttp://www.who.int/search/
semantic/en/#search=medicines%20purchased
%20over%20the%20Internet&sort=score%20de
sc&fq=%7B!noshow%3Dtrue%7Dlanguages%3
Aen>.
[5] Johnston, L.D., P.M. O’Malley, J.G. Bachman &
J.E. Schulenberg (2004) Monitoring the Future
national results on adolescent drug use:
Overview of key findings, 2003. Bethesda, MD:
National Institute on Drug Abuse. Available at
<http://www.monitoringthefuture.org/pubs/mo
nographs/overview2003.pdf>.
[6] Henney, J.E., J.E. Shuren, S.L. Nightingale &
T.J. McGinnis (1999) Ann. Intern. Med. 131:
861-2.
[7] US Food and Drug Administration (2004)
Recent FDA/U.S. Customs Import Blitz Exams
Continue to Reveal Potentially Dangerous
Illegally Imported Drug Shipments. January 27,
2004. Available at

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


432
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951110
A Survey on Secure User Data Transaction using Attribute-Based
Encryption Scheme over Cloud Data
Navin Sethiya1, Hrishikesh Patel1, Akshay Harshe1, Alekh Gaigole1, Harshvardhan Donadkar1, Prof. Priya
Karemore2
1UG Students, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,
Maharashtra, India.
2Assistant Professor, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,
Maharashtra, India.

ABSTRACT

Cloud computing is an innovation which utilizes the Internet to store, oversee, and register the information, as
opposed to a nearby server or a PC. Clients can access the cloud condition wherever they need on
compensation for each utilization way. Transferring of delicate information on un-believed servers in the cloud
is a testing assignment. Established encryption procedures are utilized when the re-appropriating of
information in the cloud to give access control and secrecy. Access control instruments are not doable in the
cloud as a result of the absence of adaptability, versatility, and fine-grained access control. As an option, So,
Attribute-Based Encryption (ABE) procedures are acquainted in the cloud with increment the security and
adaptability. Attribute-based encryption is an open key cryptographic strategy in which the private key and
figure content are based on the attributes. ABE involves IBE, KP-ABE, and CP-ABE. In Identity-Based
Encryption (IBE) the mystery key and the attributes are based on the character of the client. In Ciphertext-
Policy Attribute-Based Encryption (CP-ABE) a mystery key is gotten from the arrangement of attributes and
the access strategy is defined over the universe of attributes inside the framework. In Key Policy – Attribute-
Based Encryption (KP – ABE) the figure content is related to the arrangement of attributes and the mystery key
is based on the access structure. This paper is a study of all ABE strategies and specialty an evaluation table
about ABE keys job in cloud condition.
Keywords : Attribute-Based Encryption, Attributes Revocation, Fine-Grained Access Control, Keywords Search,
Mobile Cloud Storage.

I. INTRODUCTION different basic attributes like On-request access,


pervasive access, multitenancy, versatility, estimated
Cloud computing is a model for empowering utilization, strength. In On-request access, the client
pervasive, helpful, on-request organize access to a can utilize the computing assets in pay per use way.
common pool of configurable computing assets (Eg: Omnipresent access speaks to the capacity to access a
Network, Servers, Storage, applications and wide scope of gadgets, transport conventions,
administrations) that can be quickly provisioned and interfaces and different security innovations in cloud
discharged with negligible administration exertion or situations. Multitenancy empowers distinctive buyers
specialist organization interaction[1]. It comprises of to share the different cloud storage administrations.

CSEIT1951110 | Received : 10 Feb 2019 | Accepted : 21 Feb 2019 | January-February -2019 [ 5 (1) : 433-437 ]
433
Navin Sethiya et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 433-437

Versatility is the capacity of a cloud to direct scale IT 2. Integrity: It will keep unapproved clients from
assets as required by the cloud purchaser or cloud changing the information which is put away or
supplier. Estimated usage addresses the limitations of transmitted in the cloud.
a cloud stage to screen the utilization of its IT assets,
in a general sense by cloud buyers. Flexibility alludes 3. Availability: It enables the information to be
to excess IT assets inside a similar cloud or over accessible and usable amid an explicit timeframe.
various clouds. Three regular cloud conveyance
models are generally utilized and set up: 4. Scalability: Capability to deal with the clients
Infrastructure-as-a-Service (IaaS), Platform-as-a- included the system powerfully with no disturbance
Service (PaaS), Software-as-a-Service (SaaS). IaaS of administration.
conveyance demonstrates involved framework driven
IT assets that can be accessed and oversaw through 5. Client revocation: If the client leaves the system
cloud benefit based interfaces and instruments. PaaS the plan should disavow his access rights from the
relies upon the current condition that makes a lot of system straightforwardly. Something else, the client
apparatuses used to help the whole applications. SaaS can't utilize the information put away and the access
gives reusable cloud benefit generally accessible to is repudiated.
the scope of cloud buyers. Essentially, three sorts of
cloud arrangement models are the open cloud, 6. User revocation: Users can decode the encoded
private cloud, and third-party cloud. The open cloud information with their very own attributes since the
is claimed by an outsider cloud which is accessible by attributes are identified with polynomial arbitrary
everybody in the cloud condition. Private cloud is encryption work.
possessed by a solitary association where limits are
defined. The third party cloud is a mix of open and 7. Access Tree: Based on the attributes, the access tree
private cloud models. is produced. The access tree is included hubs. The leaf
hubs are spoken to as attributes and the middle of the
Cloud computing is a sort of distributive computing road hubs are spoken to utilizing doors likes AND,
that speaks to the usage models for remotely OR, and so on.
provisioning versatile and estimated assets [12]. At II. LITERATURE REVIEW
the point when the business information is moved to
the cloud, at that point the security issue will ascend, By expecting the believed cloud service provider
as information security ends up imparted to the cloud gives security additionally to the expansive measure
specialist organization. Fundamental wordings of delicate and significant information put away. ABE
utilized in the cloud will be: cloud supplier gives the calculations can be utilized for securing the
cloud-based IT assets. Cloud buyer utilizes the cloud- confidentiality of the put away information and
based IT assets. Cloud examiner assesses the security furthermore ABE gives the access control instrument
controls, protection effects, and execution. Central to information on the cloud. In cloud condition,
security terms important to cloud computing are: information confidentiality is imperative to shield
against insider assault, crash assault and refusal of
1. Confidentiality: It enables approved gatherings to administration assault. This area gives the current
access the information or assets that is it confines the attribute based encryption components in cloud
unapproved access to information in travel and condition.
storage.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


434
Navin Sethiya et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 433-437

ABE was presented by Sahai and Waters [2] in 2005. intended for one to numerous correspondences.
It is an open key based one to numerous encryption Decoding should be possible just when the attribute
that enables client to scramble and decode the set fulfills the client's access structure. It bolsters
information based on client attributes. The mystery access control conspire.
key and figure content are subject to the client
attributes. The decoding of ciphertext is conceivable Disadvantages:
just if the arrangement of attributes of client key Information proprietor can't choose who can decode
matches with the attributes of ciphertext. Decoding the encoded information. It can just pick graphic
should be possible just when the quantity of attributes for the information. It isn't appropriate for
coordinating keys is equivalent to the referenced broadcasting applications in light of the fact that the
limit level. ABE calculation comprise of four stages: information proprietor needs to confide in the key
Setup, key age, encryption and unscrambling. Crash backer.
obstruction is a pivotal highlights of ABE. A rival that
holds numerous keys can access the information if 3. Expressive Key Policy ABE (EKP-ABE)
the individual key matches. It is the all-inclusive variant of KP-ABE, the non-
monotonic access structures [5] are utilized. Non-
Drawbacks: monotonic access structure utilizes repudiated doors,
For encryption, the Data proprietor needs to utilize for example, NOT in the access structure. It is
each approved client's open key so it expands the increasingly adaptable to limit the unapproved use of
calculation overhead. This strategy is confined in information.
light of the fact that it utilizes access of monotonic
attributes to control client's access to the framework. Disadvantage:
The overhead is expanded on account of negative
1. Identity Based Encryption (IBE) attributes which isn't pertinent to the encoded
It was proposed by Shamir in 1984[3].In Identity information.
Based Encryption (IBE) the mystery key and the
attributes are based on the identity of the client. For 4. Ciphertext Policy ABE (CP-ABE)
instance, email id be the identity of the client. It is a turn around the form of KP-ABE [6]. In CP-
Believed third party will create the private key. ABE, the ciphertext is related with an access
structure and client's private key is based on set of
Drawback: attributes. A client can unscramble the ciphertext just
In the event that the believed third party gets traded whenever set of attributes related with clients private
off, there is no security for the framework. The client key fulfills the access policy related with the
protection can't be saved utilizing IBE. ciphertext. CP-ABE is more anchored even the
believed the third party is imperiled.
2. Key Policy Attribute Based Encryption (KP-ABE)
It was proposed by Goyal et al. [4]. Each client is Disadvantage:
assigned with some access tree over the arrangement
of attributes. Ciphertexts are based on the CP-ABE doesn't fulfilled undertaking needs of the
arrangement of attributes and the private key is based access control component. Decoding keys underpin
on the monotonic access structure that controls client attributes that are sorted out intelligently as a
which ciphertexts a client can decode. This is solitary set.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


435
Navin Sethiya et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 433-437

5. Hierarchical Attribute Based Encryption (HABE) The FHABE conspire is turned out to be secure under
The HABE was proposed by Wang et al. [7]. This the standard suspicion.
model comprises of a root ace that relates to the
confided in third party, different space ace in which 8. Ciphertext policy Weighted ABE (CPWABE)
top dimension area aces compares to various venture This plan assigns [10] load to the attributes based on
clients and various clients that relates to all work the inclination defined in the access control structure.
force in an undertaking. This plan client's Clients will have appointed the load to their
hierarchical age of keys. It utilizes randomized attributes. The information proprietor will encode
polynomial time calculation. It underpins anchored the information based on the attributes. The private
access control, adaptability and completely key is based on the weighted access structure and the
designation. It can share the ensured information for decryption [11] is conceivable just when the
clients in a cloud in an endeavor situation. It bolsters ciphertext and the weighted attribute matches with
intermediary reencryption conspire. the weighted access structure. It takes care of the key
Disadvantages: escrow issue by utilizing weighted attributes.
It is extremely hard to actualize continuously Drawback:
condition. A similar attribute will be utilized by Cost of computation is increased.
numerous area aces.
III. CONCLUSIONS
6. Multiple Authority ABE (MAABE)
This strategy utilizes numerous experts and the client This paper abridges different attribute-based
attributes are dispersed to different specialists. encryption systems for information security in cloud
Suppose [8] it comprises of N multiple attribute condition. Security will be the essential factor to be
experts and one focal specialist. The information can tended to. The storage cost of the figure writings,
be decoded by the client under attribute set an and time of encryption and unscrambling to be
unscrambling keys for an attribute set Au. It enables diminished. Based on the qualities like fine-grained
any free experts to disseminate private keys and to access control, adaptability and adaptability in cloud
deal with flawed or lost specialists. computing we can infer that FH-WABE performs
well superior to the next attribute encryption
Disadvantages: strategies.
Trouble in muti-expert plan is it expects specialist's to
keep up disjoint arrangements of attributes. IV. REFERENCES

7. File Hierarchy Attribute Based Encryption (FH- [1]. Muhammad Yasir Shabir, Asif Iqbal, Zahid
ABE) Mahmood and Ataullah Ghafoor, Analysis of
This strategy utilizes record progressive system based Classical Encryption Techniques in Cloud
encryption strategies. The documents are scrambled Computing, IEEE Tsinghua Science &
based on the layered access structure. Both figure Technology, Vol.21, February 2016.
content storage and time expenses of encryption are [2]. A.Sahai and B.Waters, Fuzzy identity based
saved [9]. encryption, in Proc. Advances in Cryptology –
Eurocrypt, 2005, pp.457-473.
Disadvantage: [3]. Adi Shamir. Identity-based cryptosystems and
signature schemes, in Proc. of CRYPTO 84 on

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


436
Navin Sethiya et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 433-437

Advances in cryptology, pages 47–53. Attribute based data sharing scheme revisited
SpringerVerlag New York, Inc., 1985. in cloud computing, IEEE Transactions on
[4]. V.Goyal, O.Pandey, A.Sahai and B.Waters, information forensics and security, Vol.11, No.8,
Attribute based encryption for fine grained August 2016.
access control of encrypted data, in proc. ACM [13]. G.K.Sandhia, G.K.Bhaskar, secure data sharing
conf. computer and communications, 2006. using attribute based encryption in cloud
[5]. J.Bethencourt, A.Sahai, and B.Waters, computing, Journal of Chemical &
Ciphertext policy attribute based encryption, in Pharmaceutical Sciences, Vol.9, December 2016.
proc. IEEE symposium security and privacy,
2007. Cite this article as :
[6]. K.Meena,S.Vinodhini, T.Pallavi & R.Vasugi,”
Surveillance Based Gcm Home Security System Navin Sethiya, Hrishikesh Patel, Akshay Harshe, Alekh
Gaigole, Harshvardhan Donadkar, Prof. Priya Karemore, "A
Using Object Motion Detection”,International
Survey on Secure User Data Transaction using Attribute -
Innovative Research Journal of Engineering Based Encryption Scheme over Cloud Data", International
and Technology,pp.44-47,2016. Journal of Scientific Research in Computer Science,
[7]. S.Rifki, Y.Park, and S.Moon, A fully secure Engineering and Information Technology (IJSRCSEIT),
ISSN : 2456-3307, Volume 5 Issue 1, pp. 433-437, January-
ciphertext policy attribute based encryption
February 2019.
with a tree-based access structure, Journal of Journal URL : http://ijsrcseit.com/CSEIT1951110
Information Science and Engineering, Vo.31,
pp.247-265,2015.
[8]. G.Wang, Q.Liu and J.Wu, Hierarchical
attribute based encryption for fine grained
access control in cloud storage services, in Proc.
ACM conf. computer and communication
security, 2010.
[9]. K.Yang, X.Jia, K.Ren and B.Zhang, Dac-Macs:
Effective data access control for multiauthority
cloud storage systems, in Proc. Of IEEE
Infocom, 2013.
[10]. Shulan Wang, Junwei Zhou, Joseph K.Liu,
Jianping Yu, Jianyong Chen and Weixin Xie,
An efficient file hierarchy attribute based
encryption scheme in cloud computing, IEEE
Transactions on information forensics and
security, Vol. 11, No.6, June 2016.
[11]. Ximeng Liu, Jiafeng Ma, Jinbo Xiong, Qi Li, Jun
Ma, Ciphertext – policy weighted attribute
based encryption for fine grained access control,
International conference on Intelligent
networkings and collaborative systems, 2013.
[12]. Shulan Wang, Kaitai Liang, Joseph K.Liu,
Jianyong Chen, Jianping Yu, Weixin Xie,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


437
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195164
Accessing Network Using Secure Android Applications
R. Gomathyjayam1, R. Thilakavathi2
1 Professor, Department of Computer Applications, Bon Secours College for Women, Thanjavur, Tamil Nadu,
India
2M.Sc., Department of Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India

ABSTRACT

Security plays a vital role in today’s mobile world. There are security issues like sniffing of data while accessing
information through open channel. Proper security measures can help to deal with the common security
threats faced by mobile phone users such as data protection, privacy, application and personal information
security. Cryptographic techniques play an important role in protecting communication links and data, since
access to data can be limited to those who hold the proper key. This paper discusses a method to securely access
information in a network by an android mobile application using AES cryptographic technique. The paper
describes a new key sharing algorithm, based on the symmetric key management, for faster and efficient
encryption of data that is suitable for use in a mobile device.
Keywords : Cognitive Radio Networks, Primary User, Secondary Users.

I. INTRODUCTION layer that directly deals with the transmission media


network support applications such as access to the
The telecommunications network allows computers World Wide Web, shared use of application and
to exchange data. The connections between nodes are storage servers, printers, and fax machines, and use of
established using either cable media or wireless email and instant messaging applications
media. Network computer devices that originate,
Route and terminate the data are called network II. LITERATURE SURVEY
nodes. Nodes can include hosts such as personal
computers, phones, servers as well as networking Reputation and Trust Management
hardware. Two such devices are said to be networked The reputation based frameworks are the nodes
together when one device is able to exchange maintain reputations of other nodes and evaluate
information with the other device, whether or not their trustworthiness are deployed to provide
they have a direct connection to each other. Scalable, Diverse and for countering different types
Computer networks differ in the transmission media of misbehavior resulting from malicious selfish nodes
used to carry their signals, the communications in the networks. The author (Jaydip et al.,2009) is
protocols to organize network traffic, the network’s proposed that Reputation systems are described along
size, topology and organizational intent. In most with their unique characteristics and working
cases, communications protocols are layered on (i.e. principles. Reputation and trust are very useful tools
work using) other more specific or more general that are used to facilitate decision making Reputation
communications protocols, except for the physical is the opinion of one entity about another it signifies

CSEIT195164 | Received : 08 Feb 2019 | Accepted : 21 Feb 2019 | January-February -2019 [ 5 (1) : 438-442 ]
438
R. Gomathyjayam, R. Thilakavathi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 438-442

the trustworthiness of an entity. Trust is the detection and energy based detection algorithms.
expectation of one entity about the action of another. Coherent sensor is an optimal linear detector for
Reputation and trust have been adapted to wireless known primary signals in presence of white Gaussian
communication Network it can solve many problems noise. However, detector implementation requires.
in the Networks. Demodulation of received signals for achieving the
optimal gains.
Combinationarial Optimization dentification
Algorithm (Coi) Agent-Based Trust Calculation
The COI (Combinationarial Optimization The author (yenumula B.Reddy et al., 2011) is
Identification Algorithm) to defend against such proposed agent based approach algorithm. The
attacks .Cooperative Spectrum sensing has been Cooperation in wireless sensor networks to detect the
shown good performance in improving the accuracy malicious node without any Infrastructure is a recent
of primary user detection .zhenquiri et al., 2013 trend in research. The current models need more
proposed a modified Design an attack model,called storage, Computation, security tools and
cooperative attack in which an attacker injects self- communication requirements. The fail wireless
consistent false data to multiple sensors sensor Network due to limitation of resources. The
simultaneously. A theorem that the center node of a proposed agent based approach eliminates the
cognitive radio network may face uncertainties under computations in the sensor nodes with appropriate
a cooperative attack, especially in the case when a trust factor. The proposed approach uses an agent
large portion of sensors are compromised. A modified based collaborative concept ensure the trust in the
COI algorithm to deal with cooperative attacks. Our successive node in the Path. The proposed agent
algorithm is a good scheme to complement IRIS for based framework uses reputation of neighboring node
cooperative attacks, and can be flexibly adjusted to as part of trust calculation in its successive node.
fulfill the detection delay requirement. Intensively
evaluate our algorithm through simulation, with the Sophisticated Detection Methods
results validating its performance. The original COI is The Spectrum scarcity is becoming a major issue for
an approach for identifying multiple instances of bad service providers interested in either deploying new
data in power system state estimation. The essential services or enhancing the capacity for existing
idea is to construct a partial decision tree using the applications. The author (Amir Ghasemi et al.,2007)is
branch-and-bound method to obtain a feasible proposed that Recent Measurements suggest that
solution with the minimum number of bad data. many portions of the licensed (primary) spectrum
remain unused for significant periods of time.
Clustered Based Algorithms Sensing-based access incurs a very low infrastructure
The spectrum sensing methods can be distributed in cost and is backward compatible with the legacy
three categories i.e. transmitter, receiver. (Dr. Talat primary systems.
Altaf et al.,2013) is propose cluster based algorithms
is based spectrum sensing and interference- III. EXISTING PROCESS
temperature based detection algorithms. From these,
transmitter detection based methods are a preferred The T-CSS (Traditional collaborative spectrum
way of sensing for presence/absence of spectral holes. sensing) protocol delays in data transmission and
These methods can be implemented through various unaware of secondary user selection for licensed
techniques including coherent detection, feature bandwidth and interference occur in unlicensed

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


439
R. Gomathyjayam, R. Thilakavathi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 438-442

secondary network, it lack in secondary power, not V. IMPLEMENTATION RESULTS


able to detect weak primary signals, in order to COGNITIVE RADIO NETWORK
protect primary receivers from interference. By Cognitive techniques have been used in wireless
proposing an energy efficient CSS protocol, namely networks to circumvent the limitations imposed by
energy efficient collaborative spectrum sensing EE- conventional WSNs. Cognitive radio (CR) is a
CSS protocol is used to transmit the data efficiently. candidate for the next generation of wireless
EE-CSS attempts to reduce the number of transmitted communications system. The cognitive technique is
reports from HSUs, based on the observation that the process of knowing through perception, planning,
HSUs agree on the spectrum usage more often than reasoning, acting, and continuously updating and
they disagree. CRN is to utilize the unused licensed upgrading with a history of learning. If cognitive
spectrum opportunistically. The SUs should protect radio can be integrated with wireless sensors. CR has
the accessing right of the PUs whenever necessary. the ability to know the unutilized spectrum in a
license and unlicensed spectrum band, and utilize the
IV. PROPOSED METHODOLOGY unused spectrum opportunistically. The incumbents
or primary users (PU) have the right to use the
The architecture consists of the following system spectrum anytime, whereas secondary users (SU) can
entities. utilize the spectrum only when the PU is not using it.
I.primary user: A user who has higher priority or
legacy rights on the usage of a specific part of the
spectrum. Primary user provides path or channel to
the secondary users
ii.secondary user: secondary user base
station(SUBS)transmit the data and reception is done
by honest secondary users(HSUs)secondary user

Figure 2. Cognitive radio Network

COLLABORATIVE SPECTRUM SENSING


TECHNIQUE

In this spectrum is used to detect spectrum band for


transformation and reception. The matched filter
detection technique requires a demodulation of the
PU's information signal, such as the modulation type
and order, pulse shaping, packet format, operating
Figure 1. Proposed System Architecture frequency, bandwidth, etc. CR Network sensing
iii. Honest secondary user: receive information from the PU's pilots, preambles,
The person who is receiving the data from the synchronization words or spreading codes etc in
secondary user base station is said to be honest figure 4.4. The advantage of the matched filter
secondary user. method is that it takes a short time and requires
fewer samples of the received signal. Sensing reports

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


440
R. Gomathyjayam, R. Thilakavathi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 438-442

provided by SUs for a given licensed band may differ


due to differences in channel fading gains, locations
of SUs and primary network transmitters, number of
signal energy quantization levels used at the sensing
SU, and sensing errors.

Figure 4. Cooperative Trust Management and avoid


malicious behavior

PRIORITY BASED SPECTRUM TRANSFORMATION

In Cognitive Radio network the users are classified


into Licensed Primary Users and Unlicensed
Secondary Users and there is no dedicated channel to
send data, sensors need to negotiate with the
neighbors and select a channel for data
communication in CR-WSNs in figure 4.6. This is a
Figure 3. Collaborative spectrum sensing technique very challenging issue, because there is no
cooperation between the PUs and SUs. PUs may
COOPERATIVE TRUST MANAGEMENT AND arrive on the channel any ime. If the PU claims the
AVOID MALICIOUS BEHAVIOR channel, the SUs have to leave the channel
immediately. CRN is implemented for short range
CR Network sensors may encounter incorrect wireless applications such as wireless sensor networks
judgments because radio-wave propagation through (WSNs) such wireless and Bluetooth, where the
the wireless channels has adverse factors, such as transmission distance is usually small (e.g., tens of
multi-path fading, shadowing, and building meters the steady-state average total number of
penetration. In addition, CR wireless sensors are sensing reports transmitted for each band And
hardware constraints and cannot sense multiple assume that the packets transmitted from the FC and
channels simultaneously. It has a malicious behavior SUs are of equal length in both EE-CSS and T-CSS.
to intermediate the signal spectrum. TRMSs record
the accuracy of previous sensing reports sent by SUs
and compute a trust value for each SU which is taken
as the trustworthiness of its future sensing reports.
And encounter the reports from SUs may be required
to mitigate against the effects of malicious behavior
of MSUs in figure 4.5. Therefore, CR wireless sensors
cooperate and share their sensing information with
each other to improve the sensing performance and Figure 5. Priority based spectrum Transformation
accuracy.
VI. CONCLUSION

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


441
R. Gomathyjayam, R. Thilakavathi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 438-442

Cognitive Radio (CR) is an adaptive, intelligent radio VIII. REFERENCES


and network technology that can automatically
detect available channels in a wireless spectrum and [1]. A. Ghasemi and E. Sousa, "Collaborative spectrum
change transmission parameters enabling more sensing for opportunistic access in fading
communications to run concurrently and also environments," in Proc. IEEE DySPAN, 2005, pp.
improve radio operating behavior. Cognitive radio 131-136.
uses a number of technologies including Adaptive [2]. A. Ghasemi and E. Sousa, "Opportunistic spectrum
Radio (where the communications system monitors access in fading channelsthrough collaborative
and modifies its own performance) and Software sensing," J. Commun., vol. 2, no. 2, pp. 71-82, Mar.
2007.
Defined Radio (SDR) where traditional hardware
[3]. S.Haykin, "Cognitive radio: Brain-empowered
components including mixers, modulators and
wireless communications," IEEE J. Sel. Areas
amplifies have been replaced with intelligent
Commun., vol. 23, no. 2, pp. 201-220, Feb. 2005.
software. A spectrum sensing scheme, was proposed
[4]. J. Mitola and G. Maguire, "Cognitive radio: Making
to improve the utilization efficiency of the radio
software radios more personal," IEEE Pers.
spectrum by increasing detection reliability and
Commun., vol. 6, no. 4, pp. 13-18, Aug. 1999.
decreasing sensing time. The proposed scheme
[5]. E. Noon and H. Li, "Defending against hit-and-run
presented spectrum sensing in effective manner So attackers in collaborative spectrum sensing of
the priority based and security-based spectrum cognitive radio networks: A point system," in Proc.
sensing is produced. This system also implemented in IEEE VTC Spring, 2010, pp. 1-5.
hardware successfully. [6]. G. Staple and K. Werbach, "The end of spectrum
scarcity," IEEE Spectr., vol. 41, no. 3, pp. 48-52,
VII. FUTURE ENHANCEMENT Mar. 2004.
[7]. H. Yu, Z. Shen, C. Miao, C. Leung, and D. Niyato,
Wireless technology is proliferating rapidly, and the "A survey of trust and reputation management
vision of pervasive wireless computing and systems in wireless communications," Proc. IEEE,
communications offers the promise of many societal vol. 98, no. 10, pp. 1755-1772, Oct. 2010.
and individual benefits. While consumer devices [8]. T. Yucek and H. Arslan, "A survey of spectrum
such as cell phones, PDAs and laptops receive a lot of sensing algorithms for cognitive radio applications,"
attention, the impact of wireless technology is much IEEE Commun. Surveys Tuts., vol. 11, no. 1, pp.
broader, e.g., through sensor networks for safety 116-130, 2009.
[9]. W. Wang, H. Li, Y. Sun, and Z. Han, "Catchit:
applications and home automation, smart grid control,
Detect malicious nodes in collaborative spectrum
medical wearable and embedded wireless devices,
sensing," in Proc. IEEE GLOBECOM, 2009, pp. 1-6.
and entertainment systems. This explosion of wireless
Cite this article as : R. Gomathyjayam, R. Thilakavathi, "Accessing
applications creates an ever-increasing demand for Network Using Secure Android Applications", International Journal
more radio spectrum. However, most easily usable of Scientific Research in Computer Science, Engineering and
Information Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
spectrum bands have been allocated, although many
5 Issue 1, pp. 438-442, January-February 2019. Available at doi :
studies have shown that these bands are significantly https://doi.org/10.32628/CSEIT195164
underutilized. These considerations have motivated Journal URL : http://ijsrcseit.com/CSEIT195164

the search for breakthrough radio technologies that


can scale to meet future demands both in terms of
spectrum efficiency and application performance.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


442
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195135

Web Site Ranking Feedback System


Shaik Muzammil*, Sai Kiran Yerramaneni
Department of Computer Science and Engineering, Vasireddy venkatadri Institute of Technology, Guntur,
Andhra Pradesh, India

ABSTRACT

We all search on google for something and get the results in the form of different websites with some
description. We generally click on first or second website links if results are not found we go down on google
page. The website ranking is given by search engines by different criteria. Last website won’t be seen by none
of the people in most cases and first website will be having a great market compared to last website. So we need
to help the last website to be moved up in results and help in generating revenue and have good rank on
searching by giving feedback. This system will provide the difference between the first website and last website
on the google results and will provide the feedback to the last website like content, links, images used by first
website which helps the last website to be used in his webpage. This system is user friendly which is built on
HTML as front end and Python flask as back end and used python package beautiful soup to parse HTML data
and to automate browser behaviour with python. This system is done on web mining which has three
categories firstly web content mining in which we scan the web pages and get to know the links, text, images
used. Secondly web usage mining in which reports are generated after analysis which contain the details of text,
images, links. Finally the web structure mining states that structural summary of website.
Keywords : Beautiful Soup Package, Web Crawling, Feedback, Python- Flask, Searching Results, Web Mining,
Website

I. INTRODUCTION perform this task, search engines use different page


rank algorithms. Some ranking algorithms use
World Wide Web is a major source to retrieve content mining and some use web structure mining.
information. It has very large amount of data. For the But all the existing algorithms have some limitations.
user, it is a difficult task to find a particular page or a The system gives feedback to the results which was
set of pages from this massive amount of data. It is clicked by user here the first the user gives query and
highly impossible to remember the URLs of all these it was searched on google automatically and user
web pages. To solve this difficulty, search engine acts have to choose the web links to show the difference
as an intermediator between the users and WWW. between it and provide the feedback to last website
to improve his ranking. In typical Search engine
The search engines are able to extract the environment Web pages having matched keywords
information from the WWW for the users’ given are fetched from the predefined database and ranked.
query. They are used to collect the web pages and Wide range of relevant Web pages having unmatched
display them to the user. The order of the URLs keywords are omitted in this procedure. Neglected
depends upon the rank given to the pages. To Web pages having relevant information are covered

CSEIT195135 | Received : 01 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 443-447 ]
443
Shaik Muzammil, Sai Kiran Yerramaneni Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 443-447

by the proposed method. Relevant Web pages are provided to webpage to increase its rank with respect
fetched and displayed irrespective of keyword to first page.
matching.
A. Implementation
Web Content Mining (WCM) concerns with retrieval
of information from WWW in a structured form and
thereby indexes the information for quick retrieval.
It is the process to extract useful information from
web documents contents. The documents consist of
images, audio, video, text files, structured records etc.
Figure 2 : Search bar to search queries
WCM technique is used on web documents and the
results page obtained from a search engine.
Here we can search for anything we get the weblinks
what the google has described in the google search.
II. METHODS AND MATERIAL

Web site ranking feedback system provides the user


to increase their rank in the search results and user
have to search the query as follows.

Figure 3. List of weblinks

Here we are get the list of weblinks which user


searched and here user can select the weblinks why
their link was least when compare to others.

The weblinks will be crawled and we will find the no


of links, images, page load time and found the
content of web page and keywords and find the social
networks present in the web page and some other
factors which help the web page to improve his
ranking and if the new one want to design a web
Figure 1 : workflows of system page he can use this system how the content should
be and how the factors to be taken in his web page to
User query is searched using google search engine have a better ranking in search results.
and extract the top n results from search query and
user selects the web links by checking the weblinks
and there system will count the following factor like
links, images, load time, content etc ,. Then
comparison will take place here the feedback is

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


444
Shaik Muzammil, Sai Kiran Yerramaneni Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 443-447

Rank factors plays a major role in ranking of webpage


here most of the users don’t know what is it? First of
all we need to scrap the webpage and then find page
load time and count the images and count the web
links for short urls and find the content it should be
minimum of 1800 words in a webpage and there
must be social connect in webpage which helps in
increase the page rank.
Figure 4 : Factors of different webpages

III. RESULTS AND DISCUSSION


Here the factors are calculated for every user checked
weblinks which helps the user why the least webpage
are having the rank with first page.

Algorithm

1. Input: user query


2. Display: Top n google results
3. user: choose the weblinks
4. calculate the rank factors
5. Display: Difference of webpages Figure 5. plotting the factors with webpages
6. Display: Feedback to least factor of
webpage
Here the graph is drawn with factors with the
Here same as google search the user enters query in different webpages which makes the user can
search bar and it display the top n results and user identify the drawbacks of the least webpage.
choose the weblinks to find why the rank is
difference and later it find the rank factors and We can easily identify the difference by using this
difference of webpages are made using the results of graph why the web page having least rank and how
rank factor and feedback is given to webpage whose to improve his ranking by using the first web page
rank factors are low. factors and he can have the good rating over the
webpage and have the better income.
Algorithm :
calculate the rank factors IV. CONCLUSION
1. Take the first link
2. scrap the webpage Web search engines are generally designed to provide
3. count the links services to all users without taking consideration of
4. count the images the user interests. Personalized web search includes
5. count the load time an individual user's interests when determining
6. repeat step 1 to second link relevant results to return. The main focus of project is
7. Difference of webpages using factors to increase the rank of last webpage of the google
result with the help of feedback given to the user

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


445
Shaik Muzammil, Sai Kiran Yerramaneni Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 443-447

checked query. Which helps the others who are International Conference on Advances in
designing the web page what to put in the webpage Computer Engineering, IEEE, 2010
and how the content resides on webpage. For the [8]. Frank McSherry "A Uniform Approach to
future work, consider the SEO tools and machine Accelerated PageRank Computation", ACM,
learning techniques for ranking the web page ranking 2005
not by factors.

V. REFERENCES Cite this article as :

Shaik Muzammil, Sai Kiran Yerramaneni, "Web Site


[1]. Gyanendra Kumar, Neelam Duhan, A. K.
Ranking Feedback System", International Journal of
Sharma "Page Ranking Based on Number of
Scientific Research in Computer Science, Engineering
Visits of Links of Web Page" 2011 2nd
and Information Technology (IJSRCSEIT), ISSN :
International Conference on Computer and
2456-3307, Volume 5 Issue 1, pp. 443-447, January-
Communication Technology (ICCCT-2011)
February 2019.
[2]. S. Geetharani, Dr M. Soranamageswari
Available at doi :
"Locationbased Ranking Method (LBRM) for
https://doi.org/10.32628/CSEIT195135
Ranking Search Results in Search Engines" at
Journal URL : http://ijsrcseit.com/CSEIT195135
2016 10th International Conference on
Intelligent Systems and Control (ISCO)
[3]. Nagappan V.K, Dr. P. Elango "Agent Based
Weighted Page Ranking Algorithm for Web
Content Information Retrieval"
[4]. Nandnee jain, Upendra Dwivedi "Ranking Web
Pages Based on User Interaction Time" at 2016
10th International Conference on Intelligent
Systems and Control (ISCO)
[5]. Tuhena Sen, Dev Kumar Chaudhary, Tanupriya
Choudhury "Modified Page Rank Algorithm:
Efficient Version of Simple Page Rank with
Time, Navigation and Synonym Factor" at 2015
International Conference on Advances in
Computing, Communications and Informatics
(ICACCI)
[6]. Ricardo Baeza-Yates and Emilio Davis ,"Web
page ranking using link attributes" , In
proceedings of 13 international World Wide
Web conference on Alternate track papers &
posters, PP.328-329, 2004.
[7]. Harmunish Taneja and Richa Gupta "Web
Information Retrieval using Query
Independent Page Rank Algorithm",

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


446
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951120

IoT Based Heart Attack Early Prediction


Shweta Gajbhiye , Bharati Vyas2, Shrushti Shrikhande 2, Anshika Janbandhu2, Komal Nagpure2,Mrunali Agashe2
1

1 Assistant Professor, Department of Computer Science and Engineering, Priyadarshini J. L. College of


Engineering, Nagpur, Maharashtra, India
2BE Students, Department of Computer Science and Engineering, Priyadarshini J. L. College of Engineering,
Nagpur, Maharashtra, India

ABSTRACT

Heart attack is most common disease that engulfs the patient’s precious life. This disease attacks a man so in a
flash that it barely gets treated. Many systems that detect heart attack came into the picture but they do have
many flaws. Some system does not work properly in cold weather and other doesn’t give accuracy. To have a
solution for Heart attack utilization of IoT and sensor system could be beneficial. Heart Attack Early Prediction
System predicts the Heart rate and accordingly sends the notification to the patient’s doctor and acquaintances.
Keywords : Heart Attack, IoT (Internet Of Things), Health Parameters, Heart Disease Prediction, Oximeter,
Heart rate sensor, Android App.

I. INTRODUCTION Monitors the Heart and even status logging and


social Sharing of recorded data.
Heart attack is most frequently occurring disease
that has a consequence of loosing a precious life. To overcome this health issue, an Android
Numerous people are mislaying their life owing to application could be designed and implemented.
heart attack and shortage of medical attention to The patient will be having the hardware device
patient at correct stage. There are many solutions with android application. The Heartbeat sensor will
that detects the heart beat like Fit Bit, Garmin and allow checking heartbeat reading and oxygen level
many more. Fit Bit and Garmin works like a smart and transmit them through Wi-Fi module.
watch. These devices track the users heart rate. Fit
Bit uses build in or connected GPS for many Heart Attack Early Prediction System will sense the
distance and tracking features. The device are Heart rate and oxygen level of the patient and will
categorized based on fitness level : Everyday, display on the Android Application. If the Heart
Active and Performance. Distributed computing rate extends the threshold value. Then the patient
helps the patient for monitoring the Health status will be considered in danger situation.
through albeit cell phones. Their is another Device
that is a laser device that monitors the Heart Beat, what is IoT?
Breathing rate and muscle activity from up to a
Internet of things simply “A network of internet
meter away with a single sensor called Zeev. Multi
connected objects able to collect and exchange
Detecting gadgets also came into the picture that
data”. It is commonly abbreviated as IOT.

CSEIT1951120 | Received : 12 Feb 2019 | Accepted : 24 Feb 2019 | January-February -2019 [ 5 (1) : 448-451 ]
448
Shweta Gajbhiye et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 448-451

IOT is the network of devices such as vehicles and et al. The models are trained and validated against a
home appliances that contain electronics, software, test dataset. Classification Matrix Methods are used
sensors, actuators, and connectivity that allows to evaluate the effectiveness of the models. Neural
these things to connect, interact and exchange data. Network is Difficult to understand. The system has
Embedded with the technology, these devices can 89.9 percent accuracy according to the Research[3].
communicate and interact through internet, and
they can be remotely monitored and controlled. Distributed computing has been incorporated in a
health care framework. The creators have
• Sensor Technology introduced a cloud-based keen healthcare
The capacity to recognize changes in the physical monitoring framework (CIHMS) for giving
status of things is additionally fundamental for restorative input to a patient through the cloud.
chronicle changes in the earth. In such manner, Albeit cell phones are constantly viewed as a
sensors assume an urgent job in overcoming any promising instrument to screen and deal with the
issues between the physical and virtual universes, patient's own health status, these gadgets have some
and empowering things to react to changes in their innate restrictions in calculation or information
physical condition. Sensors gather information concentrated undertakings. ECG monitoring system
from their condition, producing data and bringing contains WI-FI Bluetooth and Zigbee. ECG signs
issues to light about setting. For instance, sensors in were gathered and were transmitted to the IOT
an electronic coat can gather data about changes in cloud utilizing WI-FI. This information were send
outer temperature and the parameters of the coat to the family of the patient [4].
can be balanced in like manner.
Remote health care framework for monitoring
II. REVIEW OF LITERATURE electrocardiographic and temperature information
has been introduced in [5]. The framework
Milan Kumari et al. fathom cardiovascular disease comprises of three modules, in particular, an
dataset utilizing diverse information mining equipment module, Bluetooth module, and show
calculations, for example, Support Vector Machine, module. The equipment module is utilized for
Artificial neural systems (ANNs), Decision Tree, and information securing. The Bluetooth module is
RIPPER classifier. The creators break down the utilized for information transmission. At long last,
execution of these calculations through a few the information is shown by utilizing the
factual investigation factors, for example, presentation module. The obtained clinical
affectability, specificity, exactness, mistake rate. A information is sent to a database server by utilizing
choice emotionally supportive network for GPRS or Wi-Fi.
conclusion of Congenital Heart. Using data mining
and artificial neural networks ,the system predicts The exhibitions of the framework have been tried
Heart disease with nearly 100% accuracy. [1] on various patients and it typical arrhythmia
database.
Niti Guru et al. [2] connected a neural system for
forecast of heart disease, circulatory strain and has been discovered that the introduced framework
sugar. A model Intelligent Heart Disease Prediction is extremely useful for the doctors[5].
System (IHDPS) in view of information mining
procedures is introduced by Sellappan Palaniappan

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


449
Shweta Gajbhiye et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 448-451

A walking framework for monitoring the physical Heart Attack Early Prediction System could contain
restoration patients has been accounted for in [6]. Arduino board and a Wi-Fi module. The Heart Beat
The framework comprises of a multi-sensor-based sensor can begin detecting heart rate readings. If the
monitoring gadget, a cell phone with customer Heart beat will be above 120 bpm then the patient
application, an administration situated engineering will be in the danger situation. If the oxygen level
based server arrangement, and an application. The will be below 80 bpm then again, the patient will be
framework has been tried in a controlled condition considered in the dangerous situation. Likewise,
comprising of some healthy volunteers and some with the utilization of Wi-Fi module it will transmit
congestive heart disappointment patients. The test the information over web. The framework will
outcomes demonstrate that the introduced begin monitoring the heart rate of patient and
framework can identify and screen congestive heart instantly the heart rate goes above or beneath as far
disappointment and it can send criticism to the as possible the framework will send notification on
attendants for patient development [6]. the Android Application.

Ongoing universal health care framework for notification in the Android Application. An android
monitoring ECG motions by utilizing cell phone has application will track the heartbeat of specific
been exhibited in [7]. By utilizing this framework, patient and screen it effectively and give the crisis
the client can screen his ECG flag. The creators message on odds of Heart Attack.
have introduced a calculation for unusual heartbeat
detection and irregular heartbeat check delineate). III. CONCLUSION
The execution of the introduced framework has
been assessed against the MIT-BIH It has It has Heart disease is one of the major cause that leads to
been accounted for that the framework can death and early prediction of heart disease is very
distinguish at a top with a win rate of 97.8 percent important for every one. Garmin is a smart watch
and it is additionally ready to recognize strange and it predicts the heart attack but it does not work
heartbeat condition with a win rate of 78.9 percent. accurately in cold weather. Fit Bit is another
An unavoidable healthcare framework empowering alternative but even it has many flaws i.e. while the
self-administration for constant the introduced tracker is charging, patient cannot move and it is
framework in [8] comprises of patient health expensive too.Heart Attack Early Prediction System
monitoring framework, status logging, and social has many pros Over the existing solutions. It takes
sharing of the recorded data. The framework has two inputs i.e. Pulse rate and oxygen level of the
been executed by a cell phone, a wearable multi- patient to have the accuracy. It transmits the
detecting gadget contains pulse sensor and Notifications to the patients doctor and
temperature sensor, an administration situated acquaintances.
engineering for correspondence, and smaller scale
blogging administrations. The framework has been IV. REFERENCES
tried on 16 patients. The test outcomes demonstrate
that the introduced framework is anything but [1]. Milan Kumari, Sunila Godara, Comparative
difficult to learn and advantageous to use by endless Study of Data Mining Classification Methods
patients. in Cardiovascular Disease Prediction, IJCST
Vol. 2, Issue 2, June 2011.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


450
Shweta Gajbhiye et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 448-451

[2]. Niti Guru, Anil Dahiya, Navin Rajpal, Cite this article as :
Decision Support System for Heart Disease
Diagnosis Using Neural Network, Delhi Shweta Gajbhiye, Bharati Vyas, Shrushti
Business Review, Vol. 8, No. 1, January-June Shrikhande, Anshika Janbandhu, Komal Nagpure,
2007. Mrunali Agashe, "IoT Based Heart Attack Early
[3]. Sellappan Palaniappan, Rafiah Awang, Prediction", International Journal of Scientific
Intelligent Heart Disease Prediction System Research in Computer Science, Engineering and
Using Data Mining Technique, 978-1-4244- Information Technology (IJSRCSEIT), ISSN : 2456-
1968-5/08/25.00 2008 IEEE. 3307, Volume 5 Issue 1, pp. 448-451, January-
[4]. http://www.nlm.nih.gov/medlineplus/magazi February 2019.
ne/issues/winter11 Journal URL : http://ijsrcseit.com/CSEIT1951120
[5]. Bourouis, A., Feham, M., and Bouchachia,
A.(2011), Ubiquitous Mobile Health
Monitoring System for Elderly (UMHMSE),
International Journal of Computer Science
and Information Technology, Vol.2, No. 3,
June, pp. 74-82
[6]. Yuce, M. R.(2010) Implementation of
wireless body area networks for healthcare
systems, Sensor and Actuators A:Physical,
Vol. 162, No. 1, July, pp. 116-129
[7]. Lei Clifton, David A. Clifton, Marco A. F.
Pimentel, Peter J. Watkinson, and Lionel
Tarassenko (2014), Predictive Monitoring of
Mobile Patients by Combining Clinical
Observations with Data From Wearable
Sensors, IEEE Journal of Biomedical and
Health Informatics, Vol. 18, No. 3, May , pp.
722-730
[8]. Parane, K.A., Patil, N.C. ; Poojara, S.R. ;
Kamble, T.S(2014) Cloud based Intelligent
Healthcare Monitoring System, In the
proceedings of International Conference on
Issues and Challenges in Intelligent
Computing Techniques (ICICT), February 7-
8, Ghaziabad, Indian, pp. 697-701

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


451
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951119
Shakti : An Android App for Mapping of Crime against Women
Diksha Ingle, Pritee Dhale, Priyanka Palsapure, Sagar Pazare, Dhawal Chimurkar, Niket Farfad, Prof. Prasanna Lohe
Department of Computer Science, JD College of Engineering and Management, Nagpur, Maharashtra, India

ABSTRACT
Women biggest issue we are facing today. Many apps have been developed to provide safety to women through
Android phones. Many of the times government conduct different awareness events for stopping crime against
women. But that's not enough; there are various kinds of crimes women are facing today. Now it's time to aid
the roots of these crimes. This paper presents Shakti, an Android Application for mapping of crime against
women. This will help the government to identify which crime has the highest frequency in a particular area
and can implement an appropriate campaign for aiding its root cause. In case of an emergency, the victim can
send its location along with a help message just by clicking the volume button 5 times. Also, the victim can
reach-out to NGO for help. Shakti will feel like having family power with you at any time.
Keywords : Android, Women, Crime, Mapping, Location, Help message, NGO.

I. INTRODUCTION that the government runs for helping the victims are
appreciable. How about not letting any women be a
From past years, it has been recorded that the graph victim? For this, a study is needed for every area and
of crime against women has risen. Though the list of the type of crime occurring most frequently.
NGO's, Women Empowerment campaigns, safety
applications, and more steps for the safety of women "Dowry" can be aided with a literate society.
has continuously risen with time, the crime graph is "Domestic Violence" needs a little work on the
still incrementing. There are various government mindset of the family; in-laws. "Rape" and "Sexual
bodies as well as private organizations working for Assault" can be stopped if the females are equipped
women empowerment and helping the victims. But with right safety tools, and self-defense training.
the call of today is to find the root cause of these
crimes and treat it accordingly. Rape, Sexual assault, A. Existing Systems:
Domestic violence, Dowry, these are major of the As a part of a literature survey, we investigated some
crime list. Everyone is talking about empowering applications that offer the same or similar services for
women, blaming society. What is making a society Android and other platforms. The aim is to see how
guilty? Is its people with narrow thinking and evil these applications work and to see how they can be
minds, and these are the causes of illiteracy, lack of improved.
life values, in short- the mindset of people.
The existing women safety apps:
In everyday Newspaper, some areas are in headlines Women Safety: This app is offered by Awesome Apps
for "Rape" and Sexual "Assault cases". Some areas are Inc. Released on 28th Aug 2015. With just a tap of a
in the headline for "Domestic Violence". The projects button, the app sends an e-mail to preconfigured

CSEIT1951119 | Received : 12 Feb 2019 | Accepted : 24 Feb 2019 | January-February -2019 [ 5 (1) : 452-455 ]
452
Diksha Ingle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 452-455

email ID(s) along with your location and a link to Limitation : You to keep this women safety app
Google Map. The app also captures 2 pics, one with always running in the background which can drain
front camera and the other with the back camera, the battery of your phone. It does not show every
and a video or audio clip and uploads it to our server. area of information. It gets crashed in cases.
A link to the pic, audio or video is emailed to a
preconfigured email Id(s). redEye - Women Safety: This is a very useful app for
women safety. It does not need any additional action
Limitation: It is not useful in case of emergency. Also, by the user in any emergency situations. The app
there is an issue in managing phone numbers. Also, itself let your dear ones know that you are in danger
the panic message has no notification, so no use if the through an alert alarm. In this app, the user will
other sees it later. choose the source and the destination. The redEye
app then keeps tracking you. If the app finds that the
Raksha, A Women's Safety App: This app is offered user's path has deviated or the user is idle in the same
by Portal Perfect, released on 27th May 2018. This place for a configured duration, then the app will
app understands the worries and problems of dear trigger the alert. Once you reach your destination
ones, be it your child, mother, daughter, sister or safely you can manually disable the app.
wife. By using this application you can check their
current location. The special feature of this app is Limitation: This app only allows 2500 uses per day by
"Are you in trouble" button. all users. More than 5000 people have downloaded
the app, so only half or less than half users are able to
Limitation: The emergency messages are not able to use the app.
send. It also failed working many times.
Here, we introduce an android app that ensures the
My Safetipin: Personal Safety & Women Safety App. safety of women. It reduces the risk and helps us in
This app gives safety scores to an area and helps you need by identifying the location of a person who is in
to get alerts when you enter in an unsafe area. It danger.
helps in finding the safest route for you. You can also
choose the alternate safe location when you give II. PROPOSED SYSTEM
Google maps the access of this women safety app.
SafetiPin collects primary data on Women Safety in For the need of surveying various areas and
public spaces at night, on a number of parameters. implementing appropriate programs in that area to
• Lighting diminish the roots of crime against women, we have
• Visibility: Are there ‘eyes on the street'. Can developed an android application SHAKTI. In today's
others see you world android phone is a basic thing which everyone
• Diversity: Presence of women and children carries, with SHAKTI in your android phone women
around now will feel secure and protected all the time.
• Crowd: How densely crowded is the area Following are the features of our Android application:
• Public Transport: How close it is The user will rate a type of crime for their particular
• Walk Path: Is there a proper place to walk area. This data will be graphically represented on our
• Security: Are police or private security nearby. website.
• Openness: Can you see all around
• Feeling: How safe do you feel

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


453
Diksha Ingle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 452-455

In cases when the victim is unable to report against (2)Server Side: Data entered by the users (members
crime to the police, the victim can complain to the registered, crime reported, etc.) via the app is
social NGO through this application. uploaded to the web server, the crime details for
The user will save three contacts of three family various areas will be graphically represented on the
members. Whenever the user feels any danger like website and complaint is sent via text message to
crime coming its way, the user can press volume key registered NGO for further processing and
of its Android phone five times and an emergency authenticity. Also, admins get all the privileges
message for help will be delivered to their family through this server.
members along with the location URL through GPS.
Family members will also be notified if the victim's BLOCK DIAGRAM
phone battery is below 20%.
Shakti: An Android app for mapping of crime against
women is implemented into two modules: (1) Client-
side, (2) Server Side.

(1)Client Side: Client-side involves the interaction of


the user. The user can register in the application
through their mobile number. Once registered they
can log in any time. After successful login user first
needs to enter its area and its pin code, select the
frequently happening crime from the drop-down list Fig: 1. Depicts the block diagram of server side
and then rate the crime. In this way, the data about
various areas will be collected from the user. III. IMPLEMENTATION

Next, the users need to enter three family members Shakti android application is being implemented
contact number with an emergency message. In the using the following technologies: (1) Android (2) Java
situation of any crime, the user will have to press the (3) SQL (4) Cloud (5) Web Technologies (6) Database
volume button five times. Saved emergency message (7) GPS
will be forwarded to the saved contact numbers along
with the victim's location through GPS. (1)Android: Android is the most widely used mobile
OS motorized by Linux kernel. This is the first
The user can also ask for help to the NGO directly complimentary platform which is robust and is
through our application. expected to gain much popularity. It was developed
by Google team and allows writing managed code in
In cases if the user’s mobile battery drains below 20%, the Java language. The current system is developed
the app will send an alert message to the dear ones. on the basis of android platform. Android utilizes a
Also user can set a secret code in the app. When the custom virtual machine that was designed to
dear ones want to reach the user, they will send this optimize memory and hardware resources in a smart
secret code as a text message and they will phone. Android does not differentiate between the
automatically get the users mobile location through phone’s core applications and third-party applications.
GPS . Any application that is built will definitely have
equal access to a phone’s capabilities providing users

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


454
Diksha Ingle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 452-455

with a broad spectrum of applications and services (2) different areas. Thus, the social bodies working for
Java: Java is an object-oriented, class-based, the security of women can implement an aid to those
concurrent, secured and general-purpose computer- areas to halt the crimes.
programming language.
V. ACKNOWLEDGEMENT
After front-end elements are finalized java classes
and packages are used to make them work. First, and foremost we would like to thank God for
the wonderful opportunities and challenges he has
(3) SQL: Structured Query Language or SQL is a given to us. We express our sincere gratitude to our
standard Database language which is used to create, guide Prof. Prasanna Lohe, Head, Department of
maintain and retrieve the relational database. It is Computer Science and Engineering for his valuable
particularly used to work with structured data where contribution in fulfilling the requirement related to
there is a relation associated within the data itself. the project and for providing his valuable guidance,
(4) Cloud: Cloud storage involves stashing data on patience and for encouraging us to do our best.
hardware in a remote physical location, which can be
accessed from any device via the internet. Clients Cite this article as :
send files to a data server maintained by
a cloud provider instead of (or as well as) storing it on Diksha Ingle, Pritee Dhale, Priyanka Palsapure, Sagar
their own hard drives. Pazare, Dhawal Chimurkar, Niket Farfad, Prof.
(5) Web Technologies: The data collected in this Prasanna Lohe, "Shakti : An Android App for
android women safety application we be displayed Mapping of Crime against Women", International
using website. Journal of Scientific Research in Computer Science,
(6) Database: A database is an organized collection of Engineering and Information Technology
data, generally stored and accessed electronically (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
from a computer system where databases are more 452-455, January-February 2019.
complex they are often develop using formal design Journal URL : http://ijsrcseit.com/CSEIT1951119
and modeling techniques.
Android comes with in-built SQLite database. SQLite
supports all the relational database features.
(7) GPS: GPS or Global Positioning System is a
satellite navigation system that furnishes location and
time information in all climate conditions to the user.
GPS is used for navigation in planes, ships, cars and
trucks also. The system gives critical abilities to
military and civilian users around the globe. GPS
provides continuous real time, 3-dimensional
positioning, navigation and timing worldwide.

IV. CONCLUSION

Shakti will help to easily know the major crimes


against women and violence happening under

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


455
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951123

Robust Video Data Hiding in Forbidden Zone


Ruchika S. Bhambore1, Ashwini B. Gurudeo1, Shital B. Borkar1, Pallavi G. Thul1, Chanchala V. Udepurkar1, Prof. Anup
Bhange2
1BE, Department of Computer Technology, KDK College of Engineering, Nagpur, Maharashtra, India

2Assistant Professor, Department of Computer Technology, KDK College of Engineering, Nagpur, Maharashtra,

India
ABSTRACT

Video data hiding is yet a vital research subject because of the structure complexities included. As of late, there
are loads of frameworks are present. The general population designed a substantial thing to ensure the data and
there are bunches of hiding procedures are to be developed for security reason. In any case, that strategies can
be hack by unapproved clients is downside in existing frameworks so that propose the new framework for
example Data hiding behind the video utilizing forbidden zone and selective embedding. We propose another
video data hiding technique that makes utilization of eradication amendment ability of Repeat Accumulate
codes and prevalence of Forbidden Zone Data Hiding. This framework makes utilization of redress capacity of
duplication store codes and favorable position of forbidden zone data hiding is utilized. In this examination the
encryption and decoding method is utilized to give the security key. Without that key, nobody can see the first
data. This system is utilized to shield the database from unapproved and the dangerous powers .It has
substantial deletion ability of data hiding.
Keywords: Data Hiding, Forbidden Zone, Quantization Index Modulation [QIM], Repeat Accumulate Codes,
Selective Embedding

I. INTRODUCTION
Consequently, video data hiding keeps on
Data hiding is the way toward embedding data into establishing a functioning examination region. Data
a host medium. When all is said in done, visual and hiding in video arrangements is performed in two
real media are favored because of their wide nearness noteworthy ways: bitstream-level and data-level. In
and the resilience of human perceptual frameworks bitstream-level, the redundancies inside the present
included. pressure measures are misused. Ordinarily, encoders
have different alternatives amid encoding and this
In spite of the fact that the general structure of the opportunity of choice is appropriate for control with
data hiding process does not rely upon the host media the point of data hiding. In any case, these strategies
type, the strategies change contingent upon the idea exceptionally depend on the structure of the
of such media. For example, picture and video data bitstream; henceforth, they are very delicate, as in
hiding share numerous regular focuses; be that as it much of the time they can't endure any organization
may, video data hiding requires progressively transformation or transcoding, even with no
complex structures because of the extra fleeting noteworthy loss of perceptual quality. Subsequently,
measurement. this sort of data hiding techniques is by and large

CSEIT1951123 | Received : 15 Feb 2019 | Accepted : 25 Feb 2019 | January-February -2019 [ 5 (1) : 456-461 ]
446
Ruchika S. Bhambore. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 456-461

proposed for delicate applications, for example, utilize the BCH code to expand blunder amendment
verification. capacity. The creators perform 3D interleaving so as
to dispose of the nearby burst of mistakes. Also, they
Then again, data-level techniques are increasingly propose a worldly synchronization procedure to
strong to assaults. Subsequently, they are reasonable adapt to fleeting assaults, for example, outline drop,
for a more extensive scope of utilization. addition, and rehash. In this paper, we propose
Notwithstanding their delicacy, the bitstream-based another square based selective embedding type data
techniques are as yet alluring for data hiding hiding structure that exemplifies Forbidden Zone
applications. For example, the excess in square size Data Hiding (FZDH) and RA codes as per an extra
choice of H.264 encoding is misused for hiding data. fleeting synchronization instrument. FZDH is a down
In another methodology, the quantization parameter to earth data hiding strategy, which it appeared to be
and DCT (Discrete Cosine Transform) coefficients are better than the traditional Quantization Index
changed in the bitstream-level. In any case, the vast Modulation
majority of the video data hiding techniques use
uncompressed video data. Sarkar proposes a high (QIM) .RA codes are as of now utilized in picture
volume change area data hiding in MPEG-2 videos. and video data hiding because of their vigor against
They apply QIM to low-recurrence DCT coefficients eradication. Vigor permits taking care of
and adjust the quantization parameter dependent on desynchronization among embedder and decoder
MPEG-2 parameters. that happens because of the distinctions in the chose
coefficients. So as to join outline synchronization
Besides, they differ the embedding rate contingent markers, we segment the squares into two gatherings.
upon the sort of the casing. Therefore, inclusions and One gathering is utilized for edge marker embedding
eradications happen at the decoder, which causes and the other is utilized for message bits.
desynchronization. They use Repeat Accumulate (RA)
codes so as to withstand deletions. Since they adjust By methods for basic standards connected to the
the parameters as per the sort of casing, each edge is edge markers, we present a specific dimension of
handled independently RA codes are as of now power against casing drop, rehash, and creepy crawly
connected in picture data hiding. Versatile square assaults. We use precise RA codes to encode message
choice outcomes in desynchronization and they use bits and edge marker bits.
RA codes to deal with eradication. Additions and
deletions can be additionally taken care of by Each piece is related to a square living in a gathering
convolutional codes. The creators use convolutional of casings. Irregular interleaving is performed
codes at embedder. In any case, the weight is put on spatiotemporally; thus, reliance on nearby attributes
the decoder. Different parallel Viterbi decoders are is diminished. Host flag coefficients utilized for data
utilized to address desynchronization mistakes. In hiding are chosen at four phases. In the first place,
any case, it is seen that such a plan is effective when outline determination is performed. Casings with an
the quantity of chose host flag tests is significantly adequate number of squares are chosen. Next, just
less than the absolute number of host flag tests. 3-D some foreordained low-recurrence DCT coefficients
DWT space is utilized to conceal data. They use LL are allowed to conceal data. At that point, the normal
subband coefficients and don't play out any versatile vitality of the square is relied upon to be more
determination. In this way, they don't utilize blunder prominent than a foreordained limit. In the last stage,
redress codes strong to eradication. Rather, they the vitality of every coefficient is contrasted with

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


457
Ruchika S. Bhambore. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 456-461

another edge. The unselected squares are marked as non-covering, which can be accomplished by
eradication and they are not handled. For each utilizing a base quantizer and moving its recreation
chosen square, there exists a variable number of focuses relying upon m, like Dither Modulation. A
coefficients. These coefficients are utilized to install run of the mill embedding capacity that utilizes a
and unravel a solitary message bit by utilizing the uniform quantizer. we need to consider that the In
multi-dimensional type of FZDH that utilizes cubic the unrivaled field, the hiding advancement, for
cross-section as its base quantizer. example, least critical bit(LSB) extra, is done in the
predominant field, while change space techniques;
II. LITERATURE REVIEW conceal data in another area, for example, wavelet
space. Least noteworthy piece (LSB) is the
The writing review is the most critical advance in the unassuming type of Steganography. LSB depends on
product improvement process. Before building up the infusing data in the littlest huge piece of pixels,
device it is important to decide the time factor, which data to a minor change on the spread picture,
economy n organization quality. When these things r which is not noticeable to the human eye. Since this
fulfilled, ten following stage is to figure out which strategy can be effectively part, it is more fragile to
working framework and language can be utilized for assaults. LSB framework effectively affects the
building up the instrument. When the software arithmetical data of picture like a histogram.
engineers begin assembling the instrument, the
developers need a great deal of outside help. This Safeguards could be the alarm of a shrouded
help can be acquired from senior software engineers, correspondence by simply testing the Histogram of a
from a book or from sites. Before building the picture. A decent answer for reject this deformity
framework the above thought r considered for was LSB indistinguishable. LSB Identical was an
building up the proposed framework. Forbidden incredible advance forward in Steganography
Zone Data Hiding (FZDH) is presented in [8]. The strategies and numerous others get plans from it.
technique relies upon the Forbidden Zone (FZ) idea, Downsides of Existing System are that the strategy
which is characterized as the host flag go where no given in Existing framework is effectively broken.
change is permitted amid data hiding process. The measure of the put-away data is little.

FZDH makes utilization of FZ to alter the power III. PROPOSED SYSTEM


intangibility exchange off the mapping capacity in (2)
expresses that the host flag is adjusted by including While there are various programming improvement
an extra term, which is a scaled adaptation of the models the idea of our task has disentangled the
quantization distinction. In 1-D, this extra term is determination procedure. Programming
scalar, though in N-D have flag is moved along the advancement procedures, for example, the cascade
quantization contrast vector and towards the and V display have been dismissed on account of
remaking purpose of the quantizer. Consequently, their inflexible structures, selecting a lithe
embedding contortion is diminished and ended up programming improvement approach which is
littler than the quantization mistake. increasingly responsive and iterative in its tendency.

So as to satisfy the prerequisite of common Building up a steganography system that is fit for
prohibition, the reproduction purposes of the hiding data in a video le lies at the core of this
quantizers that are recorded by various m ought to be undertaking, and it is alluring that the resultant

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


458
Ruchika S. Bhambore. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 456-461

arrangement is as secure as could be expected under We talk about parts of our underlying examination,
the circumstances. Lithe plan strategies enable us to which included a wide examination of steganography
assess and enhance our executed systems as we (utilizing sound, picture and video compartments).
advance. Moreover, our overview of the writing and The understanding given by this exploration in
our starter explore (see A.2) yielded incredibly uenced the plan choices that pursue.
constrained data on the best way to plan and execute
a video steganography arrangement of this nature. In Steganography and Steganalysis are in a ceaseless
light of this, coordinated advancement would enable fight. At whatever point a decent steganography
us to roll out considerable improvements to the system is built up, another Steganalysis method is
structure and execution anytime something that can likewise created endeavouring to overcome it
only with significant effort be accomplished with Steganalysis is the workmanship and exploration of
other programming designing models. identifying mystery messages shrouded utilizing
steganography. Once there is proof that a message is
concealed, the objective of steganography is crushed
regardless of whether the message was not separated.
In spite of the fact that steganography methods may
clearly straightforward to the human eyes, assaults on
them are yet conceivable. Any embedding procedure
definitely leaves follows in the stego-item and
modifies a portion of its properties, which present
irregular attributes and some debasement as far as
quality. Consequently, Steganalysis can be arranged
into two classes: inactive Steganalysis and dynamic
Steganalysis. Aloof Steganalysis distinguishes the
nearness or nonappearance of shrouded message or
recognizes the embedding calculation utilized. While
dynamic Steganalysis change, extricate or obliterate
the concealed message or concentrate a portion of its
traits, for example, message length.

It is at this crossroads that we might want to bring up In the first step, frame selection is completed and the
that our exploration has been considerably more test a selection of frames are managed block-wise. For
than we could have predicted. Our decision of a each block, only a first bit is hidden. After obtaining
coordinated improvement process has been a 88 DCT of the block, energy check is performed on
demonstrated gift on various events. While most of the coefficients that are predefined in a mask. Selected
this section will report the structure and execution of proficient of variable length are used to hide data bit
our last arrangement, we will likewise feature how m. m is a supporter of message bits or frame
our plan and usage have changed with the emphases. synchronization markers. Message sequence of each
Being in contact at the last phase of usage has been an group is obtained by using RA encryptions for T
unpredictable procedure and a few essential exercises consecutive frames. Each block is assigned to one of
have been learned en route. these groups at the start. After the inverse convert
host frame is obtained. Decoder is the dual of the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


459
Ruchika S. Bhambore. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 456-461

embedded, with the exception that frame selection is E. Coefficients Selection


not performed. Energy of each constant is compared to another
beginning T2. If the energy is above T2, then it is used
Marked frames are detected by using frame during data inserting together with other selected
synchronization markers. Decoder employs the same constants the same block.
system parameters and concludes the marked signal
values that will be fed to data insertion step. Not- F. Block Partitioning
selected blocks are handled as erasures. Erasures and Two split data sets are inserted; message bits (m1) and
decoded message data possibilities (om) are passed to frame synchronization markers (m2). The block
RA decoder for sequential frames as a entire and then locations of m2 are resolute randomly depending on a
the secret data is decrypted. random key. The rest of the blocks are kept for m1.
A. Selective Embedding The same splitting is used for all frames. m2 is
Host signal models, which will be used in data hiding, inserted frame by frame. Continuously the other
are single-minded adaptively. The selection is hand, m1 is single to T sequential frames. Both of
performed at four stages: border variation, frequency them are found as the outcomes of the RA encoder.
band willpower, block selection, and coefficient
selection. Of data. The proposed system architecture G. Erasure Handling
is shown in below figure. Due to adaptive block collection, de-synchronization
occurs between embedded and translator. Because of
attacks or even embedding operation, decoder may
not perfectly determine the selected blocks at the
embedded. In order to overcome this problem,
mistake modification codes strong to erasures, such as
RA codes are used in video data hiding in previous
hard work. RA code is a low complication turbo-like
code. It is collected of repetition code, interleave, and
a convolutional encoder. The source bits (u) are
repetitive R times and accidentally permuted
B. Frame Selection dependent on a key. The interleaved sequence is
A number of numbers of blocks in the whole frame is passed through a convolutional encoder with a
calculated. If the percentage of selected blocks to all transfer function 1/ (1 + D), where D represents a
blocks is above a certain value (T0) the frame is _rst-order stay. In efficient RA code, input is 16
handled. Otherwise, this frame is avoided. placed at the start of the output as shown in Fig. 1.

C. Frequency Band In this paper, we apply logical RA codes to find m1 as


Only certain DCT constants are operated. Middle u1+v1 and m2 as u2+v2. Here, u1 denotes the
frequency band of DCT constants. encrypted message bits and u2 is the encrypted frame
management marker bits. RA code is decrypted using
D. Block Selection sum-invention algorithm. We apply the message-
Energy of the constants in the mask is added. If the passing algorithm given in.
vitality of the block is overhead a assured value (T1)
then the block is managed. Otherwise, it is avoided.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


460
Ruchika S. Bhambore. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 456-461

IV. CONCLUSION V. REFERENCES

In this paper, we proposed another video data [1] Ersin Esen, A. Aydin Alatan,\Robust Video Data
hiding structure that makes utilization of deletion Hiding Using Forbidden Zone Data Hiding And
redressability of RA codes and prevalence of FZDH. Selective Embedding ",IEEE ,VOL. 21, NO. 8,
The strategy is likewise powerful to outline control AUGUST 2011.
assaults through casing synchronization markers. To [2] K.Mohan, S.E.Neelakandan \Secured Robust
begin with, we thought about FZDH and QIM as the Video Data Hiding Using Symmetric Encryption
data hiding strategy for the proposed structure. We Algorithms ", IJIRE ,VOL.6,DECEMBER 2012
saw that FZDH is better than QIM, particularly for [3] R. Ravi Kumar V., Kesav Kumar \Selective
low embedding mutilation levels. The structure was Embedding and Forbidden Zone Data Hiding for
tried with MPEG-2, H.264 pressure, scaling and Strong Video Data Thrashing ",IJETT ,VOL.
casing rate transformation assaults. Ordinary 4,SEPTEMBER 2013
framework parameters are accounted for mistake-free [4] Mr.Sudheer Adepu, Mr.P. Ashok , Dr.C.V.Guru
translating. The outcomes show that the structure Rao \A Security Mechanism for Video Data
can be effectively used in video data hiding hiding " ,IJCTT,VOL.4, August 2013
applications. For example, Tardos fingerprinting, [5] Resoju Omprakash and D. Jyothi \Block Based
which is a randomized development of double unique Adaptive Videodata Hiding Technique", IJMSTH,
mark codes that are ideal against agreement assault, 2012
can be utilized inside the proposed structure with the [6] Mr. Mritha Ramalingam \Stego Machine Video
accompanying settings. The length of the Tardos Steganography using Modified LSB Algorithm ",
unique mark is AC2 0 In 1 ε1, where A will be an World Academy of Science, Engineering and
element of false positive likelihood (ε1), false Technology,2011
negative likelihood, and the greatest size of the [7] W. Bender D. Gruhl,N. Morimoto,A. Lu,
colluder alliance, (Co). We likewise analyzed the \Techniques for data hiding ", IBM SYSTEMS
proposed structure against the standard JOURNAL, VOL.35, NOS 3 and 4, 1996
watermarking technique, JAWS, and a later
quantization-based strategy. The outcomes Cite this article as : Ruchika S. Bhambore, Ashwini B.
demonstrate a huge predominance over JAWS and a Gurudeo, Shital B. Borkar, Pallavi G. Thul, Chanchala
V. Udepurkar, Prof. Anup Bhange, "Robust Video
practically identical execution with. The analyses
Data Hiding in Forbidden Zone", International
likewise revealed insight into conceivable
Journal of Scientific Research in Computer Science,
enhancements for the proposed technique. To begin Engineering and Information Technology
with, the system includes various edges (T0, T1, and (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
T2), which are resolved physically. 456-461, January-February 2019.
Journal URL : http://ijsrcseit.com/CSEIT1951123
The scope of these limits can be examined by
utilizing a preparation set. At that point, a few
heuristics can be derived for the legitimate choice of
these edge esteems. Moreover, joining of the human
visual framework based spatiotemporally adjustment
of data hiding technique parameters as in stays as a
future bearing.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


461
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/IJSRCSEIT

The Internet-of-Things (IoT) Security : A Technological


Perspective and Review
Dr. Yusuf Perwej*1, Firoj Parwej2, Mumdouh Mirghani Mohamed Hassan3, Nikhat Akhtar4
*1Assistant Professor, Department of Information Technology, Al Baha University, Al Baha, Kingdom of Saudi Arabia
(KSA)
2Research Scholar, Department of Computer Science & Engineering, Singhania University, Jhunjhunu, Rajasthan, India
3Assistant Professor, Department of Computer Science, Al Baha University, Al Baha, Kingdom of Saudi Arabia (KSA)
4Research Scholar-Ph.D, Department of Computer Science & Engineering, Babu Banarasi Das University, Lucknow, India

ABSTRACT

Recent years have seen the swift development and deployment of Internet-of-Things (IoT) applications in a
variety of application domains. In this scenario, people worldwide are now ready to delight the benefits of the
Internet of Things (IoT). The IoT is emerging as the third wave in the evolution of the Internet. The 1990s’
Internet wave connected 1.2 billion subscribers while the 2000s’ mobile wave connected another 2.4 billion.
Actually, IoT is expected to consist of more than 84 billion connected devices generating 186 zettabyte of
data by 2025, in the opinion of IDC. It includes major types of networks, such as distributed, ubiquitous, grid,
and vehicular, these have conquered the world of information technology over a decade. IoT is growing fast
across several industry verticals along with increases in the number of interconnected devices and diversify of
IoT applications. In spite of the fact that, IoT technologies are not reaching maturity yet and there are many
challenges to overcome. The Internet of Things combines actual and virtual anywhere and anytime, fascinate
the attention of both constructor and hacker. Necessarily, leaving the devices without human interference for a
long period could lead to theft and IoT incorporates many such things. In this paper, we are briefly discussing
technological perspective of Internet of Things security. Because, the protection was a major concern when just
two devices were coupled. In this context, security is the most significant of them. Today scenario, there are
millions of connected devices and billions of sensors and their numbers are growing. All of them are expected
secure and reliable connectivity. Consequently, companies and organizations adopting IoT technologies require
well-designed security IoT architectures.
Keywords : Internet of Things (IoT), Sybil Attack, IoT Authentication, IoT Trust, IoT Protocols, IoT Security

I. INTRODUCTION connected to common Internet using tiny computing


devices with communication technology. At present,
The size of computer systems has diminished our world includes billions of computing devices and
drastically over the years, from mainframes sensors that are continually sensing, collecting,
encompassing whole rooms, via desktop computers, integrate [2], and analyzing significant amount of our
and down to smart cell phones [1]. At the turn of the personal information. This would permit anything to
century, a new concept appear called the Internet of speak to everything, making everyday life trouble-
Things, envisioning all “things” in the world free for everybody. The massive use case for the

CSEIT195193 | Received : 12 Feb 2019 | Accepted : 25 Feb 2019 | January-February -2019 [ 5 (1) : 462-482 ]
462
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

Internet of Things today is consolidation of data, and exposed to attacks by malicious opponent [1]. An
responding to the collected data in a useful way [3]. invasion may compromise an IoT device and use it as
While connecting all our things to the Internet will a platform for launching invasion on other IoT
permit us to advantage insight into our lives and devices. IoT security introduces technological
environment, we can potentially permit others to defiance at the device, network and platform level.
advantage the same insight if security is not managed Therewith, there is the process challenge of
correctly. A couple of security anxiety on a single organizing the security technologies in an end-to-end
device like as a mobile phone can swiftly turn to 60 manner. One thing is certain, when evaluating
or 70 anxiety when considering multiple IoT devices security necessity is that, then IoT is still very much a
in an interconnected business or home. In light of the work in progress. This paper provides an overview of
significance of what IoT devices have access to, it is Internet of Things (IoT) stage in section 2. We are
an essential to understand their security peril [1]. briefly discussing Internet of Things (IoT) threats in
section 3, and in section 4, we describe Internet of
The Internet of Things (IoT) [4] enables everywhere Things scheme. We categorize the reliance in section
communication between various devices. From 5. In section 6, we discuss public key infrastructures
entering patient details to watch post-surgery, from in Internet of Things. In section, 7 and 8 Internet of
parking vehicles to tracking vehicles, from childcare Things design thought for digital certificates and
to elder care, from smart cards to near field cards, protocols. In the last section discuss guidelines for
sensors are making their impendence felt. Sensors secure the IoT devices.
play an important role in the IoT as well. In spite of,
the functionality and operations of the IoT heavily II. INTERNET OF THINGS (IoT) STAGE
depend on the underlying network connectivity
structure. In the opinion of Gartner [5], it is required The IoT requires five stages, from data collection to
that the number of Internet connected devices will data delivery to the end [1] users on or off demand, as
increase from around 28 billion to 50 billion by 2020. shown in figure 1.
In spite of the IoT features everywhere
communication [6] among all kinds of electronic
devices, it inevitably raises security concerns due to
seamless infiltration and automated integration
among all sorts of applications. The IoT works across
miscellaneous networks and standards. Security and
privacy are considered the most important IoT
challenges [7]. In particular, no network is free from
security threats and vulnerabilities. Each of the IoT
layers is uncovered to various types of threats. In the
opinion of Gartner [8] security and risk, concerns
will continue to be the greatest hindrance to IoT
adoption. The market for IoT distinguished security Figure 1. The Internet of Things (IoT) Stage
solutions will dramatically expand in 2017 as current
security providers aggressively retool present In the upstanding stage, sensors collect data from the
capabilities to address IoT security risks IoT devices environment or object under measurement and turn
have often limited resources and may be more it into useful data [9]. Actuators can also intervene to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


463
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

modify the physical conditions that generate the data. security problems. In the future, maybe around the
The thing may be a constant body (body sensors or year 2020 with IPv6 and the 5G network, billions of
RFID tags) or a dynamic vehicle (sensors and chips) miscellaneous things will be part of the IoT. Privacy
[10]. The actuating and sensing stage involucre and security will be the major factors of worry at that
everything from legacy industrial devices to robotic time [12]. IoT security threats and attacks will
camera systems, water-level detectors, heart rate escalate as the IoT devices become everyday events.
monitors, accelerometers, and air quality sensors. In The security threat is high enough for Gartner to
the secondly data collected on stage first should be estimate spending on IoT security is required to reach
stored and making it obtainable for analysis. If the near $600 million in 2018. In this [13] report,
thing has its own local memory, data can be stored prophecy that 30% of attacks in enterprises will
[7]. The different technologies are appropriate for include IoT. In this section, we are discussing
this purpose, depending on the scenario. The high security problems in three dimensions, components,
performance storage solutions are suggested for based on stage, and architecture.
companies performing data analysis with the
distributed system Hadoop. Normally, IoT 3.1 The Components Based Incursion
components are installed with low memory and low
processing potential [1]. The cloud takes over the As a technology, IoT is peerless since it has a role to
answerability for storing the data in the case of play in consumer, industrial worlds and enterprise.
stateless devices. In the thirdly stage, IoT analyzes The things are diverse in nature, communicating
the data stored in the cloud DCs and provides sensitive data over a distance [13]. The Internet of
intelligent services for work and life in hard actual Things (IoT) is extremely heterogeneous, highly
time as well as analyzing and responding to queries, dynamic, always available, [7] and consequently
then IoT also handle things. The IoT proposal for always vulnerable to attack. Apart from attenuation,
intelligent processing and control services to all stealing, loss, violation, and disaster, data can also be
things homogeneously. In the fourth, data [1] concocted and modified by compromised sensors.
transmission happen in all stage for instance sensors, The figure 2 shows the types of attacks at the
RFID tags [10], processors to controllers, devices, DCs component level in IoT.
to processing units, and end users. In the fifth stage,
delivery of processed data to things on time without
errors or transformation is a sensitive task that must
always accomplish.

III. INTERNET OF THINGS (IoT) THREATS

The IoT devices have several applications that are


designed to make life convenient and effortless.
Think of engineers being able to access a device,
perform remote diagnosis and remediation any
problem. This is after the device has informed the
engineering team of an imminent problem before it Figure 2. The Types of Attacks at the Component
becomes a major problem [11]. As you can predict, Level in Internet of Things (IoT)
with this data exchange over the internet come

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


464
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

3.2 The Stage Based Incursion needed to make sure connected devices on the IoT
can be trusted to be what they intend to be.
The diversify kind of attacks on the five stages of IoT. Accordingly, each IoT device needs a distinctive
Data effusion, sovereignty, [14] violation, and identity that can be authenticated when the device
authentication are the major concerns in the data effort to connect to a main server or gateway. With
perception phase. this distinctive identity in place [15], IT system
administrators can track every device throughout its
3.2.1 Data Effusion lifecycle, communicate securely with it, and inhibit it
from executing detrimental processes. If a device
The data is a precious resource, and if recent events shows unforeseen behavior, administrators can
surrounding the allegations against (for instance, simply revoke its perquisite [16]. In addition, it is
Cambridge Analytica) demonstrate anything, it is compulsory to verify that the data do not change
that you cannot always trust companies to manage during transit. Data authentication could endow
such data responsibly. Specifically, device originality and integrity.
manufacturers are in a prime position to potentially
abuse [14] IoT generating data. Data effusion can be 3.2.4 Data Mislay
internal or external, intentional or unintentional,
authorized or spiteful, involving hardware or Data mislay considered a primus risk for the Internet
software. Data effusion is a solemn threat to of Things. The data mislay dissimilar from data
reliability. As the cloud data move from one renter to effusion in that the latter is a sort of revenge taking
several other renter of the cloud, there is a sedate risk action on the employer or administrator. Data mislay
of data effusion. is losing the work accident due to hardware or
software, lack of success and natural calamity.
3.2.2 Data Sovereignty
3.2.5 Attack on Availability
Internet of Things (IoT) connections growing from 9
billion in 2017 to 24 billion by 2020, it is vital to The Internet of Things continues to increase, which
understand the potential influence data sovereignty in turn spread your organization’s attack surface.
will play for organizations collecting and sharing Many types of attacks have been on every side for a
customer’s individual information. International very long time. What is new is the scale and relative
organizations are now often moving big data across naivety of attacks in the Internet of Things the
borders for analysis and consolidation [2]. Data millions of devices that are a potential sufferer to
sovereignty is the idea that information which has traditional style cyber-attacks, but on a much huge
been transformed and stored in binary digital form is scale and often with limited, if any shield [17]. IoT is
subject to the laws of the country in which it is all about connecting and networking devices that up
located. The IoT encloses all things across the globe until now not on a mandatory basis been connected.
and is hence responsible to sovereignty. This means that all of those devices, whether it is
your new connected sensor or your connected
3.2.3 Data Authentication vehicle, are creating a new entry point in the
network and here upon posing an increasing security
The data can be perceived from any device at any and privacy hazard. In this context, denial of service
moment. The powerful IoT device authentication is (DoS) attack happens when a service that would

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


465
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

usually work is not available. There can be many of vicinal gets formed and these new vicinal [19] are
factors for unavailability, but it generally refers to an all from other end of wormhole tunnel and not in
infrastructure that cannot cope due to capacity transmission range of node therefore, during the
overload [18]. In a Distributed Denial of Service attack lots of control packets are going to exchange
(DDoS) attack, a huge number of systems spitefully from one end of the tunnel to the other in that
attack one target. This is often done via a botnet, vicinal advertisement. Wormhole attack is very
where many devices are programmed to entreaty a strange and arduous to identify. The figure 3 shows
service at the same time. the wormhole attack on IoT.

3.3 The Architecture Based Incursion

The IoT will generate data at various locations for


various end users, including the enterprise, its
subscriber and partners, network segmentation and
segment-based topologies are expected to protect
against extensively attacks. The various vendors and
applications adopt their own layers. Now, we are
discussing the possible threats to each layer in IoT. Figure 3. The Wormhole Attack in Internet of
Things
3.3.1 Exterior Attack
3.3.3 Brute-Force Attack
The IoT is a bit of a buzzword right now. It is a used
as a catch all for everything that layers physical A brute force attack is the cyber-attack compeer of
devices like as computing infrastructure, sensors, trying every key on your key ring, and in conclusion
networking, storage, application [14] capabilities. The finding the right one. This type of attack depends on
organizations purposely offload both confidential and a trial and error technique in order to get information
non-confidential data to obtain the services. like as user passwords or personal identification
However, they do not know of the location where number (PIN). The brute force attack uses [20]
their data will be processed or stored. It is possible automated software to generate a large number of
that the provider may share this information with sequential guesses to decrypt the cipher text. The
others, or the provider itself may use it for spiteful criminals to crack encrypted data may use brute force
actions. attacks.

3.3.2 Wormhole Attack 3.3.4 Man-in-the-Middle Attack

The wormhole attacks considered as the grievous The man-in-the-middle concept is where an attacker
attacks during IoT routing. In this attack, tunnel is or hacker is looking to intervene and violation
established between two nodes and the packet is communications between two distinct systems. It can
forwarded among each other. These distant spiteful be a hazardous attack because it is one where the
nodes make believe that they are very close to each attacker secretly intercepts [21] and transmits
other so that vicinal nodes forward packets via them. messages between two parties when they are under
If wormhole attack is triggered in the more number

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


466
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

the faith that they are communicating directly with


each other. As the attacker has the genuine The impersonation is a threat in which malicious
communication, they can trick the recipient into nodes alter the data flow route and tempt the nodes
thinking they are still getting a rightful message. to the wrong positions. In Sybil attack, a malicious
Man-in-the-middle attacks can be succinct in many user dissimulates to be a separate user after acquiring
ways, including MITM, MitM, MiM and MIM. multiple identities and tries to create a relationship
with a truthful user. If the malicious user is successful
3.3.5 Sinkhole Attack in compromising one of the truthful users, the
attacker gains unauthorized privileges that help in
The sensors, which are leave unattended in the the attacking process [24].
network for long periods, are mainly susceptible to
sinkhole attack. The compromised node attracts the 3.3.9 Social Engineering Attack
information from all the neighboring nodes. Thereby,
the intruder posts other attacks, [21] such as selective Social engineering is the act of manipulating people
forward, fabrication, and transformation. so they give up secret information [25]. The types of
information that criminals are seeking can vary, but
3.3.6 Botnet Attack when individuals are targeting, the criminals are
usually an effort to deceive the user into giving them
A botnet is a collection of Internet-connected passwords or bank information. Alternatively, they
computers whose security defenses have been infarct could be effort to access a computer in order to
and control ceded to a spiteful party. Each such stealthily install malicious software that will then
compromised device, known as “bot” is created, give them access to confidential information, as well
when a computer is infiltrated by [22] software from as giving them control over the computer.
a malware distribution, else cognize as malicious
software. The controller of a botnet is able to direct 3.3.10 Hello Flood Attack
the concern with of these compromised computers
via communication channels formed by standards- In Hello flood news attacks, every object will
based network protocols such as hypertext transfer familiarize with Hello messages to all the vicinal that
protocol (http) and internet relay chat (IRC). are reachable at its frequency level. A malicious node
will cover a wide frequency area, and hence it
3.3.7 Side-Channel Attack becomes a vicinal to all the nodes in the network.
Thereupon, this malicious node will also broadcast a
Most IoT objects, for security intent, will be Hello message to its entire vicinal, make an
integrated with some of security mechanisms such as impression the availability. Flooding attacks cause
an encryption to protect their confidential data. The non-availability [26] of resources to rightful users by
side channel attack, [23] is intended to break such distributing a large number of nonsense requests for
mechanisms by analyzing side channel information a few services.
emitted by IoT objects. In this context, power and
time analysis attacks are some examples of such type 3.3.11 Data Insertion Attack
of attacks.
During the process of transposing data transmitted
3.3.8 Sybil Attack between two objects equipped with NFC protocol, an

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


467
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

attacker could insert some data into this data only, if


the object needs a long time to reply [27]. The well-
turned insertion could only happen if the inserted
data can be conveyed, before the original device
starts with the answer. If both data streams overlap,
the data will be unserviceable.

3.3.12 Flash Crowd Attack

A flash crowd is fundamentally an unexpected


increase in the overall traffic to any specific web page
or website on the Internet and the unexpected Figure 4. The IP Spoof Attack
occurrence of any event that triggers that particular
massive traffic of people accessing that web page or 3.3.14 Eavesdropping Attack
website. Less robust sites are unable to cope with the
large increase in traffic and become not available Eavesdropping is obstructing of network traffic to
[14]. The general causes of flash crowd or lack of gain prohibited access. It can outcome in lack of
sufficient data bandwidth, servers that collapse to success of confidentiality. The man in the middle
cope with the huge number of requests, and traffic attack is also a category of eavesdropping.
quotas. Eavesdropping attack observes the packets
transmitted between objects and servers during the
3.3.13 IP Spoof Attack firmware [29] upgrade process. The attacker could
only get confidential data if the packets are either
Spoofing is a type of attack in which the attacker weakly keeping safe or not keep safe at all.
dissembles to be someone else in order to gain access
to prohibit resources or thieve information. This type 3.3.15 Object Tampering Attack
of attack can take a [28] diversify of different forms,
for example; an attacker can act the IP address of a The likelihood of accessing IoT objects physically by
authorize user to get into their account. IP address attackers is very high because some IoT objects may
spoofing, or IP spoofing, refers to the creation of IP bring into service in stoic environments.
packets with a counterfeit source IP address, called Consequently, such objects are vulnerable to
spoofing, with the purpose of concealing the identity hardware attack, the most [30] renowned ones are
of the sender or imitate another computing system. the extraction of cryptography keys, the modification
The figure 4 shows the IP Spoof attack. of operating system or firmware, and the circuit
transformation.

IV. INTERNET OF THINGS (IoT) SCHEME

The IoT is a computing idea that depicts the thought


of everyday physical objects being related to the
internet and being able to identify themselves to
other devices. It enables multifarious physical objects

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


468
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

to share information and coordinate decisions. In spite


of, the functionality and operations of the IoT heavily 4.3 Friend Relationship Sybil Scheme
depend on the underlying network connectivity
structure [1]. The IoT characteristic ubiquitous In a mobile network, due to the mobility and the
communication among all kinds of electronic devices, deficiency of global social graph information, Sybil
it inevitably raises security concerns due to seamless defense is completely different and arduous
infiltration and automated integration among all sorts compared with that in the online networks [24].
of applications. Consequently, efficient and effective Quercia and Hailes propose an MSD scheme to match
defense mechanisms are of the paramount importance mobile user communities and label the users from the
to make sure the reliability of the IoT [31]. Sybil community as Sybil attackers [35]. In, one
imagination is that each mobile maintains two lists,
4.1 Sequential Scheme
first familiar list containing the trusted mobile users,
and second foe list with the unfaithful users in it. In
A sequential defense scheme is proposed by that
view of the fact that two users are encountered in the
sequentially collects feedbacks from high degree
network, they match their communities. If a user
nodes for attack presumption. The benefit of
were not in the trusted communities, this user would
sequential defense is that there is no necessity to
be reckoned as a Sybil user. In [36], Chang et al. also
acquire feedbacks from all nodes, and it abolishes the
propose a Sybil defense scheme in MSNs, assuming
collection process once sufficient feedbacks have
that the Sybil users and normal users happen in
been collected for attack presumption. The huge
various communities, and rely on the community
network size renders concurrent data transmissions
analogous to detect the Sybil users [24]. Therefore,
infeasible, particularly for wireless networks with
leveraging rapprochement is an effective solution to
scarce radio resources [32]. Besides, due to the huge
detect Sybil attackers. Although, this type of FR-MSD
network size and limited computational power,
schemes need mobile users to maintain the faithful
analyzing the collected information from all nodes
community information in advance.
incurs terrible computation overheads and it may
default to [33] provide timely safeguard. It is
4.4 Fusion Scheme
reaffirming in that a relatively small fraction of
feedbacks is enough to detect fatal attacks on the
A fusion-based defense mechanism is proposed to
network prior to network interruption.
deduce the presence of an attack [37] based on the
response from each node. The response information
4.2 Topological Scheme
can be as simple as a binary status report reflecting
that each node is, or is not, under attack, based on the
A topological defence scheme permits alteration of
node-level detection capacity [38]. The network level
network topology to enhance network resilience [34].
attack presumption plans are carried out at the fusion
It has been established in that by swapping a small
center [39]. A two-player game between the
number of edges in the network topology, one is able
protector and the attacker is naturally formed, given
to significantly make better network resilience
the critical value of network resilience and the node-
without including extra edges. Moreover, the
level detection configurations. Intuitively, from the
proposed edge rewiring technique in can be executed
adversary’s point of view, too few node elimination
in a distributed fashion, which is especially
cause hardly any harm to the network connectivity,
advantageous for the IoT due to scalability.
while too many node elimination is prone to be

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


469
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

detected by the fusion center, which means that the arduous endeavor. On the physical level, device
attack is eventually in futile. From the protector enclosures are frequently not tamper proof devices
point of view, inferring attacks using all feedbacks can be unlocked and their hardware can be accessed
might treat the topological attack as a fabled alarm, via investigate and pin headers [45]. The device
since only a small subset of nodes are targeted. On central processing units (CPUs) are low-cost
the other hand, inferring attacks using only a few components that frequently have no sophisticated
feedbacks might tolerate from information means to protect their code, data, and tokens for
insufficiency and therefore fabled to detect the external access. This permits an attacker to clone
presence of attacks [40]. Accordingly, there exists a entire devices or manipulate software and data. If the
balance point at which both attacker and defender device is brought into service in an unsupervised
are contented with their own master plan, which is environment, it may be accessed and frame up by a
precisely the notion of Nash equilibrium in game malicious third party without information. Besides,
theory. As an outcome, the game payoff at game IoT devices are frequently based on low-power
equilibrium can be used to study the robustness of a hardware and may only be able to process tokens
network. with a low complexity. This can have an implication
on the robustness of a token, as it can be
V. INTERNET OF THINGS (IoT) RELIANCE reengineered or recovered through a brute force
attack. As an outcome of this, any faith management
The Internet of Things (IoT) can deliver important system for IoT deployments must have the ability to
advantage to our society and economy, enabling dynamically withdraw faith of individual devices.
superior health services, cleaner and secure societies, When faith and security credentials given out at the
more dexterous industries and [41] stimulating new time of manufacturing or deployment, a device is
businesses. On the other hand, data is a sensitive seen as initially honest. A network wide update
subject and customers and end users are wary of mechanism will ideally incorporate a smooth and
sharing data for apprehension that it will fall into the effective patching process, which contains robust
hands of criminals, or using in ways they consider integrity and authenticity examine, minimize service
inappropriate [42]. Faith is the oxygen, which will outages, and permits for a version withdraw if
breathe life into the internet of things. Industry needed.
necessity to show data is secure and that it is properly
treated [43]. Faith and security are based on tokens or 5.2 Secure Key Storage
credentials, provided by a faith management
infrastructure, which are embedded in and The secure storage provision to increase the
potentially shared between devices. The aim of this robustness of reliance tokens used both within an IoT
section is to investigate frameworks that ensure faith system and within its reliance management
as well as communication security between nodes in infrastructure [46]. Passive keystores endow a means
an IoT deployment [44]. to securely save and retrieve credentials cryptographic
operations are carried out [10] outside these stores by
5.1 Reliance and Security from a Device Point of the device’s CPU. Active keystores in contrast, permit
View the internal execution of cryptographic operations
through an application program interface (API), so
IoT devices are vulnerable in many sides, so the credentials are never disclosed.
providing and maintaining faith and security is an

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


470
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

5.3 Reliance and Security from a Network Point of the number one users of the Internet and necessity to
View digital identities for secure operation. As enterprises
seek to transform their business models to stay
During the operation of a network, devices set up competitive, swift adoption of IoT technologies is
stable or unstable communication links with other creating increasing demand for Public Key
peers. These links can be either point-to point or Infrastructures (PKIs) to confer digital certificates for
incorporate a group of nodes. From a device’s point of the increase number of devices and the software and
view, the challenge [47] is to validate the authenticity firmware they run [49]. The secure IoT deployments
and authority of the other peer and to set up a secure need not only faith the devices to be authentic, but
communication link to keep away from attack also faith that the data they collect is real and not
scenarios. For this objective, reliance tokens are changed. If one cannot fault the IoT devices and the
interchange and validated or new session tokens bring data, there is no point in collecting, running
into being. The promise of data integrity, optionally analytics, and executing decisions based on the
in combination with data confidentiality through [48] information gathered.
encryption, provides trustworthiness in the data a
node sends or receives. Data integrity and 6.1 The PKIs Elements
confidentiality provide a foundation to deal with the
principal attack vectors of interception, A public key infrastructure confers the revocation,
transformation and disruption. The data secrecy is distribution, revocation and verification of public
generally conferred through symmetric encryption keys used for public key encryption, and enables
with the advanced encryption standard algorithm as a linking of identities with public key certificates. The
de facto industry standard often implemented directly public key infrastructure describes the policies, the
in hardware. While data integrity confers through procedure, the hardware, the software, and the
message authentication codes or cryptographic hashes people that are used to handle digital certificates [50].
that are attached to the data payload [48]. The peer This means the process of creating, distributing,
authenticity belonging to the problem of how a peer handle, storing, and revoking certificates are all
can validate another peer’s identity before a encapsulate under PKI. The PKI also mention to the
communication link is established. Peer authenticity associations we make with a public key to a person,
can go hand in hand with system availability. For or a public key to a device. It is based on faith, and a
instance, denial-of-service (DoS) way attacks are certificate authority creates this faith. This key
typically external attacks, so the ability to qualify and management lifecycle starts with the creation of a
if necessary to discard data or connection requests [14] key [51]. We make up one's mind on a distinctive
at an early stage can help to reduce such attacks. Proof strength of the key, which would be a certain
of authorization provides commitment that a peer has number of bits, and we make up one's mind what
the authority to communicate with another peer and cipher we would like to use to create the key. Now,
conduct a few actions. we are discussing the PKI elements.

VI. PUBLIC KEY INFRASTRUCTURES (PKIs) IN 6.1.1 The Certification Authorities


INTERNET OF THINGS
A believable party confers the root of faith for all PKI
At present, there are more things (devices) online certificates and confers services that can be used to
than there are people on the planet. The devices are authenticate the identity of individuals, computers

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


471
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

and other existence. Generally known as certificate typically through online certificate status protocol
authorities (CA), these entities provide promise about services.
the parties identified in a PKI certificate. Each CA
keeps going its own root CA, for use only by the CA. 6.1.4 The Central Directories
The certification authority browser forum, also called
a CA & browser forum, is an industry coalition The central directories protect location in which are
founded in 2005 and whose members contain [51] stored and index keys [50]. The central directories
CAs, browser software publishers and other system make certificates accessible to other entities. Since
contributor, who use X.509 digital certificates for other data, such as policies or CRLs necessity to be
authentication and encryption. The CAs forms the published as well, central directories store and make
backbone and the faith anchors of a PKI. They all these data accessible. They are frequently
problem certificates and, in many cases, repeal status executing as lightweight directory access protocol
data regarding the certificates they problem, and servers.
publish both types of products.
6.1.5 The Time Stamping Authorities
6.1.2 The Registration Authorities
When the date and time of the phenomena is
The registration authorities (RAs) act as the front end recorded, we say that it is time stamped. A digital
of certification authorities. A registration authority, camera will record the time and date of a photo being
often called a subordinate CA, problem PKI taken, a computer will record the time and date of a
certificates. The registration authorities are certified document being saved and emended. These are all
by a root certificate authority and authorized to issue instances of a timestamp. Timestamps are essential for
certificates for distinguished uses permitted by the keeping records on when information is being
root. That they are accountable for identifying and reciprocity, created, or destroyed online. In many
authenticating entities that request certificates, and situations, these records are simply utilitarian for us
then dispatching certificate requests to CAs and to know about. However, in some situations, a
routing back the certificate(s) to the implore entity. timestamp is more valuable. The time stamping
In some instance, RAs are just a unique component of authorities are characterized by their ability to
CAs. problem PKI based believable timestamps. Believable
time stamping is a process that keeps track of the
6.1.3 The Validation Authorities creation and alteration of data. This data can be a
program or a document. This process is done in a
A certificate database stores information about trouble free manner and recorded so that no one cans
controversy certificates. Therewith, to the certificate alteration the data, including the owner, without
itself, the database includes validity period and the being informed and it assurance the integrity of the
status of each PKI certificate. The validation data.
authorities (VAs) permit for the validation of
certificates. Validating a certificate in fact comprises 6.1.6 The Certificate Revocation Authorities
many steps for instance; possibly acquire certificates,
verifying signatures, checking the revocation status. A certificate revocation list (CRL) contains digital
It’s normally supposed that VAs only provides certificates that have been invalidated by the
services in connection to check revocation status, emanate certificate authority before their scheduled

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


472
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

expiration date and should no longer be believed. The The connected IoT endpoints share coequal security
CRLs is a type of blacklist and are used by different requirements [47]. A faithful or device identity,
endpoints, including web browsers, to confirm confirm applications and data secured in motion and
whether a certificate is valid and trustworthy. In at rest. These needs translate to authentication,
general, invalidate duties are carried out by a signature and encryption [52]. The industrial sector
dedicated service that belongs to each certificate has placed greater trust on digital certificates for IoT
authority. While either the number of issuing system elements such as controls, applications,
certificates is high, or the complexity of invalidate sensors, devices, switches, and data [23]. The
procedures increases, or so does the number and countless commercial operations, systems and
diversification [51] of CAs, specialized authorities, infrastructure are before controlled through the
CRAs, come to play, whereby a single, centralized Internet. Many devices that use the sensors and
CRA can replace equivalent invalidate services on actuators should follow particular policy and proxy
multiple CAs. When a web browser makes a rules for authentication to authorize the sensors to
connection to a site using TLS, the web server's public their information [53]. Meantime, low cost
digital certificate is investigated for anomalies or solutions in this field have not been conferred as
difficulty, part of this process involves investigating much as needed. At the present, if we want to confer
that the certificate are not catalogued in a certificate the security for the sensors we have to use high-cost
revocation list. These investigate are arduous steps in solutions, which is a dispute with the primary goal of
any certificate-based transaction because they permit IoT to provide lightweight protocols. The need for
a user to calibrate the identity of the owner of the particular commitment and related controls is greatly
site and discover whether the CAs still considers the increased for both industrial and consumer-driven
digital certificate believable. IoT devices. The security is very critical, and one of
the key distinctions between an enterprise PKI and
6.2 The IoT Challenges in PKIs an IoT PKI is the necessity for protecting data and
safe authentication at multiple endpoints. In this
PKI for IoT needs to be dissimilar than an enterprise context, digital certificates are an increasingly famous
PKI. The majority of devices collect transmit and has solution for signing, encryption and authentication.
at least one piece of private, confidential or
proprietary information [47]. Some devices will 6.4 Need the IoT Device Authentication
cross-unencrypted network and cloud services with
varying security levels and requirements. Some The powerful IoT device authentication is needed to
devices will be physically inaccessible. Other devices make sure connected devices on the IoT can be
may be attractive to impersonate in order to [14] gain believed to be what they purport to be. Accordingly,
access to IoT system resources. Comprehension the each IoT device needs a distinctive identity that can
following difficult situation can help organizations be authenticated, when the device effort to connect to
plan certificate policies upon which the PKI a central server. With this distinctive ID in place, IT
environment will be based. system administrators can track each device via its
lifecycle, communicate securely with it, and inhibit it
6.3 In the IoT Authentication, Integrity & from executing detrimental processes. If devices,
Confidentiality manifest unforeseen behavior, administrators can
simply revoke its prerogative.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


473
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

6.5 Need the IoT Device Connected Trouble Freely certificate policy the suitable amount of promise so
that relying parties know why the certificate is
To trouble freely participate in the IoT, each considered faith, as well as to understand what the
connected device needs a distinctive identity even device is authorized to do and not to do.
before it has an IP address. This digital credential
establishes the root of faith for the device’s entire VII. TERNET OF THINGS (IoT) DESIGN IDEA
lifecycle, from initial design to deployment to FOR DIGITAL CERTIFICATES
freedom from work. The every device a distinctive
identity using the powerful cryptographic processing, In a digital identity certificate, both its owner and the
key shielding, and key management available. A CA that signed the certificate must be distinctively
digital certificate is injected into every device enable identified. While there will be a comparatively small
to firstly, authentication of each device introduced to number of CAs, there is a need for a scalable naming
the organization’s architecture. Secondly, verification scheme appropriate for billions of nodes. In the design
of the integrity of the operating system and idea, device identifier establishment plans can be
applications on the device level. Thirdly, secure based on a different technique. These techniques
communications between devices, gateway, and incorporate either a hierarchical identifier, the
cloud. Lastly, authorized software and firmware encoding of extra information, unsystematic data, and
updates, based on approved code. the use of cryptographic operations [54]. In the design
idea, certificate validity the X.509 certificates have a
6.6 The PKI Security Guarantee in the IoT limited life span, which is enciphered in the validity
field [55]. The field accommodates the two date’s
The guarantee in the PKI space can be defined as the notAfter and notBefore, both accommodates a
amount of confidence that a person or system has timestamp in the UTCTime encoding format.
that the identity being introduced in a certificate in Investigation the validity of a certificate requires
fact be suited to the device. Maintaining specific access to actual time [56], and since low cost
levels of guarantee across millions of deployed device oscillators found in embedded, systems have an
identities is no trivial task. Risks must be important drift in the order of up to many. Seconds
comprehensible and mitigation plans established and per day, the use of time synchronization protocols
executed. The IoT system may incorporate millions of like network time protocol or precision time protocol
diverse connected devices sharing data to complete should be believed. In the design idea, the public key
various tasks, it makes sense that peril be inclined to cryptosystems confer pairs of keys, whereby the
be greater for an IoT system than for an enterprise public encryption key dissimilar from the secret
use case. The higher IoT peril’s profile, the PKI decryption key [51]. Such cryptosystems are at the
environment needed to issue believe certificates core of PKI, as they confer a means to digitally, the
across the system endpoints warrants very careful hash value of a digital certificate using a CA’s private
security planning to make sure a distinctly defined key; provide a means to confirm the integrity of a
level of promise for device authentication. The digital certificate, through decoding the already. The
reliability promise applies to both enterprise and IoT encoded hash value using a CA’s public key and
PKIs and various use cases may rely on varying levels comparing it with the hash value calculated over the
of promise. Some use cases need to define, detailed presented certificate; and permit a device to digitally
and strict promise requirements, while others may sign or decrypt information. In the design idea, hash
require less definition. Organizations must define in functions are one-way functions that modify a bit

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


474
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

string of variable length into a fixed-length hash value. scenarios, a number of technologies have been
They are utilizing to digital signs a certificate [57]. In developed to achieve information privacy and security
the hash, functions have four essential mathematical objective, such as transport layer security, which
and algorithmic characteristic, first they should have a could also make better the confidentiality and
small computational complexity, second irreversible integrity of the IoT.
“one-way” functions, third infeasible to alter an input
without changing the hash and fourth it must be 8.1 In IoT Secure Device Provisioning and
infeasible to find two dissimilar inputs with the same Authentication Using Azure
hash. There are a number of various future-proof hash
algorithms in use, most particularly SHA-2 and SHA- The security token technique provides authentication
3 with customizable hash lengths of between 224 and for every call made by the device to the IoT hub by
512 bits. associating the symmetric key to each call. The X.509-
based authentication permits authentication of an IoT
VIII. INTERNET OF THINGS (IoT) PROTOCOLS device at the physical layer as part of the TLS
FOR ACCESS NETWORKS connection establishment [60]. The security token-
based method can be used without the X.509
The Internet of Things (IoT) goal to make better our authentication, which is a less safe pattern. The
lives by increasing the interconnectivity of an preference between the two techniques is primarily
increased variety of embedded computing devices dictated by how safe the device authentication needs
using components of existing Internet infrastructure to be, and availability of safe storage of the device. IoT
[1]. This will permit for communications between hub uses security tokens to authenticate devices and
sensors in cars, laptops, factory, home appliances, services to keep away from sending keys on the
mobile phones, machineries, and many other devices network. Besides, security tokens are limited in time
that are already capable of network access through validity and scope. Azure IoT SDKs automatically
existing protocols such as 3G, Wi-Fi, Bluetooth [58], generate tokens without need any special
and ZigBee. The IoT scenarios remain a challenge, configuration [61]. Some scenarios need the user to
mainly due to the large number of miscellaneous generate and use security tokens outright. These
devices as well as data exchanged via insecure scenarios contain the direct use of the AMQP, MQTT,
connections. Moreover, the concepts of security are and HTTP surfaces, or the implementation of the
extended not only to device-to-device token service pattern.
communications, but also to network aspects. As an
example, many hackers create fake networks (termed After each IoT hub has a recognize registry that can
botnets) to steal data and user privacy information. be used to create per-device resources in the service,
Normally, various security requirements should be like as a queue that contains in flight cloud-to-device
addressed to promise network, and data security. First, messages, and to permit access to the device facing
confidentiality is essential to limit network access and endpoints. The IoT hub identity registry provides safe
data only to authorized users. Second, data integrity storage of device identities and security keys for a
and authentication should be promise so that solution. Independent or groups of device identities
messages are triumphantly transmitted and are can be added to a permit list, or a block list, enabling
reliable to the receiver. In the end, data complete control over device access. The use of a
authentication and availability should be provided, as device-based X.509 certificate and its associated
well as detection of malicious interloper [59]. In IoT private and public key pair permits extra

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


475
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

authentication at the physical layer. The private key close to the gateway and negotiator, propagation
is stored securely in the device and is not detectable delays can be ignored. In addition, for the
outside the device. The X.509 certificate contains unidirectional case, the security keys have a valid
information about the device, like as device time interval sufficiently long to transmit one or more
identification, and other organizational details. A packets and to absorb possible retransmissions or any
signature of the certificate is originated by using the other undesirable latency.
private key. Internet connection between the IoT
device and IoT hub is secured using the transport IX. GUIDELINES FOR SECURE THE INTERNET OF
layer security standard. Azure provides for IoT THINGS DEVICES
transport layer security standard, namely TLS 1.2,
TLS 1.1, and TLS 1.0, in this order. The Internet of Things is comprised of an
indiscriminately diverse range of device types from
8.2 The Secure Access to Unidirectional Data in IoT small to huge, from simple to complex, from
consumer gadgets to state of the art systems found in
The unidirectional devices cannot perform any safe DoD, [1] utility and industrial and manufacturing
procedure for secure key exchange with the systems. IoT devices face the same types of privacy
negotiator. The transmitter just sends a message and security problem that many traditional end-user
without any feedback, that is, it lapses to receive any devices face. There are approx six million new things
signal, and is equipped with an internal clock, which being connected every day in 2016, as we head
is supposed not to be on the mark. Then, a toward more than 22 billion by 2020, according to
nonspecific non-IP unidirectional terminal runs the Gartner [5]. End users do not have the technical
following move to send data to the gateway and specialist to assess the privacy and security
negotiator in a safe way. In the first move, it implications of any particular IoT device, or they may
generates the encryption key locally, based on the lack interest in doing so [63]. The subscribers already
time measured by a local clock. In the second move, have trouble identifying and troubleshooting the
it creates the message and encrypts it with the devices that are currently connected to their home
generated key, this message includes the payload and networks [64]. IoT devices will worsen these
any other data to be used to make better security. In circumstances, as subscribers connect an increasingly
the third move, it computes the hash values using the wide variety of devices to their home networks. The
message text and the generated key and attaches consumer will likely lose track of what devices are
them to the message. At the end move, it sends the connected to the Internet over time, which will make
message to the gateway and negotiator. defend them even more challenging. In this section,
we are discussing guidelines for secure the IoT
8.3 The Secure Access to Bidirectional Data in IoT
devices.

For bidirectional terminals, the negotiator can 9.1 Don’t Connect Your Devices Unless You Necessity
periodically broadcast its clock timing in a dedicated
message, and its identification in the plain part of the The first step is to consider what functionality you
message. The terminals can align their local clocks to necessity from the device. In view of the, your TV or
the gateway and negotiator terminal, and then fridge can connect to the internet, doesn’t mean you
generate the security keys in accordance with the surely want to hook it up. Take a good look at the
algorithm already described [62]. Since devices are

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


476
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

features it offers and learn precisely what internet sent through email or SMS before access is permitted.
connectivity brings before you connect. When used properly, two-factor authentication can
halt the bad people obtain access to your accounts
9.2 To Secure Communications Using Encrypted and taking control of your IoT devices.
Protocols
9.5 Using the Latest Firmware
Encryption practices of IoT devices are lower and
unsafe. A small number devices use encrypted If you want to make sure you have the latest security
communications as part of their beginning patches and diminish the possibility of a successful
configuration. Instead, most use ordinary web attack, then you need to keep your firmware fully
protocols that communicate across the Internet in updated. Vulnerabilities and exploits will be
plain text, which makes them simple targets for extricated as they emerge, so your IoT devices and
hackers keep an eye on network traffic to identify your router need to be frequently updated. Automate
debility. At the very least, all webs, traffic should be this wherever possible or set a schedule to investigate
using HTTPS [65], transport layer security, secure file for updates every two months or so.
transfer protocol, DNS security extensions, and other
secure protocols for communications with 9.6 During Processing Using a Encrypt Data
management stations and across the Internet.
Therewith, devices that connect to mobile apps or Infrequently the party processing the data should not
other remote gateways should use encrypted be able to read the data or the computational
protocols as well as encrypt data stored on flash outcome. The operational data while they are in
drives. encrypted form. For instance, identical encryption is
a form of encryption that permits computations to be
9.3 Block Incoming Traffic When Possible carried out on cipher-text, thus generating an
encrypted outcome that, when decrypted, matches
The several IoT devices ship with open ports to the outcome of operations performed on the plain
support management functions rather than the text.
standard functionality available through a user
interface. Even some passwords allow telnet access 9.7 Create Impressive and Inoffensive Password
with only an IP address. Afterwards, the point here is Policies
to decrease your attack surface as much as feasibly
possible. That might mean perfectly blocking all Most network infrastructure needs the administrator’s
incoming traffic with a firewall. However, in other default password to be altered when first accessed. In
cases, that will mean only keeping open which TCP spite of that, most devices, like as home routers,
and UDP ports you necessity. network printers, and sensors, lack strong
authentication and access technique. Furthermore,
9.4 Two-Factor Authentication the concept of using multifactor authentication using
a diversity of mechanisms to log in besides an easy
Supposing any of your devices offer two-factor
password, such as with an SMS code sent to a cell
authentication, use it. Two-factor authentication is
phone is a rarity in the IoT world. Actually, some IoT
an extra security layer on top of a device password
devices do not need any authentication. A subscriber
that need secondary authentication a one-time code
can navigate with a web browser to a specific IP

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


477
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

address and control the device’s configuration and read up on the provider’s privacy policy and look for
operation. commitment about encryption and data safety.

9.8 Investigate if Physical Access Assents Intrusion 9.12 Put IoT Devices on Their Personal Firewalled
and Monitored Network
It is obliged to understand how your attack surface
dissimilar in the case that a hacker is remote versus, When it comes to linked consumer grade IoT devices
when they are corporeally in the office location. in the enterprise, you need to take a proactive

There are a number of attached devices that are procedure. You want to have them segmented away
vulnerable subsequently doing a hard reset. If there and rear a firewall. You can block incoming traffic to
are any, consider locking them away, when feasible. it so people cannot attack from the inboard and you
can handle and monitor it closely.
9.9 Decrease Data Granularity
9.13 Disable UPnP Characteristics

The IoT applications should appeal the minimum


The IoT devices tend to have Universal Plug in and
level of granularity that is needed to perform their
Play (UPnP) characteristics, enabling various devices
most important tasks. A higher level of granularity
to explore and connect to one another. Whilst this is
could lead to secondary data usage and in conclusion
favorable and removes the need to configure each
privacy violations.
device individually, the protocols rely on local
9.10 Ameliorate Failover Design networks to connect to each other and these are
vulnerable to third party attackers.
The devices should function when Internet
connectivity is vanished or interrupted. However, 9.14 Conduct Risk Assessment
few IoT devices are designed to face with the lack of
success, such as Internet continuity or data After choosing the devices, check-up the network and

disconnections. Failover design is especially vital for its potential lapse points, as well as the IoT and Cloud
IoT devices that involve user protection, like as video platform used for handling and storage of data. Many
monitoring, door lock mechanisms, and simpler IoT devices have no computing power and
environmental monitors and alarms. These devices communicate with a gateway and the idea should be
should have manual overrides or particular functions given to choosing a remote monitoring service or IoT
for disconnected operations. service platform.

9.11 Careful of the Cloud Services X. CONCLUSION


In the time to come, every object in our daily life will
Many IoT devices rely on cloud services, but the need be connecting to the Internet. In this context, mobile
for an internet connection in order for something to phones will be used as the center point or the remote

function can be a real difficulty. Not only will it not control for all objects in the physical world commonly
work when the network is very slow, but it may also called as IoT. The Internet of Things opens up a new
be synchronized sensitive data or offering another universe of connected and intelligent devices that can
potential route into your home and make sure you work together to provide virtually unlimited
capabilities, and the majority of these new capabilities

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


478
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

will be personalized. Much of the value of the IoT Technology,” International Transaction of
comes from the capability to customize products and Electrical and Computer Engineers System
services to a client individually and immediately, (ITECES), USA, ISSN (Print): 2373-1273 ISSN
necessity. The analysts have a prophecy that (Online): 2373-1281, Vol. 4, No. 1, page 26-38,
hundreds of thousands of new IoT services will May 2017, DOI: 10.12691/iteces-4-1-4.
connect billions of new IoT devices over the next [4]. Luigi Atzori, Antonio Iera, and Giacomo
decade. Industry and academia are both Morabito. The internet of things: A survey.
concentrating on moving ahead in attempts to Computer Networks, 54(15):2787–2805, 2010.
improve usability, maintainability, and security via [5]. Gartner Inc. Press Release (2014)
standardization and development of best practices. http://www.gartner.com/newsroom/id/2905717
The Internet of Things (IoT) embodies the [6]. G. Wu, S. Talwar, K. Johnsson,N. Himayat, and
convergence of the physical and virtual worlds. It is K. D. Johnson, “M2M: from mobile to embedded
the important nexus between data-oriented internet,” IEEE Communications Magazine, vol.
applications and device-oriented sensor networks 49, no. 4, pp. 36–43, 2011.
facilitated by Internet technologies. The biggest [7]. L. Atzori, A. Iera, and G. Morabito, “The
challenges to the IoT will come in securing sensitive internet of things: a survey,” Computer
information from unauthorized access as well as Networks, vol. 54, no. 15, pp. 2787–2805, 2010.
authorizing access to only the information we're [8]. Gartner Inc., “Hype cycle for the internet of
comfortable disclose. This paper provides an things 2017,” Technical report, July 2017.
overview of Internet of Things (IoT) threats, scheme [9]. V. Karagiannis, P. Chatzimisios, F. Vazquez-
and faith. Afterwards, we are briefly discussing Gallego, J. Alonso-Zarate, "A survey on
public key infrastructures and guidelines for secure application layer protocols for the internet of
the IoT devices. things", Transaction on IoT and Cloud
Computing, vol. 3, no. 1, pp. 11-17, 2015.
XI. REFERENCES [10]. Xiaolin Jia, Quanyuan Feng, Taihua Fan,
Quanshui L. , ”RFID technology and its
[1]. Yusuf Perwej, Mahmoud A. AbouGhaly, Bedine applications in Internet of Things (IoT)”, 2nd
K. Hani Ali M. Harb, “An Extended Review on International Conference on Consumer
Internet of Things (IoT) and Its Promising Electronics, Communications and Networks
Applications”, Communications on Applied (CECNet), IEEE, Yichang, China, April 2012
Electronics (CAE), ISSN: 2394-4714, Foundation [11]. Nasser S. A., Andrew J., Olga A.,” Internet of
of Computer Science FCS, New York, USA, Things Security: A Review of Risks and Threats
Volume 9, Number 26, Pages 8– 22, Feb 2019, to Healthcare Sector ”, IEEE International
DOI: 10.5120/cae2019652812 Conference on Internet of Things (iThings) and
[2]. Yusuf Perwej, “An Experiential Study of the Big IEEE Green Computing and Communications
Data,” International Transaction of Electrical (GreenCom) and IEEE Cyber, Physical and
and Computer Engineers System (ITECES), USA, Social Computing (CPSCom) and IEEE Smart
ISSN (Print): 2373-1273 ISSN (Online): 2373- Data (SmartData), Exeter, UK, June 2017
1281, Vol. 4, No. 1, page 14-25, March 2017, [12]. Wei Z., Yan Jia., Anni P., Yuqing Z., Peng
DOI:10.12691/iteces-4-1-3. L.,”The Effect of IoT New Features on Security
[3]. Nikhat Akhtar, Firoj Parwej, Dr. Yusuf Perwej, and Privacy: New Threats, Existing Solutions,
“A Perusal of Big Data Classification and Hadoop

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


479
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

and Challenges Yet to Be Solved”, IEEE Internet [22]. Elisa Bertino, Nayeem Islam,” Botnets and
of Things Journal, June 2018 Internet of Things Security ”. Computer, IEEE,
[13]. J. Gubbia, R. Buyya, S. Marusic, M. Palaniswami, Volume 50 , Issue 2, Feb 2017
"Internet of Things (IoT): A vision architectural [23]. A. Mohsen Nia and N. K. Jha, “A Comprehensive
elements and future directions" in Future Study of Security of Internet-of-Things,” IEEE
Generation Computer Systems, Elsevier, vol. 29, Transactions on Emerging Topics in Computing,
pp. 1645-1660, 2013. vol. PP, no. 99, p. d, 2016.
[14]. R. H. Weber, "Internet of Things New security [24]. Kuan Zhang, Xiaohui Liang, Rongxing Lu,
and privacy challenges", Computer Law & Xuemin Shen,” Sybil Attacks and Their Defenses
Security Review, vol. 26, no. 1, pp. 23-30, Jan. in the Internet of Things”, IEEE Internet of
2010. Things Journal , Volume 1 , Issue 5 , Oct. 2014
[15]. M. O. Lehtonen, F. Michahelles, E. Fleisch, [25]. Ian G. Harris,” Social Engineering Attacks on the
"Trust and Security in RFID-Based Product Internet of Things ”, September 7, 2016
Authentication Systems", IEEE Systems Journal, [26]. L. Atzori, A. Iera, G. Morabito, "The Internet of
vol. 1, no. 2, pp. 129-144, Dec. 2007. Things: A survey", Comput. Netw., vol. 54, no.
[16]. Yusuf Perwej, Kashiful H., Uruj J., Firoj Perwej, 15, pp. 2787-2805, 2010.
“Block ciphering in KSA, A major breakthrough [27]. E. Haselsteiner and K. Breitfuß, “Security in
in cryptography analysis in wireless networks” Near Near Field Communication (NFC)
International Transactions in Mathematical Strengths,” Semiconductors, vol. 11, no. 71, p.
Sciences and Computer, India, ISSN-0974-5068, 71, 2006.
vol. 2, No. 2, pages 369-385, July-December [28]. Seo JW, Lee SJ,”A study on the detection of
2009 DDoS attack using the IP Spoofing”, J Korea Inst
[17]. G. Gan, Z. Lu, J. Jiang, "Internet of Things Inf Secur Crytol 25(1):147-153, 2015
Security Analysis", 2011 International [29]. D. Miessler, “Securing the Internet of Things:
Conference on Internet Technology and Mapping Attack Surface Areas Using the
Applications, pp. 1-4, Aug. 2011. OWASP IoT Top 10.
[18]. I. Gudymenko, K. B. Pfitzmann, K. Tietze, [30]. G. Hernandez, O. Arias, D. Buentello, and Y. Jin,
"Privacy implications of the internet of things" “Smart Nest Thermostat : A Smart Spy in Your
in Constructing Ambient Intelligence, Springer, Home,” Black Hat USA, pp. 1–8, 2014.
pp. 280-286, 2012 [31]. Pin-Yu Chen, Shin-Ming Cheng, and Kwang-
[19]. Marianne Azer, Sherif El-Kassas, Magdy El- Cheng Chen. Optimal control of epidemic
Soudani, “A full of the Wormhole Attack”, information dissemination over networks. IEEE
International Journal of Computer Science and Trans. Cybern., 44(12):2316–2328, December
Information Security, 2009. 2014.
[20]. Kelly Jackson, “Hacker’s Choice: Top Six [32]. Pin-Yu Chen and Shin-Ming Cheng. Sequential
Database Attacks. defense against random intentional attacks in
[21]. Ş. Okul ; M. Ali Aydın,” Security Attacks on IoT complex networks. Phys. Rev. E, 91:022805,
”, International Conference on Computer February 2015.
Science and Engineering (UBMK), IEEE, [33]. Shan C, Hui X, Da L, et al. A Vision of IoT:
Antalya, Turkey, Oct. 2017 Applications, Challenges, and Opportunities
With China Perspective{J}. IEEE Internet of
Things Journal, 1(4), 349-359, 2014

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


480
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

[34]. Pin-Yu Chen and Alfred O. Hero. Assessing and [44]. J. Guo, R. Chen, "A Classification of Trust
safeguarding network resilience to nodal attacks. Computation Models for Service-Oriented
IEEE Commun.Mag., 52(11):138–143,November Internet of Things Systems", IEEE International
2014. Conference on Services Computing (SCC), pp.
[35]. D. Quercia and S. Hailes, “Sybil attacks against 324-331, June 2015.
mobile users: Friends and foes to the rescue,” in [45]. Ie yuan, Xiaoyong li,” A Reliable and
Proc. IEEE IEEE Conf. Comput. Commun. Lightweight Trust Computing Mechanism for
(INFOCOM), 2010, pp. 336–340. IoT Edge Devices Based on Multi-Source
[36]. W. Chang, J. Wu, C. Tan, and F. Li, “Sybil Feedback Information Fusion”, special section
defenses in mobile social networks,” in Proc. on security and trusted computing for industrial
IEEE Conf. Global Commun. (GLOBECOM), internet of things, volume 6, May 16, 2018
2013, pp. 1–6. [46]. P. Porambage, C. Schmitt, P. Kumar, A. Gurtov,
[37]. Pin-Yu Chen and Kwang-Cheng Chen. M. Ylianttila, "Two-phase authentication
Intentional attack and fusion-based defense protocol for wireless sensor networks in
strategy in complex networks. In Proc. IEEE distributed IoT applications", Proc. IEEE
Global Telecommunications Conference Wireless Commun. Netw. Conf. (WCNC), pp.
(GLOBECOM), pages 1–5, December 2011. 2728-2733, Sep. 2014
[38]. Pin-Yu Chen, Shin-Ming Cheng, and Kwang- [47]. E Vasilomanolakis, J Daubert, M Luthra, V
Cheng Chen. Information fusion to defend Gazis, A Wiesmaier, P. Kikiras, "On the Security
intentional attack in internet of things. IEEE and Privacy of Internet of Things Architectures
IoT-J., 1(4):337–348, August 2014. and Systems", 2015 International Workshop on
[39]. J. Tian, W. Zhao, R. Du, and Z. Zhang, “A New Secure Internet of Things (SIoT), pp. 49-57, Sep
Data Fusion Model of Intrusion Detection- 2015
IDSFP,” in Parallel and Distributed Processing [48]. Chen, D., Chang, G. R., Sun, D. W., Li, J. J., Jia.
and Applications, vol. 3758 of Lecture Notes in J., and Wang. X. W.,”TRM-IoT: A Trust
Computer Science, pp. 371–382, Springer Berlin Management Model Based on Fuzzy Reputation
Heidelberg, Berlin, Heidelberg, 2005. for Internet of Things”, Computer Science and
[40]. Martin Osborne and Ariel Rubinstein. A Course Information Systems.,1207-28, October 2010
in Game Theory. MIT, Cambridge, MA, 1999. [49]. Michael Schukat, Pablo Cortijo,” Public key
[41]. S. Sicari, A. Rizzardi, L. A. Grieco, A. Coen- infrastructures and digital certificates for the
Porisini, "Security privacy and trust in Internet Internet of things”, 26th Irish Signals and
of Things: The road ahead", Computer Systems Conference (ISSC),IEEE, Carlow,
Networks, vol. 76, pp. 146-164, Jan. 2015. Ireland, June 2015
[42]. T. Eder, D. Nachtmann, D. Schreckling, "Trust [50]. Zhiyong Zhang, Qingqi Pei, Jianfeng Ma, Lin
and Reputation in the Internet of Things", Yang, "Security and Trust in Digital Rights
Conference Seminar (SS2013) - Real Life Management: A Survey", International Journal
Security (5827HS), Dec. 2013. of Network Security, vol. 9, no. 3, pp. 247-263,
[43]. Z. Yan, P. Zhang, A. V. Vasilakos, "A survey on 2009.
trust management for Internet of Things", [51]. Younggyo Lee, Jeonghee Ahn, Seungjoo Kim,
Journal of Network and Computer Applications, Dongho Won, "A PKI System for Detecting the
vol. 42, pp. 120-134, Jan. 2014. Exposure of a User's Secret Key Public Key

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


481
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

Infrastructure Springer Berlin, Heidelberg", vol. [60]. Borislav S. Đorđević, Slobodan P. Jovanović,
4043, pp. 248-250, June 2006. Valentina V. Timčenko, ” Cloud Computing in
[52]. Yun-kyung Lee, Hong-il Ju, Jee-hye User, Amazon and Microsoft Azure platforms:
authentication mechanism using authentication Performance and service comparison ”, 22nd
server in homnetwork Advanced Telecommunications Forum Telfor
Communication Technology, pp. 504-506, 2006. (TELFOR),IEEE, Belgrade, Serbia, Nov. 2014
[53]. Gianmarco Baldini, Trevor Peirce, Maria Chiara [61]. R. Tudoran, A. Costan, G. Antoniu, L. Bougé, "A
Tallachini, "Internet of Things: IoT Governance performance evaluation of azure and nimbus
European Research Cluster on the Internet of clouds for scientific applications", Proceedings of
Things, Jan. 2014 the 2nd International Workshop on Cloud
[54]. M. Bauer, P. Chartier, K. Moessner, Catalogue of Computing Platforms, pp. 4, 2012.
IoT Naming, Addressing and discovery schemes [62]. R. Giuliano, A. Neri, and D. Valletta, “End-to-
in IERC projects V1.7, IERC-AC2-D1, 2013. end secure connection in heterogeneous
[55]. AS. Wazan, R. Laborde, F. Barrère, A. Benzekri, networks for critical scenarios”, WIFS 2012,
"Validating X. 509 certificates based on their Proc. of the 2012 IEEE Intl. Workshop on
quality", Young Computer Scientists 2008. Information Forensics and Security, pp. 264–
ICYCS 2008. The 9th International Conference 269, Tenerife, Spain.
for IEEE, pp. 2055-2060, November 2008. [63]. Ka-Ping Yee, "Aligning security and usability."
[56]. J. Shannon, H. Melvin, A. G. Ruzzelli, Dynamic IEEE Security & Privacy 2.5, pp 48-55, 2004
flooding time synchronization protocol for [64]. Rebecca E. Grinter, et al., "The work to make a
WSNs, IEEE GLOBECOM, 2012. home network work." ECSCW 2005. Springer
[57]. P. Camion, J. Patarin, "The knapsack hash Netherlands, 2005
function proposed at Crypto'89 can be broken", [65]. A. P. Castellani, M. Gheda, N. Bui, M. Rossi, and
Adv. in Cryptology Proc. Eurocrypt'91, pp. 39- M. Zorzi, "Web Services for the Internet of
53, 1991. Things through CoAP and EXI," in IEEE RWFI,
[58]. Yusuf Perwej, Kashiful H., Uruj J., Sharad S., Kyoto, Japan, Jun. 2011
“Some drastic improvements found in the
analysis of routing protocol for the Bluetooth Cite this article as :
technology using scatternet” International
Conference on Computing, Communications Dr. Yusuf Perwej, Firoj Parwej, Mumdouh Mirghani
and Information Technology Applications Mohamed Hassan, Nikhat Akhtar, "The Internet-of-
(CCITA-2010), Ubiquitous Computing and Things (IoT) Security : A Technological Perspective
Communication Journal (UBICC) Seoul, South and Review ", International Journal of Scientific
Korea, ISSN Online 1992-8424, ISSN Print 1994- Research in Computer Science, Engineering and
4608, Volume CCITA-2010, Number 5, pages Information Technology (IJSRCSEIT), ISSN : 2456-
86-95, 2010 3307, Volume 5 Issue 1, pp. 462-482, January-
[59]. J. Granjal, E. Monteiro, J. Sá Silva, "Security for February 2019. Available at doi :
the Internet of Things: A survey of existing https://doi.org/10.32628/CSEIT195193
protocols and open research issues", IEEE Journal URL : http://ijsrcseit.com/CSEIT195193
Commun. Surveys Tuts., vol. 17, no. 3, pp. 1294-
1312, 3rd Quart. 2015

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


482
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195199
Secure Micro Payment Using Physical Unclonable Function
1
M. Jalasri, 2S. Nalini, 3N. Magesh Kumar, 4J. Elumalai
1,2,3,4
Assistant Professor, Information Technology, Jeppiaar Maamallan Engineering College, Sriperumbudur,
Tamil Nadu, India
ABSTRACT

In online aggressors regularly go for taking such client information. In these situations, malware that can take
card information when they are perused by the gadget has thrived. This framework gives secure micro-
payment and enhances date approaches in terms of security. In proposed system, utilize two principle capacities
PUF (Physical Unclonable Function) and FRoDo(Fraud Resilient Device for Off-Line Micro-Payments). These
two can be given security to installment process and client account. These produce double encryption to
protect the system from attackers. In proposed system gives more security for transaction.
Keywords : Mobile Secure Payment, Architecture, Protocols, Advance Encryption, Fraud-Resilience

I. INTRODUCTION identification and authentication but very difficult to


exchange key to card issuers [3].
Mobile commerce is one of the future research areas,
especially for mobile payment systems. In the FRoDO highlights a character component to confirm
existing system, the payments are performed on the the client, and a coin component where coins are not
fixed cellular network which degrades security. To privately put away, but rather are registered on-the
enhance security on transaction and payment process fly when required.
by performing e-payment systems in the distributed
development of the wireless ad-hoc network [1]. Payment transaction does not directly read customer
coins. The vendor only communicates with the
In Modern cryptographic, only authorized parties can identity element in order to identify the user.The
acquire secret keys and access information. Be that as main benefit is a simpler, faster, and more secure
it may, different sorts of altering techniques have interaction between the involved actors/entities. The
been concocted to separate secrecy keys from identity element used to improve the security of the
restrictive access frameworks, for example, users. This provide more secure off-line payments.
smartcards and ATMs. This server authenticates the .
PUF by two kinds of error probabilities such as false II. RELATED WORK
alarm rate (FAR) and false detection rate (FDR) [2].
A TPM (Trusted Platform Module) is a standard for
The main goal of this paper is to introduce the variety securing hardware device with the help of
of devices for influence magnetic stripe card data and incorporated cryptographic keys. Here identify
demonstrate magnetic stripe cards by fraudulently attacks only in memory and bus probing not real
cloned and analyze how such devices can be word attacks [4].
processed. This technique used for personal

CSEIT195199 | Received : 12 Feb 2019 | Accepted : 25 Feb 2019 | January-February -2019 [ 5 (1) : 483-486 ]
483
M. Jalasri et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 483-486

Near Field Communication (NFC) technology is a A. CLIENT MODULE:


short-range wireless communication between the
Fig. 2. Client module. User is registered first before
devices. NFC used in credit card, e-ticket, and mobile
login. User and Vendor are waiting for activation.
payment. It provides security for m-payment systems
Customer gives the payment request after Activation.
and also eases of use for the user. The main issue in
Users are activated by Frodo. Gives bank name,
this system is unable to receive phone calls when
permanent card number and its pin number to
processing payment operation [5].
registration. User Details are Stored in Database.

Symmetric key cryptography which reduces


computation at all the entities of the protocol.
Kungpisdan et al and Tellez said this security
protocol satisfies the existing standard protocols like
SET. But lack of complexity produced during
transaction and identification of users [6]. However,
the solutions are limited in that protocols require at Fig. 2. Client module
least one of the two parties to be online, i.e.
connected either to a trusted third party or to a B. VENDOR MODULE:
shared database.
In fig. 3 Vendor is registered first before login.
Vendor Details are Stored in Database. Vendor is
FORCE (Fully off-line secure credits for mobile
waiting for activation. Vendors get transaction after
micro payments ) exclusively perform the operations
activation. Users and Vendors are activated by Frodo.
at catching the topological impact introduced by the
Gives bank name, Branch and password to
wormhole. Detection of wormholes is successfully
registration.
done without any overhead or need of external
devices to monitor the network. But, difficult to
share database to trusted third party it didn’t notice
any real time attacks such as skimmers, scrapers and
data vulnerabilities [7].
Fig. 2. Vendor module
III. IMPLEMENTATION WORK
C. KEY GENERATION
Architecture Diagram for secure micro payment
using physical unclonable function in Figure 1. User Frodo view the user who gives request to activation
is registered first before login. Vendor is registered and Frodo activate the user. Identity element is
first before login. Vendor Details are Stored in generating private key to the user and coin element
Database. Vendor is waiting for activation. Vendors generate binary code for user account number into
get transaction after activation. Frodo view the user binary code. This two-step makes Payment more
who gives request to activation and Frodo activate secure. Key Generator used generate the private key
the user. After activation user login into account to of the identity element and symmetric and
gives payment request. After PUF verification asymmetric cryptographic algorithms applied to data
Payment transaction will success. received in input and sent as output by the identity
element in fig, 3.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


484
M. Jalasri et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 483-486

verification. After PUF verification Payment


transaction will success.

Fig. 3. Key Generation


Identity Element Coin Element

Vendor
FRoDo
Activate the
User

Generate
Private Key For Generate Binary
User Code for Account
FRoDo Login Number

Check Availability
Balance

Request Sent to
Fig. 4. Payment request
PUF

User
User Database User Login Enter the Payment Encrypted Cash
Registration
Amount

Vendor Login
In cryptography, (Advance Encryption Standard)AES
Payment Verification Identity and Coin
Element Verification Cash decryption
is broadly embraced and upheld in both equipment
Waiting for PUF Activate the
View Vendor
Verification Vendor

PUF Login and programming. Cryptanalytic attacks against AES


Payment Success

PUF
has not been discovered practically. A degree of
Fig. 4. Architecture Diagram for secure micro ‘future-proofing’ allowed by AES which has
payment using physical unclonable function. flexibility of key length.

The Rivest-Shamir-Adleman(RSA) public key In any case, similarly with respect to DES, the AES
cryptography algorithm which is used to encrypt a security is guaranteed just in the event that it is
message. The RSA algorithm performs public key effectively actualized and great key administration is
encryption and digital signatures. utilized.
1. Represent the message as a whole number in the
vicinity of 0 and (n-1). Huge messages can be E. PUF PROCESS:
separated into various squares. Each piece would then
be spoken to by a whole number in a similar range. Heart of FRoDO proposal lies a read-once strong
2. Encrypt the message by raising it to the eth physical unclonable function. In fig.5 PUF check
control modulo n. The outcome is a ciphertext vendors details involved in payment and activate the
message C. vendors.
3. To unscramble ciphertext message C, raise it to
another power d modulo n

The encryption key (e,n) is made open. The


unscrambling key (d,n) is kept private by the client.

D. PAYMENT REQUEST
Fig. 5. Puf process
After activation user login into account to gives
payment request in fig. 4. Enter the required In payment verification PUF check both Identity and
payment and Check the availability of payment for Coin Element. After verification payment decrypted
customer requirement. Required amount encrypted using public key and the payment transaction
using key and after payment request goes to PUF completed successfully and whole transaction will be

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


485
M. Jalasri et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 483-486

secure. Finally receives the Private Response, the last [2]. D. Lim, J. W. Lee, B. Gassend, G. E. Suh, M. van Dijk,
step only requires the coin just read to be validated. and S. Devadas, “Extracting secret keys from
Then, the whole payment transaction can be integrated circuits,” IEEE Trans. Very Large Scale
Integr. Syst., vol. 13, no. 10, pp. 1200–1205, Oct.
authorized and committed.
2005.
[3]. G. Hong and J. Bo, “Forensic analysis of skimming
IV. CONCLUSION
devices for credit fraud detection”, 2nd IEEE Int.
Conf. Inf. Financial Eng., Sep. 2010
Security analysis shows proposed system does not [4]. P. Choi and D. K. Kim, “Design of security enhanced
impose trustworthiness assumptions.Further, it is also TPM chipagainst invasive physical attacks”,IEEE Int.
the first solution in the literature where no customer Symp. Circuits Syst., pp. 1787–1790,2012.
device data attacks can be exploited to compromise [5]. W. Chen, G. Hancke, K. Mayes, Y. Lien, and J.-H.
the system This has been accomplished basically by Chiu, “Using 3G network components to enable NFC
utilizing a novel erasable PUF engineering and a mobile transactions and authentication”, IEEE Int.
Conf. Progress Informat. Comput., Dec. 2010, vol. 1,
novel convention outline. Furthermore, our proposal
pp. 441–448.
has been thoroughly discussed and compared against
[6]. V. C. Sekhar and S. Mrudula, “A complete secure
the state of the art. Our investigation demonstrates
customer centric anonymous payment in a digital
that it is the main suggestion that appreciates every ecosystem”, Int. Conf. Comput., Electron. Elect.
one of the properties required to a safe small scale Technol., 2012, pp. 1049–1054.
installment arrangement, while likewise presenting [7]. V. Daza, R. Di Pietro, F. Lombardi, and M. Signorini,
adaptability while thinking about the installment “FORCE: Fully off-line secure credits for mobile
medium (kinds of advanced coins). micro payments”, 11th Int. Conf. Security
Cryptography, 2014, pp. 125–136.
Finally, some open issues have been identified that [8]. S. Golovashych, “The technology of identification
and authentication of financial transactions. from
are left as future work. Specifically, examining the
smart cards to NFC-terminals”, IEEE Intell. Data
likelihood to enable computerized change to be spent
Acquisition Adv. Comput. Syst., Sep. 2005, pp. 407–
over various disconnected exchanges while keeping
412.
up a similar level of security and ease of use. In [9]. Slava Gomzin, “Hacking Point of Sale: Payment
FRoDO, develop to access money without standard Application Secrets, Threats, and Solutions,
coin element. In future, try to enhance the capability Computing & Technology 1st ed. New York, NY,
of the AES algorithm to a generate private key and USA: Wiley, 2014.
create more security for Electronic payments. [10]. Meng-Day Yu, Srinivas Devadas “Secure and robust
Another approach will involve smarter error correction for physical unclonablefunctions”,
implementation of secure offline transfer of money IEEE Design & Test of Computers , vol. 27, no. 1, pp.
48–65, Jan. 2010.
by using PUF.

Cite this article as : M. Jalasri, S. Nalini, N. Magesh Kumar,


V. REFERENCES
J. Elumalai, "Secure Micro Payment Using Physical
Unclonable Function", International Journal of Scientific
[1]. N. Chitra Kiran, G. Narendra Kumar, “Building Research in Computer Science, Engineering and
robust m-commerce payment system on offline Information Technology (IJSRCSEIT), ISSN : 2456-3307,
wireless network”,2011 Fifth IEEE International Volume 5 Issue 1, pp. 483-486, January-February 2019.
Available at doi : https://doi.org/10.32628/CSEIT195199
Conference on Advanced Telecommunication
Journal URL : http://ijsrcseit.com/CSEIT195199
Systems and Networks (ANTS) , March 2012

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


486
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951102

Reversible Data Hiding in Homomorphic Encrypted Domain by


Monitoring Ciphertext Group
V. Santhi *1, M. Abinaya 2
*1 Assistant Professor, Dept. of Computer Applications, Bon Secours College for Women Tamil Nadu, India
2 M.Sc Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India

ABSTRACT
Since few years, a new problem is trying to combine in a single step, compression, encryption and data hiding.
So far, few solutions have been proposed to combine image encryption and compression for example. Nowadays,
a new challenge consists to embed data in encrypted images. Since the entropy of encrypted image is maximal,
the embedding step, considered like noise, is not possible by using standard data hiding algorithms. A new idea
is to apply reversible data hiding algorithms on encrypted images by wishing to remove the embedded data
before the image decryption. Recent reversible data hiding methods have been proposed with high capacity,
but these methods are not applicable on encrypted images. In this paper we propose an analysis of the local
standard deviation of the marked encrypted images in order to remove the embedded data during the
decryption step. We have applied our method on various images, and we show and analyze the obtained results.
Keywords: Standard Data Hiding, Embedding Images

I. INTRODUCTION two technologies can be used complementary and


mutually commutative. Sinha and Singh proposed a
Nowadays, the transmission of images is a daily technique to encrypt an image for secure image
routine and it is necessary to find an efficient way to transmission. In their approach the digital signature
transmit them over networks. To decrease the of the original image is added to the encoded version
transmission time, the data compression is necessary. of the original image. The encoding of the image is
The protection of this multimedia data can be done done using an appropriate error control code. At the
with encryption or data hiding algorithms. Since few receiver end, after the decryption of the image, the
years, a problem is to try to combine compression, digital signature can be used to verify the
encryption and data hiding in a single step. For authenticity of the image. Encryption and
example, some solutions were proposed in to watermarking algorithms rely on the Kickoffs
combine image encryption and compression. Two principle: all the details of the algorithm are known,
main groups of technologies have been developed for and only the key to encrypt and decrypt the data
this purpose. The first one is based on content should be secret.
protection through encryption. There are several
methods to encrypt binary images or gray level II. LITERATURE SURVEY
images. In this group, proper decryption of data
requires a key. The second group bases the protection Reversible data hiding techniques The quality of the
on digital watermarking or data hiding, aimed at image gets disturbed when the data is embedded into
secretly embedding a message into the data. These the image. So it is expected that after the data

CSEIT1951102 | Received : 12 Feb 2019 | Accepted : 25 Feb 2019 | January-February -2019 [ 5 (1) : 487-490 ]
487
V. Santhi et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 487-490

extraction the image quality should be maintained Reversible data hiding is a technique to embed
just like the original image. But the image which is additional message into some distortion-unacceptable
obtained contains some distortions. With regards to cover media, such as military or medical images, with
distortion in image, Kalker and Williams established a eversible way so that the novel cover content can
a rate-distortion copy for RDH, through which they be perfectly restored after extraction of the hidden
showed the rate-distortion bounds of RDH for message. As an effective and popular means for
without memory covers and proposed a recursive privacy fortification, encryption changes the ordinary
code development which, however, does not move signal into incomprehensible data, so that the general
towards the bound. Another promising strategy for signal processing typically takes place before
RDH is histogram shift (HS), in which the space is encryption or after decryption.
saved where data can be embedded by shifting the
bins of histogram of gray values. In this process, the However, in some circumstances that a content
embedding of data is done in three steps. Step1. The owner does not trust the supplier, the ability to
histogram is drawn. Step2. The peak point is taken influence the encrypted data when maintaining the
into consideration. Step3. The whole image is plain content secret is needed. When the secret data
scanned row by row. After these steps, the image is to be broadcasted are encrypted, a supplier without
scanned again. If the greyscale value 154 is any information of the cryptographic key may
encountered, then the embedded data sequence is compress the encrypted data due to the limited
checked & we get the marked image. Finally, the data channel resource. Some attempts on RDH in
extraction is done. To get the original quality of the encrypted images have been made. Zhang divided the
cover, the process of histogram shift is applied again. encrypted image into numerous blocks. By spinning
The original cover is then obtained back. Basically, LSBs of the half of pixels in every block, space can be
data hiding is the process to hide the data into some created for the embedded bit. The data extraction and
covering media i.e. it is the concatenation of two image recovery proceed by finding which part has
blocks of data, first is the embedding data & second is been spinned in one block. This process can be
the covering media. But in most of the cases the grasped with the help of spatial correlation in
covering media gets distorted after the data is decrypted image.
embedded & the covering media is not inverted back
to its original form after data is removed from it. Hong et al. ameliorated Zhang’s method at the
Some reversible data hiding methods use the concept decryption side by further making use of the spatial
of differential expansion transform which is based on correlation using a different estimation equation and
haar wavelet transform. Another concept used is the side match method to gain much lower error rate.
histogram shift. The differential expansion is the These two methods explained above rely on spatial
difference between two neighbouring pixels for correlation of original image to extract data. That is,
hiding one bit of data. In this process, the histograms the encrypted image should be decrypted first before
are drawn first. Then the peak values are taken into data extraction.
consideration. Then two peak values are considered
& difference is calculated. Then according to the Zhang et al. recovered the recursive code
result the bit by bit data is embedded into the image. development for binary covers and proved that this
In this way the distortion analysis is done & it is development can gain the rate-distortion bound as
helpful to remove the distortion in the covering long as the compacting algorithm reaches entropy,
media & to get the original cover back. which launches the correspondence between data

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


488
V. Santhi et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 487-490

compression and RDH for binary covers. A more IMPLEMENTATION RESULTS


popular method is based on difference expansion
(DE), in which the difference of each pixel group is Authentication module
expanded, e.g., multiplied by 2, and thus the least An authentication module is a plug-in that collects
significant bits (LSBs) of the difference are all-zero user information such as a user ID and password, and
and can be used for embedding messages. So in this compares the information against entries in a
way the additional data can be embedded into the database. If a user provides information that meets
covering media which is an improvement to the the authentication criteria, the user is validated and,
existing methods. assuming the appropriate policy configuration,
granted access to the requested resource. If the user
III. METHODS AND MATERIAL provides information that does not meet the
authentication criteria, the user is not validated and
EXISTING PROCESS denied access to the requested resource.
There are existing systems having the key tool for
information hiding which is Vacating the room after Data Analysis
encryption. It consists of problems such as, the The Data Analysis module will focus on strategies
extracted data may contain errors. If there is no and procedures, both quantitative and qualitative, for
availability of sufficient space then some data may be analyzing social data. The module will complement
lost & that is why the data is missing at the receiver the material on data collection covered in the
side which can be termed as data with error. Again Research Appreciation module. Though the
the un-availability of memory space is a big problem. preferable sequence of progression would be from
Some space is created at the time of data embedding Research Appreciation to the Data Analysis module,
which is a time consuming process. After data the latter module can stand by itself enabling if
extraction the image recovered does not contain the necessary students to undertake these modules in
qualities of the original cover. Some distortions are either sequence.
introduced into the image. But it is possible in future
that the quality may be improved as compared to Data Encryption
existing system. Data encryption translates data into another form, or
code, so that only people with access to a secret key
PROPOSED METHODOLOGY (formally called a decryption key) or password can
The extracted data may contain errors because if read it. Encrypted data is commonly referred to as
there is no availability of sufficient space then some cipher text, while unencrypted data is called
data may lost & that’s why there is data missing at the plaintext.
receiver side which may called as data with error.
Again the unavailability of memory space is the big Data Decryption
problem, as some space is created at the time of data The conversion of encrypted data into its original
embedding which is the time consuming process. form is called Decryption. It is generally a reverse
After data extraction the image recovered does not process of encryption. It decodes the encrypted
contain the qualities as was the original cover. Some information so that an authorized user can
distortions are there into that image. only decrypt the data because decryption requires a
secret key or password.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


489
V. Santhi et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 487-490

Mail [4]. W. Liu, W. Zeng, L. Dong, and Q. Yao, “Efficient


The Mail module enables the sending of emails from compression of encrypted grayscale images,” IEEE
within Magnolia. The module can be used to send Trans. Image Process., vol. 19, no. 4, pp. 1097-1102,
Apr. 2010.
plain text and HTML messages, and template-based
[5]. V. Sachnev, H. J. Kim, J. Nam, S. Suresh, and Y.-Q.
messages. Sending emails from Magnolia is typically
Shi, “Reversible watermarking algorithm using
an automated process. An event acts as a trigger. For
sorting and prediction, ”IEEE Trans. Circuits Syst.
example, a verification message is sent to a user when Video Technol., vol. 19, no. 7, pp. 989-999, Jul.
they fill a registration form. 2009.
[6]. W. Zhang, B. Chen, and N. Yu, “Capacity-
IV. CONCLUSION approaching codes for reversible data hiding,” in
Proc 13th Information Hiding(IH’2011),LNCS 6958,
Data hiding is gaining the area of interest due to its 2011, pp. 255-269, Springer-Verlag.
provision for secured environment. Data hiding in [7]. J. Tian, “Reversible data embedding using a
difference expansion, ”IEEE Trans. Circuits Syst.
reversible manner in encrypted images is providing
Video Technol., vol. 13, no. 8, pp. 890- 896, Aug.
double security for confidential data by using
2003.
techniques such as image encryption. The existing
[8]. D.M. Thodi and J. J. Rodriguez, “Expansion
system contains some disadvantages so the future embedding techniques for reversible watermarking,”
scope is to remove the disadvantages by adding IEEE Trans. Image Process., vol. 16, no. 3, pp. 721-
reversible manner i.e. data extraction and recovery of 730, Mar. 2007.
image are free of errors. The PSNR will be improved [9]. X. L. Li, B. Yang, and T. Y. Zeng, “Efficient
to get original cover back. reversible watermarking based on adaptive
prediction-error expansion and pixel selection,”
V. FUTURE ENHANCEMENT IEEE Trans. Image Process., vol. 20, no. 12, pp.
3524-3533, Dec. 2011.
[10]. Z. Ni, Y. Shi, N. Ansari, and S. Wei, “Reversible data
In future it may be possible that memory space can
hiding,” IEEE Trans. Circuits Syst.Video Technol.,
be reserved before encryption which requires less
vol. 16, no. 3, pp. 354- 362, Mar. 2006.
amount of time for data extraction & image recovery. [11]. L. Luo et al., “Reversible image watermarking using
interpolation technique,” IEEE Trans. Inf. Forensics
Security, vol. 5, no. 1, pp. 187-193, Mar. 2010.
VI. REFERENCES [12]. X. Zhang, “Reversible data hiding in encrypted
images,” IEEE Signal Process. Lett., vol. 18, no. 4,
[1]. P. Tsai, Y. C. Hu, and H. L. Yeh, “Reversible image pp. 255-258, Apr. 2011.
hiding scheme using predictive coding and
histogram shifting,” Signal Process.,vol. 89, pp. Cite this article as :
1129-1143, 2009.
[2]. M.Johnson, P.Ishwar, V.M.Prabhakaran, V. Santhi, M. Abinaya, "Reversible Data Hiding in
D.Schonberg, and K.Ramchandran, “On com- Homomorphic Encrypted Domain by Monitoring
pressing encrypted data," IEEE Trans. Signal Ciphertext Group ", International Journal of Scientific
Process, vol. 52, no. 10 Oct 2004., pp. 2992-3006. Research in Computer Science, Engineering and
[3]. W. Hong, T. Chen, and H.Wu, “An improved Information Technology (IJSRCSEIT), ISSN : 2456-3307,
reversible data hiding in encrypted images using Volume 5 Issue 1, pp. 487-490, January-February 2019.
side match,” IEEE Signal Process. Lett., vol. 19, no. Available at doi : https://doi.org/10.32628/CSEIT1951102
4, pp. 199-202, Apr. 2012. Journal URL : http://ijsrcseit.com/CSEIT1951102

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


490
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951118

A Review : Venue Booking System


Shruti N. Ghule*1, Harish M. Mohurle2, Akshay P. Shegaonkar3, Zeba Parvin Abdual Saleem4,
Pradhnya S. Borkar5
1-4Priyadarshini J. L. College of Engineering, Nagpur, India
5 Assistant Professor in Priyadarshini J.L. College of Engineering, Nagpur, India
ABSTRACT
Ecommerce refers to both online retail as well as electronic transactions. Ecommerce has hugely increased in
popularity over the last decades, and in ways, it's replacing traditional brick and mortar stores. This application
enables you to book venues as per the choice of customer and his / her requirements. Venue booking system is a
web application in which service providers or a venue owner can registered their venues and customers can
book venues for events as per their location, availability and interest. This makes the workload simple for
customers to book a venue for an event they need to search for it without visiting actual place. In this
application, the user needs to be registered and log in for further advantages. The customer can search for
venues by specifying names or by giving the location if venues are present in that particular location.
Keywords: E-Commerce, Venue Booking

I. INTRODUCTION enter into a single system for reservation which


provides availability and profile of different venues..
In the 21st century most of the users are using If a particular venue is not available they have to go
internet for purchasing the product online, paying for to other website which is a time consuming process.
bills, booking for tickets etc. Due to rapid use of venue and is time consuming as the user checks with
internet the user have reduced purchasing their multiple services on their own. All this consumes lot
product through stores. E-Commerce based websites of time. Consider a user who wish to book a wedding
are increasing massively. The venue or hotel Industry hall for the wedding. The first thing that user does is
is one of the largest and most profitable industries in to visit the owner of wedding hall, discuss with him
most countries. It helps to oust the economy of both about the services offered by the hall and then
developed and developing countries. negotiate the deal for booking the hall.

People are always looking around’ when vacationing Furthermore there may be some services which are
for the cheapest venues for celebration’s and arrange not provided by the wedding hall owner, so for that
bookings in advance to guarantee their stay for the the user has to inquiry and get there services from a
period of their celebration. However these activities third party, such services can be like decoration of
generally require individuals to do the bulk of the hall, orchestra, photographer etc. So in he proposed
search and booking. This is not only time consuming system we aim to gather the entire venue for all types
exercise, but can be frustrating and costly as well. of function and also the services. By this user will
have much more options under one roof.
The objective of the system is to reserve a venue for
functions, parties, weddings, meetings etc., user can

CSEIT1951118 | Received : 12 Feb 2019 | Accepted : 25 Feb 2019 | January-February -2019 [ 5 (1) : 491-494 ]
491
Shruti N. Ghule et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 491-494

II. LITERATURE SURVEY people are viewing it and more and more people are
queuing.
“LARS: A Location-Aware Recommender System”
authored by Justin J. Levandoski ,Mohamed Sarwat, In [4] the author tells about Position based Services
Ahmed Eldawy, Mohamed F. Mokbel contains a which offers many benefits to the mobile consumers
position alert recommender structure that uses to recover the information about their current
position-based ratings to generate references. LARS position and manner that data to get more useful
achieves Customer rating positions through user information near to their position. With the help of
partitioning, a technique that influences GPS in phones and through Net Services using GPRS,
recommendations with ratings spatially close to Position based Facilities can be fulfilled on Android
querying users in a manner that maximizes system based smart phones to provide these value-added
scalability while not sacrificing recommendation services like handset profiling, scheduling alarms.
quality. Position based services are a general class of
computer program level services used to include
“Position Detection and Tracking System” published specific controls for position and time data as control
by Mahesh Kadibagil, and Dr. H S Guruprasad features in computer programs.
proposed a Self-directed location recognition and
tracking method which enhances the correctness of III. PROPOSED
finding friends and family member’s locations by
using GPS and typical web technology. This design The main aim of the proposed concept is to develop a
includes a mobile client, a repository, a web client web application that will allow the user to search and
and a map service. The mobile user is used to book the menu. The application should be able to
discover location and conduct a notification SMS to provide a platform for venues providers to showcase
user when his/her friends or family members come their services. There should be the options for
nearby the user’s zone of path. This location material disbursing auto discount for existing customers. The
can be sent to the server and the same information application should provide details of all booked
can be achieved and observed using the web venues with dates and user can query if they want to.
consumer by other customers. Following fig1. shows work flow VBS which has two
parts service provider (venue owner) and second is
In [3] the author wants to convey that with user having their own functionality of booking venue
the hasty improvement of Internet, people's day-to- or registering the venue.
day life has basically been attached from the Internet.
After the Internet, without resolution, whether
entities or industries, both clients and industrialists,
are made on the network of expectation, this new
expectation is to get through the network of the
greatest things the inexpensive, most practical to find
the fastest information, the most money-making
products to sell up. As people's lives are getting better,
it comes with many lifestyle changes. As the film are Fig1. Work Flow of VBS
still the same and has not changed predilections and
as more and more film are churned out more an more

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


492
Shruti N. Ghule et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 491-494

• Admin
Admin can add, update the categories in which a
venue will be added. He/ She can manage all the
functionality related to user and venue owner.
The admin acts as a intermediate between this
two modules.
• Booking
The booking module will handle the booking
functionality also it manages all functions related
to the enquiries of different venues and
availability of dates according to bookings.
Fig 2. DFD for Customer • Category
Above data flow diagram(DFD) i.e. fig2 will show This module will used to saving the categories
how user can interact with the application by sign in, from admin module. It contains different
sign up, searching venues etc. categories related to venues which can be decided
by venue owner according to his / her venue.
• Discount
This module will handle the functionality related
with discount. It provides discounts to the
existing user which will decided by venue owner.
• Gateway
This is the dummy implementation of gateway
for payments. It would be the full payment to
owner or some amount of down payment.
• Messages
This will used for showing and storing the
Fig 3. DFD for Venue Owner inquiry messages. This is basically act as enquiry
Fig3 will show how venue owner can interact with section where user can ask their inquiries related
the application by registering their venues and to venues.
provide various services as per their venue. • User
The user module will provide functionality for
This application will consists of following modules. signup and login etc. This will be customer
1) Admin Module section.
2) Booking Module • Venue
3) Category Module The venue module will used for showing and
4) Discount Module managing the venues which are registered and
5) Payment Gateway their information and services provided by venue
6) Inquiry Module owner.
7) Customer Module • DAO
8) Venue Owner Module Data access object will used for managing the
Following information shows functionality of database connectivity
different modules.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


493
Shruti N. Ghule et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 491-494

The proposed system will be implanted as a web Pages 196-199 July 16-17, 2011 IEEEComputer
based application on J2EE platform of Java with Society Washington, DC, USA ©2011 ISBN:978-
MySQL as database. 0-7695-4443-4 10.1109/ISCCS.2011.61
[4]. VEDANG MOHOLKAR,PRATHAMESH
Thus we have been able to design a venue booking HULE,MANDAR KHULE,SUMIT SAURABH
system and have been able to carry on partial computer department,pvpit,pune,India
implementation as. “AUTOMATED LOCATION BASED
SERVICES”, International Journal of Advanced
IV. CONCLUSION Research in Computer Science and Software
Engineering, Volume 4, Issue 2, February 2014
In this paper we discuss about managing different ISSN: 2277 128 X.
venues according to user need. It helps user to book
venues as per their requirements. It can be different Cite this article as :
halls, lawns, meeting rooms etc. The user can show
their interest by booking venues. This application Shruti N. Ghule, Harish M. Mohurle, Akshay P.
will overcome the problem of previous system in Shegaonkar, Zeba Parvin Abdual Saleem, Pradhnya S.
which all the work were done manually. It gives Borkar, "A Review : Venue Booking System",
better results by booking venues directly without International Journal of Scientific Research in
visiting the venue. Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
V. REFERENCES 5 Issue 1, pp. 491-494, January-February 2019.
Available at doi :
[1]. Justin J. Levandoski ; Mohamed Sarwat ; Ahmed https://doi.org/10.32628/CSEIT1951118
Eldawy ; Mohamed F. Mokbel “LARS: A Journal URL : http://ijsrcseit.com/CSEIT1951118
Location-Aware Recommender System” ICDE
'12 Proceedings of the 2012 IEEE 28th
International Conference on Data Engineering
Pages 450-461 April 01 - 05, 2012 IEEE
Computer Society Washington, DC, USA ©2012
ISBN: 978-0-7695-4747-3 10.1109/ICDE.2012.54
[2]. MAHESH KADIBAGIL PG SCHOLAR, DEPT.
OF ISE, DR. H S GURUPRASAD PROFESSOR
AND HEAD, DEPT. OF CSE, BMSCE,
BANGALORE, INDIA POSITION DETECTION
AND TRACKING SYSTEM IRACST -
International Journal of Computer Science and
Information Technology & Security (IJCSITS),
Vol. 4, No. 3,June 2014
[3]. BO HANG “DESIGN AND IMPLEMENTATION
OF CINEMA ONLINE BOOKING SYSTEM”,
ISCCS ‘ 11 Proceedings of the 2011 International
Symposium on Computer Science and Society,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


494
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951103

A Review on IOT Based Garbage Monitoring System


Himanshu Gupta1, Amit Chaturkar1, Abhay Bagde1, Hritik Raj1, Ankur Hemane1, Prof. Prerana Deshmukh2
1 BE Students, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,
Maharashtra, India
2Assistant Professor, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,
Maharashtra, India
ABSTRACT

Ordinarily, in our city, we see that the garbage bins or dustbins put at open spots are over-burden. It makes
unhygienic conditions for individuals just as grotesqueness to that place leaving a terrible stench. To dodge
every single such circumstance we are going to execute a venture called IoT Based Smart Garbage and Waste
Collection Bins. As the populace is expanding step by step, the earth ought to be spotless and sterile. In a large
portion of the urban areas, the flooded garbage bins are making an unhygienic situation. This will additionally
prompt the emerging of various kinds of maladies. This will corrupt the way of life. To beat these circumstances
a productive savvy squanders collection framework must be created. As the extent of IoT is creating step by
step successful techniques can be discovered effectively. Different plans were proposed and have points of
interest just as drawbacks. This paper is a review dependent on Smart Waste Collection System dependent on
IoT.
Keywords : Internet of Things, Smart Garbage Monitoring, Smart City, Microcontroller

I. INTRODUCTION the fuel utilization, CO2 outflows and city traffic


decrease. Taking into account that the courses will be
The garbage collection issue on megalopolis turned advanced so as to keep away from a truck moving to a
into a genuine issue crosswise over numerous urban vacant waste. For instance to send a truck to a course
communities around the globe. One of the incredible in which the whole of the occupations of the dumps
regions of concentrate inside Smart City and Internet don't extrapolate the limit of the truck, which will
of Things (IOT), includes the proposition of answers permit enhancement of the quantity of trucks
for regular issues of the urban communities, one of utilized in a given territory. The metropolitan waste
these issues is accurately the Garbage Collection. collection has just been streamlined through
Pointers demonstrate that Brazil has a low exhibition programming and numerical techniques [3], anyway
with regards to the collection and transfer of urban the outcomes were not agreeable, because of the
waste [1]. substantial measure of factors that brought about an
incredible computational time which may make the
The inquiries identified with garbage collection and utilization of remote sensors difficult and reasonable.
that should be unraveled are a few, in this work the As per [4] as the Internet and different innovations
issue that will be tended to is simply the one that kept on creating and advance in the principal decade
bargains with the collection itself, that is, the age of of the twenty-first century, a few arrangements rose
the courses for the trucks so as to permit decrease in up out of monsters showcases that made Internet of

CSEIT1951103 | Received : 13 Feb 2019 | Accepted : 25 Feb 2019 | January-February -2019 [ 5 (1) : 495-500 ]
495
Himanshu Gupta et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 495-500

Things a suitable alternative for a decent number of developing and populated urban communities as it
urban communities. contains contaminants which result in flimsiness,
issue and uneasiness to the biological system.
The board of waste is a major testing issue in urban Numbness and absence of neatness are ruining nature.
regions for the greater part of the nations all through The correct waste expulsion and the executives are
the world and is seen in a large portion of the incredibly viable to improve the wellbeing and
creating nations than in the created nations. A prosperity of the city's populace.
productive administration of waste is a prerequisite
for keeping up a perfect and green condition as there The primary objective of this paper is to take a shot at
is increment in a wide range of squanders tossed by natural issues because of ill-advised waste transfer
numerous spots like mechanical, agrarian, home and unravel them for better wellbeing and
waste, and so on. Squander collection and reusing is cleanliness of the general population. The proposed
done through different innovations. Collection of framework fits into the classification of IoT
data is huge and unwieldy. The present development connected to outer and open situations and it satisfies
in country with vast local location and an interest for the accompanying essential prerequisites of IoT
modernization in the city makes a testing administrations:
undertaking for waste administration individuals [5].
• Unwavering quality: Communication is crucial in
India especially creates around 1, 33,760 tons of IoT for administration provisioning, connected to
Municipal strong waste (MSW) the executives every the outside and open condition. Dependable and
day, of which roughly 91,152 tones are gathered, and solid correspondence is required arranged by
a colossal entirety of cash is spent on collection [6]. completing correspondence between gadgets as
World waste generation is relied upon to be roughly this sort of IoT has an extensive administration
27 billion tons for each year by 2050, 33% of which space. Consequently, the receptacle utilized in
will originate from Asia, with significant the proposed framework associated with one
commitments from China and India [7]. Squander age another, in light of a remote work arranges
in urban regions of India will be 0.7 kg per individual (WMN), giving dependability.
every day in 2025, around four to multiple times • Portability: IOT gadgets might be required to
higher than in 1999 [8]. move in the external environment. The proposed
framework works with a battery rather than the
II. MOTIVATION AND BACKGROUND changeless power source, bringing about an
extraordinary dimension of versatility. The
In the previous history, since the human populace versatility of the framework is secured with a
thickness and dimensions of misuse of characteristic power-based power supply.
assets were less, the measure of strong waste created • Administration Continuity: Data communications
was noteworthy. Be that as it may, the expansion in and administrations ought to be directed
the populace in this day and age has expanded the flawlessly whenever and any area in IoT with an
garbage. To keep nature spotless and solid, there is a extensive administration space. Bins are situated
necessity of legitimate garbage transfer. Inappropriate at the customary space of separation to guarantee
garbage transfer raises contamination, medical the administration progression.
problems, different dangers and in result it influences • Client Convenience: The presentation of IoT has
nature. Contamination extremely influences the improved client comfort. For client facilitate, the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


496
Himanshu Gupta et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 495-500

proposed framework lessens the method postpone Another technique [10], there are different
the time of the existing garbage gathering dustbins situated all through the city or the Campus.
frameworks, which empowers clients to set free These dustbins are furnished with ease implanted
their garbage without a long pause and auspicious gadget which helps in following the dimension of the
end of filled bins. garbage bins and a one of a kind ID will be
• Vitality Efficiency: IoT connected to outer and accommodated each dustbin in the city so it is
open conditions depends on a dependable on the anything but difficult to recognize which garbage
framework and requires versatility, causing a lot container is full. The undertaking module is isolated
of vitality utilization. To take care of this issue, into two sections Transmitter area and recipient
the sensors work utilizing vitality proficient segment. Here in the transmitter segment we are
procedures, expanding their battery lifetimes. utilizing 8051 microcontrollers, RF Transmitter and
sensors these are appended to the dustbin. Where
III. LITERATURE REVIEW
sensor is utilized to identify the dimension in the
dustbin whether the dustbin is full or void.
The garbage the executives in urban communities
must be successfully and productively actualized. The
Another strategy [11] is that, when the garbage
different recommendations were advanced and some
achieves the edge level ultrasonic sensor will trigger
of them effectively actualized. Be that as it may, it
the GSM modem which will constantly alarm the
can't be considered as a powerful one. Therefore, an
required expert until the garbage in the dustbin is
overview was done among various proposition and
squashed. When the dustbin is squashed, individuals
this review paper incorporates study among various
can reuse the dustbin. At customary interims dustbin
strategies for Waste Collection System dependent on
will be squashed. In this strategy, GSM 900A modem
IoT.
is utilized to send the messages.

The paper [9] proposed waste collection framework


Another strategy for garbage the executives is
depends on waste dimension information from
presented [12] as pursues. A dustbin is interfaced
trashcans in a metropolitan territory. The
with microcontroller based framework having IR
information gathered by sensors is sent over the
remote frameworks alongside focal framework
Internet to a server where it is put away and handled.
demonstrating current status of garbage, on portable
The creator gathered information is then utilized for
internet browser with HTML page by Wi-Fi.
checking and enhancing the every day choice of
Consequently the status will be refreshed on to the
trashcans to be gathered, figuring the courses in like
HTML page.
manner. Consistently, the specialists get the recently
determined courses in their route gadgets. The key
In paper [13] Infrared sensor (IR sensor) is utilized
element of this framework is that it is intended to
which is a multipurpose sensor, which can recognize
gain as a matter of fact and to settle on choices on the
the dimension of garbage. IR sensor produces the
every day squander level status as well as on future
light, which is imperceptible to stripped eye yet the
state estimate, traffic clog, adjusted cost-proficiency
electronic segments can distinguish it.
capacities, and other influencing factors that from the
earlier people can't predict.
In Paper [14] System screens the garbage bins and
educates about the dimension of garbage gathered in
the garbage bins by means of a page. For this the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


497
Himanshu Gupta et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 495-500

framework utilizes ultrasonic sensors put over the partition from wrong dimension estimation and are
bins to identify the garbage level and contrast it and interfaced with the microcontroller.
the garbage bins profundity. The framework makes
utilization of AVR family microcontroller, LCD To distinguish awful stench a gas sensor is set at the
screen, Wi-Fi modem for sending information and a base of the dustbin and is adjacent to interfaced with
signal. the controller to remember it off the waste filled in
the dustbin. The two sensors send the signs to the
In paper [15] guarantees the cleaning of dustbins controller. Arduino accumulates information gotten
soon when the garbage level achieves its most by the gatherer and trade nearby page through the
extreme. In his administration framework IOT as the Ethernet shield.
working in the field for arranged radio-recurrence
distinguishing proof (RFID), following the collection
vehicle, Dustbin observing and other developing
detecting advancements.

Creators in [16] consider dynamic booking over a lot


of recently characterized collection trips. The
fundamental target of the methodology is to limit the
complete operational and fixed truck costs.

A numerical detailing procedure is proposed in [17]


building up an arrangement of administrative regions, Figure 1. System Architecture
characterizing directing, and planning booking
mulling over conceivable new elective arrangements The ultrasonic sensor is utilized to check the
in dealing with the framework in general. dimension status of a dustbin so as to pick whether it
is full or passed greatest limit esteem. Dynamic status
In [18] creators assess dynamic arranging strategies of dustbin is appeared on the site using relationship
connected for the waste collection of underground through the Ethernet shield. Checking the page will
bins. Demonstrate diminishes the measures of carbon help the waste collection office with following for the
dioxide discharged in the earth from trucks by correct area and proportion of the junk. The waste
making dynamic steering progressively powerful. vehicles would then have the ability to debilitate the
garbage from a specific domain.
IV.PROPOSED METHODOLOGY
The farthest point of GSM module is to establish a
In this framework, dustbins are organized at level 1 of connection on the waste social event division when it
a building made under savvy city activity. It will gets full. The garbage bins put at level 1 of the
gather the misfortune through keen pipe framework structures can be effectively emptied utilizing engines
set in the building. The perceptive dustbins are to pivot it by 180⁰ while the gatherer truck is at an
interfaced with the web through GSM to get the area under dimension 1 (ground level).
present status. Two sensors are settled and no more
lifted inspiration driving the dustbin to keep up a key

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


498
Himanshu Gupta et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 495-500

V. CONCLUSION [8] Modak P, Jiemian Y, Hogyuan Yu, Mohanty CR


2010 Municipal solid waste management
Despite the fact that online pharmacy is more subject turning waste into resources in Shanghai
than nearby pharmacy particularly in remote regions, manual: a guide for sustainable urban
it ought to be planned to build up standards and development in 21st century, pp. 1-36
direction like transfer of filtered solutions. Patients [9] Kumar JS, Subbaiah KV, Rao PVVP. 2014
ought to know on utilization of doctor prescribed Municipal solid waste management in India.
prescriptions. It is great to embrace the plans and Austr J. Eng. Res. 2, 1-8 (DOI: 10.7603/s40632-
projects of driving nations to structure and build up 014-0001-4)
the e-pharmacy framework. [10] Jose M. Gutierrez, Smart Waste Collection
System Based on Location Intelligence.
VI. REFERENCES Procedia Computer Science 61 ( 2015 ) 120 –
127..
[1] Melo, Alexander Bento, et al. "Optimization of [11] Parkash, Prabu, IoT Based Waste Management
Garbage Collection Using Genetic Algorithm." for Smart City. IJECS Vol. 4, Issue 02 February
Mobile Ad Hoc and Sensor Systems (MASS), 2016.
2017 IEEE 14th International Conference on. [12] Monika K, Smart Dustbin-An Efficient Garbage
IEEE, 2017. Monitoring System. IJECS Volume 6 Issue No.
[2] C. A. Mucelin, “LIXO E IMPACTOS 06 June 2016.
AMBIENTAIS PERCEPTÍVEIS NO [13] S.S. Navghane, IoT Based Garbage, and Waste
ECOSSISTEMA Collection Bin. IJARECE Volume 5, Issue 5,
[3] URBANO Garbage and perceptible May 2016.
environmental impacts in urban ecosystem,” [14] Meghana KC, Dr. K R Nataraj IOT Based
vol. 20, no. 1, pp. 111–124, 2008. Intelligent Bin for Smart Cities. IJRITCC May
[4] R. Fujdiak, P. Masek, P. Mlynek, J. Misurec, 2016.
and E. Olshannikova, “Using Genetic [15] Abdul Atif Khan, Study Of Smart City Using
Algorithm for Advanced Municipal Waste Internet Of Things. Ijritcc March 2016.
Collection in Smart City,” 2016. [16] Vishesh Kumar Kurrel, Smart Garbage
[5] M. N. K. Boulos and N. M. Al-shorbaji, “On the Collection Bin Overflows Indicator using the
Internet of Things , smart cities and the WHO Internet of Things. Volume 3, Issue 05 May
Healthy Cities,” pp. 1–6, 2014. 2016.
[6] Jain, Aaditya, and Ranu Bagherwal. "Design [17] J. Q. Li, D. Borenstein, P. B. Mirchandani,
and implementation of a smart solid waste “Truck Scheduling forSolid Waste Collection in
monitoring and collection system based on the City of Porto Alegre, Brazil”, Omega,
Internet of Things." Computing, Elsevier, 2008, vol. 36, pp. 1133-1149.
Communication and Networking Technologies [18] P. T. R. Ramos, M. I. Gomes, and A. P. B.
(ICCCNT), 2017 8th International Conference Povoa, “Assessing and improving Management
on. IEEE, 2017. Practices when Planning PackagingWaste
[7] Mirchandani, Sahil, et al. "IoT enabled Collection Systems”, Resources Conservation
dustbins." Big Data, IoT and Data Science, 2017 and recycling, Elsevier, 2014, vol. 85, pp. 116-
International Conference on. IEEE, 2017. 129.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


499
Himanshu Gupta et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 495-500

[19] A. Stellingwerff, “Dynamic Waste Collection:


Assessing the Usage of Dynamic Routing
Methodologies”, Master Thesis, Industrial
Engineering & Management, University of
Twente, Twente Milieu, 2011.

Cite this article as :

Himanshu Gupta, Amit Chaturkar, Abhay Bagde,


Hritik Raj, Ankur Hemane, Prof. Prerana Deshmukh,
"A Review on IOT Based Garbage Monitoring
System", International Journal of Scientific Research
in Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 495-500, January-February 2019.
Journal URL : http://ijsrcseit.com/CSEIT1951103

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


500
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951122
A Survey on Rice Crop Yield Prediction in India Using Improved Classification
Technique
Kolin Sukhadia*1, M. B. Chaudhari2
*1PG Scholar, Department of Computer Engineering, Government Engineering College Modasa, Modasa, Gujarat, India
2Head of Department, Department of Computer Engineering, Government Engineering College Modasa, Modasa, Gujarat,
India

ABSTRACT

India is an agricultural country. Agriculture is the important contributor to the Indian economy. There are
many classification techniques like Support Vector Machine(SVM), LADTree, Naïve Bayes, Bayesnet, K Nearest
Neighbour(KNN), Locally Weighted Learning(LWL) on rice crop production datasets. They have some
drawbacks like low accuracy and more errors. To achieve more significant result, To increase classification
accuracy and reducing classification errors, our research uses classification method Bayesnet based adaboost
will be proposed in work. Rice crop yield depend on environment’s parameters like Rainfall, minimum
temperature, average temperature, Maximum temperature, Vapour Pressure, potential evapotranspiration,
reference crop evapotranspiration, cloud cover, wet day frequency for the kharif season. our dataset containing
these environmental parameters for accurate prediction of Rice crop yield.
Keywords : AdaBoost Technique, BayesNet Classifier, Rice Crop Yield Prediction, Kharif Season

I. INTRODUCTION used by Government, policy makers, agro-based


industries, traders and agriculturists. The
Rice crop production contributes more than 40% to Government use crop yield prediction in
overall crop production. Its production is reliant on procurement, distribution, buffer-stocking, import
climatic conditions. Improving the ability to predict export, price fixation and marketing of agricultural
crop productivity in under different climatic commodities[8]. Agriculture is the most important
parameters like rainfall, temperature, cloud cover, branch in Indian economy and 70 percentage of rural
vapour pressure, potential evapotranspirsation etc. population livelihood depends on agricultural work.
can assist farmers and other stakeholders in making Crop yield depends on environment’s factors like
important decisions in terms of agronomy and crop precipitation, temperature, evapotranspiration, etc.
choice[43]. The prediction of agricultural yield is a Based on previous experience, Farmers cultivate crop.
challenging and desirable task for every nation. But nowadays, the uncertainty increased in
Prediction of yield can be done based on historic crop environment. So, accurate analysis of historic data of
cultivation data and environment data and environment parameters should be done for
environment data crop yield can be predicted. Crop successful farming. To get more harvest, we should
yield prediction is the forecasting of harvest in Kg per also do the analysis of previous cultivation data. The
Hectare. To analyse data various data mining Prediction of crop yield can be done based on historic
technique can be used. Crop yield prediction can be

CSEIT1951122 | Received : 15 Feb 2019 | Accepted : 26 Feb 2019 | January-February -2019 [ 5 (1) : 501-507]
501
Kolin Sukhadia, M. B. Chaudhari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 501-507

crop cultivation data and weather data using data Regression AdaBoost (Adaptive Boosting), Support
mining methods[14]. Vector Machine Regression and a Modified
Nonlinear Regression (MNR) equation and then
There are many classification techniques like Support comparing the modified equation with the other
Vector Machine(SVM), Naïve Bayes, BayesNet, K three methods to check its accuracy For Aus and
NearestNeighbour (KNN), Locally Weighted Boro Rice our modified Nonlinear Regression
Learning(LWL), LAD Tree are applied on rice crop equation not only yields better RMSE, MSE and MAE
production datasets. The provision of accurate and values, it also has the highest Rsquare value, which
timely information can assist farmers to make the proves that the MNR equation is the best fit for this
best decision for their cropping situations This could study. this study also proves that weather conditions
benefit them to attain greater crop productivity[42]. play very vital role in predicting the yield. In [2],
This study examines the application of machine author used classifiers for research are J48, LWL,
learning for the prediction of rice crop yield. The LAD Tree and IBK. The experimental results showed
current study use the most rice producing that J48, LADTree achieved the highest accuracy,
states(Haryana, Punjab, Tamilnadu, Andhra Pradesh, sensitivity and specificity. In [3], author used
Bihar). The dataset parameters are state name, district classifiers were BayesNet and NaiveBayes. The results
name, crop year, Rainfall, minimum temperature, showed that the performance of BayesNet was much
average temperature, maximum temperature, vapour better compared with NaiveBayes for the dataset. In
pressure, reference opevapo transpiration, [4], proposed approach builds a model to predict the
Potential evapotranspiration, diurnal temperature weather parameters applying Neural Networks, then
range, area, production, yield, class. The dataset was estimates the rice yields applying Support Vector
then sorted on the basis of yield to classify the Regression. In [5], The authors have used pre-
records in to low, moderate and high. The low yield processed data that was clustered using k-means
was from 0. 1 to 1 tonnes/hectare, moderate from 1 to clustering algorithm. The association rule mining
2 tonnes/hectare and high from 2 to 5 process will apply on clustered data to find the rules.
tonnes/hectare. The training phase ends with number of generated
rules. In the testing phase, the yield value is predicted
The further document is as follows. In Section 2 based on the generated rules. The parameters
(Literature Survey), the literature review done on considered for research were precipitation, minimum
Rice Crop Yield Prediction, Bayesnet Method, temperature, reference crop evapotranspiration
AdaBoost method and Rice Crop. In Section 3 average temperature, maximum temperature.
(Conclusion) conclusion derived based on literature production, area and yield for the Kharif season
survey. including months from June to November.

II. LITERATURE SURVEY This research discusses the experimental results


obtained by applying SMO classifier using the WEKA
A. CROP YIELD PREDICTION tool on the dataset of 27 districts of Maharashtra state,
India in [6]. In [7], The parameters considered for the
IN [1] Umid Kumar Dey, Abdullah Hasan Masud, present study were precipitation, reference crop
Mohammed Nazim Uddin does research to predict evapotranspiration average temperature, minimum
the yield of rice in the regions of Bangladesh during temperature, maximum temperature, area,
the aforementioned seasons using Multiple Linear production and yield for the Kharifs season (June to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com | UGC Approved Journal [Journal No : 64718]
502
Kolin Sukhadia, M. B. Chaudhari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 501-507

November) for the years 1998 to 2002. In this Nigeria. The climate data used was for a period of 36
research WEKA tool is used. A Multilayer Perceptron years from 1980 to 2015. This study[14] describes the
Neural Network was implemented as ANN. Cross role of data mining in Agriculture and crop yield
validation method was used to validate data. In [8], prediction. This paper also describes Groundnut crop
The proposed method uses yield and weather data yield prediction analysis and Naive Bayes Method.
collected from United States Department of This Study [15] describes multi-model heterogeneous
Agriculture. The parameters in the dataset are rainfall, ensemble approach. This study [16] examine the
humidity, temperature, yield and. regression-based introduction about Bayesian network, Directed
algorithms to predict the crop yield. These were acyclic graph, Conditional Probability table. This
multivariate polynomial regression, support vector study [17] examine various data mining classification,
machine regression and random forest. support clustering and association rule mining technique.
vector machine regression to obtain the best possible This study [18] examine detail information about
results for predicting the crop yield. In [9], The naïve bayes classification technique. This study [19]
prominent classification methods in this study are the examine Bayesian network classification technique,
Support VectorMachine(SVM) and Naive Bayes. In structure learning technique, parameter learning
this paper, AdaSVM and AdaNaive are the proposed technique. This study[20] examine various adaptive
ensemble model use to project the crop production Learning Algorithms for Bayesian Network Classifiers.
over a period of time. This ensemble model is This study[21] examine Bayesian Network Classifiers
compared with Naive Bayes and SVM methods. in Weka. This study[22] examine the introduction
accuracy and the classification error are two about bayesian approach to Probability and Statistics.
parameters used for output predictions. The finding This study[23] examine bayesian method for
yields that AdaSVM and AdaNaive are agreeable than constructing probabilistic networks from databases.
SVM and Naive Bayes for the data set analyzed. In they focus on constructing Bayesian belief networks.
this study[10], regression based adaptive boosting This study[24] examine the overview of rice crop
prediction model is presented, There are three coastal yield prediction. Examines Different data mining
districts belonging to Odisha are located in India techniques utilized for foreseeing rice crop yield. In
using the datasets of Rabi and Kharif seasons. This this study[25], there is overview of rice crop yield
study experiments on weak regressors, such as: SVR prediction. This paper Examines Different data
regression, linear, lasso, ridge proposes strong mining techniques utilized for foreseeing rice crop
predictors by avoiding the shortcomingss of yield. In this study[26], rice yield prediction is
individual weak regressors. In [11] study examines performed by back propagation algorithm. In this
the variations of climate variables and rice yield and study[27], there is Comparative Study of
examines the relationships using multiple linear Classification Algorithms for Forecasting Rainfall.
regression, principal component analysis, and support This study[28] examine Modeling Rainfall Prediction
vector machine (SVM) analysis in southwest Nigeria. Using Data Mining Method: A Bayesian Approach
In this study[12], Adaptive boosting of weak data mining method for modeling rainfall prediction.
regressors for forecasting of crop production In [29], Rice Crop Yield Forecasting of Tropical Wet
considering climatic variability. In this study[13] and Dry Climatic Zone of India using data mining
examines the variations of climate variables and rice Techniques is applied. For data set results show that
yield and quantifies the relationships using multiple J48 and LADTree achieved the highest accuracy,
linear regression, principal component analysis, and sensitivity and specificity. Classification by LWL
support vector machine (SVM) analysis in southwest classifier displayed the lowest accuracy, sensitivity

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com | UGC Approved Journal [Journal No : 64718]
503
Kolin Sukhadia, M. B. Chaudhari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 501-507

and specificity results. In [30], This study presents an Accuracy is highest in case of Soybean. A study by
analysis using data mining techniques for estimating [37] has applied association rules for Tamilnadu. Rice
the future yield prediction in tea cultivation with crop yield predicted using regression analysis,
climatic change trends observed in last 30 years experiment implemented in MATLAB tool. Yield
(1977-2006). The patterns of crop production in predicted from Annual Rainfall, Area under
response to the climatic (rainfall, temperature, Cultivation, and Food Price Index. Relationship
relative humidity, evaporation and sunshine) effect established between AR, AUC and FPI[38]. A study
across the four tea growing regions (South Bank, [39] described literature review in tabular form. Some
North Bank, Upper Assam and Cachar) of Assam research study only summarized the data mining
were developed using Multiple Linear Regression techniques for crop yield prediction. Support Vector
(MLR) technique. Machine, Random Forest, Neural Network, REPTree,
Bagging, Bayes algorithm described in the study [40].
The tea production estimation equations developed These classifiers used for Soybean Crop Yield
for the regions were validated for the future yield Prediction and results are compared. Bagging
prediction (2007, 2009 and 2010) and were found to algorithm gives maximum accuracy for soybean crop
be significant. yield prediction among REPTree, Random Forest,
Neural Network, SMO classifiers. Bayesian modelling
A study by [31] has done Crop Yield Prediction of used for wheat crop yield prediction using small
Wheat Using Fuzzy C Means Clustering and Neural dataset[41]. A study also identified drought related
Network. A research work [32] has implemented agronomic factors that significantly affect the
artificial neural networks (ANNs) for predicting crop variation in the rain fed wheat yield and relative
yields for an agricultural region in Nepal. In research, importance of these factors.
Agricultural data was collected from paddy field
cultivation in the Siraha district, an eastern region in Applications
Nepal. climatic parameters are rainfall, maximum
temperature and minimum temperature, fertilizer use Predicts crop yield production of rice for most Rice
were also used as input values. The result shows that producing states, India.
the trained neural network produced a minimum Predicting productivity of crop in various climatic
error. In [33] researcher has implemented neural conditions can help farmer and other partners in
networks for rice yield prediction in mountainous essential basic leadership as far as agronomy and
regions. A study by [34] reaching review of research product decision. It also helps to increase the yield of
devoted to applications of machine learning in crop.
agrarian generation frameworks. The review reports
the application of machine learning techniques with III. PERFORMANCE EVALUATION METRICS
artificial neural networks, Bayesian networks,
support vector machines and k-means. In [35], Accuracy: It is defined as the overall success rate of
researchers predict Rice Production in Phimai the classifier.
District, Thailand Decision Support System Using Accuracy = (TP + TN) / (TP + FP + FN + TN).
Artificial Neural Network. A study by [36] Where, True Positive (TP) depicts the number of
implemented C 4. 5 Algorithm to predict yield of instances where system detects for condition when it
crops like paddy, maize, wheat, soybean. Interactive is really present[1].
web pages developed for crop yield prediction.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com | UGC Approved Journal [Journal No : 64718]
504
Kolin Sukhadia, M. B. Chaudhari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 501-507

True Negative (TN) depicts the number of instances V. REFERENCES


where system does not detect a condition when it is
absent[1]. [1]. Shruti Mishra, Priyanka Paygude, Snehal
False Negative (FN) depicts the number of instances Chaudhary, Sonali Idate"Use of Data Mining in
where system does not detect a condition when Crop Yield Prediction", Proceedings of the
actually it is present[1]. Second International Conference on Inventive
False Positive (FP) depicts the number of instances Systems and Control (ICISC 2018), pp. 796-802,
where system detects a condition when it is really 2018
[2]. Umid Ku mar Dey, Abdullah Hasan Masud,
absent[1].
Mohammed Nazim Uddin, "Rice Yield Prediction
RMSE(Root Mean Square Error): It is defined as the
Model Using Data Mining", International
difference between the values predicted by the model
Conference on Electrical, Computer and
and the actual values noted[1].
Communication Engineering (ECCE), pp. 321-
MAE(Mean Absolute Error): It is another factor in
326, 2017
statistics which measures the difference between two
[3]. Niketa Gandhi, Leisa J. Armstrong, Manisha
continuous variables[1]. Nandawadekar, "Application Of Data mining
RAE(Relative Absolute Error): This measure gives the Techniques For Predicting Rice Crop Yield Semi-
total absolute error between the variables[1]. Arid Climate Zone Of India", 2017 IEEE
International Conference on Technological
IV. CONCLUSION AND FUTURE SCOPE Innovations in ICT for agriculture and Rural
Development, pp. 116-120, 2017
In rice crop yield prediction, various classification [4]. Ayush Shah, Akash Dubey, Vishesh Hemnani,
techniques are applied like support vector machine, Divye Gala, D. R. Kalbande, "Smart Farming
naïve bayes, bayesnet, locally weighted learning, System: Crop Yield Prediction Using Regression
ladtree etc. are applied on rice crop production Techniques", Proceedings of International
datasets. They have some drawbacks of low Conference on Wireless Communication, pp. 49-
classification accuracy and high errors. To get more 56, 2018
significant result and to get better classification [5]. Niketa Gandhi, Leisa J. Armstrong, Owaiz Petkar,
accuracy and to reduce errors like RMSE(Root Mean Amiya Kumar Tripathy, "Rice Crop Yield
Square Error), MAE(Mean Absolute Error), Prediction in India using Support Vector
Machines", 2016 International Joint Conference
RAE(Relative Absolute Error), boosting technique
on Computer Science and Software Engineering
Adaptive boosting technique is used in conjuction
(JCSSE)
with classification method Bayesnet in my research
[6]. Niketa Gandhi, Leisa J. Armstrong, Owaiz Petkar,
work and that will determine better accuracy of
"Predicting Rice Crop Yield Using Bayesian
prediction of rice yield in future. This survey
Networks", 2016 Intl. Conference on Advances in
concludes that the more accurate rice crop yield
Computing, Communications and Informatics
prediction model can be constructed by using (ICACCI), pp. 795-799, 2016
different data mining techniques. [7]. Akter Hossain, Mohammed Nazim Uddin,
Mohammad Arif Hossain, Yeong Min Jang,
"Predicting Rice Yield For Bangladesh By
Exploiting Weather Condition", 2017
International Conference on Information and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com | UGC Approved Journal [Journal No : 64718]
505
Kolin Sukhadia, M. B. Chaudhari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 501-507

Communication Technology Convergence Journal of Applied Research in Higher Education,


(ICTC), pp.589-594, 2017 Vol. 10 Issue: 1, pp.61-75
[8]. Dipika H. Zala, M.B. Chaudhari, ” Review on Use [16]. ADNAN MASOOD, "Bayesian Networks", A
of “BAGGING” Technique in Agriculture Crop BRIEF INTRODUCTION
Yield Prediction”, International Journal for [17]. Jiawei Han, Micheline Kamber, "Data
Scientific Research & Development| Vol. 6, Issue Mining:Concepts and Techniques"
08, 2018, pp.675-677. [18]. https://en.wikipedia.org/wiki/Naive_Bayes_classif
[9]. Niketa Gandhi, Leisa J. Armstrong, "Applying ier
Data Mining Techniques to Predict Yield Of Rice [19]. Friedman, N., Geiger, D. and Goldszmidt M.
in Humid Subtropical Climatic Zone Of India", (1997), ” Bayesian network classifiers”, Machine
2016 International Conference on Computing for Learning 29, 139–164.
Sustainable Global Development, pp.1901-1906, [20]. Castillo, G. (2006) , ”Adaptive Learning
2016 Algorithms for Bayesian Network Classifiers”,
[10]. Sagar S. Nikam , ”A Comparative Study of PhD Thesis, Chapter 3.
Classification Techniques in Data Mining [21]. Bouckaert, R. (2007), ” Bayesian Network
Algorithms”, ISSN April 2015. Classifiers in Weka”, Technical Report
[11]. Narayanan Balakrishnan, Dr.Govindarajan [22]. Heckerman, D. (1996), ” A Tutorial on Learning
Muthukumarasamy, ”crop Production - Ensemble with Bayesian Networks”, Microsoft Technical
Machine Learning Model for Prediction”, Report MSR-TR-95-06.
International Journal of Computer Science and [23]. G. Cooper, E. Herskovits (1992), ”A Bayesian
Software Engineering (IJCSSE), Volume 5, Issue 7, method for the induction of probabilistic
July 2016, pp. 148-153. networks from data, ”Machine Learning. 9(4):309-
[12]. Subhadra Mishra, Debahuti Mishra, Gour Hari 347.
Santra, "Adaptive boosting of weak regressors for [24]. N.Gandhi, L.J. Armstrong and O. Petkar, “Rice
forecasting of crop production considering Crop Yield Prediction in India using Machine
climatic variability: An empirical assessment", Learning Techniques”, communicated, 2016.
Journal of King Saud University - Computer and [25]. Weka 3:Data Mining Software in Java, Machine
Information Sciences, pp. 1-23. Learning Group at the University of Waikato,
[13]. Philip G. Oguntunde, Gunnar Lischeid, Ottfried Official
Dietrich, "Relationship between rice yield and Web:http://www.cs.waikato.ac.nz/ml/weka/index
climate variables in southwest Nigeria using .html, accessed on 26th March 2016.
multiple linear regression and support vector [26]. S. Puteh, M. Rizon, M. Juhari, J. Nor Khairah, S.
machine analysis", August 2017, springer. Siti Kamarudin, B. Aryati, R. Nursalasawati, “Back
[14]. Haresh L. Siju, Pinal J. Patel, "Review on Crop propagation algorithm for rice yield prediction”,
Yield Prediction using Data Mining Focusing on Proceedings of 9th of the Ninth International
Groundnut Crop and Naive Bayes Technique", Symposium on Artificial Life and Robotics,
International Journal of Scientific Research in Beppu, Japan, Oita, pp. 586-589, 2004.
Computer Science, Engineering and Information [27]. Deepti Gupta, Udayan Ghose, " A Comparative
Technology, Volume 3 | Issue 1, ISSN : 2456- Study of Classification Algorithms for Forecasting
3307, pp.1947-1954 Rainfall ", IEEE 4th International conference on
[15]. Olugbenga Wilson Adejo, Thomas Connolly, Reliability, Infocom Technologies and
"Predicting student academic performance using Optimization (ICRITO) (Trends and Future
multi-model heterogeneous ensemble approach", Directions), 2015, pp. 1-6.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com | UGC Approved Journal [Journal No : 64718]
506
Kolin Sukhadia, M. B. Chaudhari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 501-507

[28]. V. B. Nikam and B. B. Meshram, "Modeling [36]. Veenadhari, B. Misra, and C. D. Singh, "Machine
Rainfall Prediction Using Data Mining Method: A learning approach for forecasting crop yield based
Bayesian Approach", Fifth International on climatic parameters, " 2014 Int. Conf. Comput.
Conference on Computational Intelligence, Commun. Informatics Ushering Technol.
Modelling and Simulation, Seoul, 2013, pp. 132- Tomorrow, Today, ICCCI 2014, pp. 1-5, 2014.
136. [37]. E. Manjula, S. Djodiltachoumy, ” A Model for
[29]. Niketa Gandhi, Leisa J. Armstrong, “Rice Crop Prediction of Crop Yield”, International Journal
Yield Forecasting of Tropical Wet and Dry of Computational Intelligence and Informatics,
Climatic Zone of India using data mining Vol. 6: No. 4, March 2017, pp.298-305.
Techniques”, 2016 IEEE International Conference [38]. Sellam and E. Poovammal, "Prediction of Crop
on Advances in Computer Applications (ICACA) Yield using Regression Analysis, " vol. 9, no.
[30]. Rupanjali D. Baruah, R.M. Bhagat, Sudipta Roy, October, 2016.
L.N. Sethi, ”Use of Data Mining Technique for [39]. B. H. Dhivya, R. Manjula, S. B. S, and R.
Prediction of Tea Yield in the Face of Climate Madhumathi, "A Survey on Crop Yield Prediction
Change of Assam, India”, 2016 International based on Agricultural Data, " pp. 4177-4183, 2017.
Conference on Information Technology, pp. 265- [40]. Savla and A. Mandholia, "Survey of classification
269. algorithms for formulating yield prediction
[31]. Akanksha Verma, Aman Jatain, Shalini Bajaj, ” accuracy in precision agriculture, " 2015.
Crop Yield Prediction of Wheat Using Fuzzy C [41]. Yildirak, Z. Kalaylioglu, and A. Mermer,
Means Clustering and Neural Network”, "Bayesian estimation of crop yield function :
International Journal of Applied Engineering drought based wheat prediction model for tigem
Research ISSN 0973-4562 Volume 13, Number 11 farms, " Environ. Ecol. Stat., vol. 22, no. 4, pp.
(2018) pp. 9816-9821. 693-704, 2015.
[32]. Tirtha Ranjeet, Leisa Armstrong, ” An Artificial [42]. Mrs. K. E. Eswari, Ms. L.Vinitha, ” Crop Yield
Neural Network for Predicting Crops Yield in Prediction in Tamil Nadu using Baysian
Nepal”, Proceedings of AFITA 2014 - Copyright Network”, International Journal of Intellectual
ASICTA Inc, pp. 376-386. Advancements and Research in Engineering
[33]. B. J I , Y. SUN, S.YANG, J. WAN1, ” Artificial Computations, Volume-6, Issue-2, pp.1572-1576.
neural networks for rice yield prediction in [43]. Niketa Gandhi, Owaiz Petkar, Leisa J. Armstrong,
mountainous regions”, Journal of Agricultural "Rice Crop Yield Prediction Using Artificial
Science (2007), pp. 249–261. Neural Networks", 2016 IEEE International
[34]. Ankita Bissa, Meena Kushwaha , Mayank Patel, Conference on Technological Innovations in ICT
”A Survey of Machine Learning Appliacations In For Agriculture and Rural Development (TIAR
Decision Making To Improve Farming”, 2016), pp.105-110
International Journal of Computer Sciences and
Engineering, Vol.-6, Issue-9, Sept. 2018, pp. 669- Cite this article as : Kolin Sukhadia, M. B. Chaudhari, "A
683. Survey on Rice Crop Yield Prediction in India Using
[35]. Saisunee Jabjone, Sura Wannasang, ” Decision Improved Classification Technique ", International Journal
Support System Using Artificial Neural Network of Scientific Research in Computer Science, Engineering
and Information Technology (IJSRCSEIT), ISSN : 2456-
to Predict Rice Production in Phimai District,
3307, Volume 5 Issue 1, pp. 501-507, January-February
Thailand”, International Journal of Computer and
2019.
Electrical Engineering, Vol. 6, No. 2, April 2014,
Available at doi : https://doi.org/10.32628/CSEIT1951122
pp. 162-166.
Journal URL : http://ijsrcseit.com/CSEIT1951122

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com | UGC Approved Journal [Journal No : 64718]
507
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951131

A Review on Various Road Safety Systems


Chandrashekhar Rathod1, Sneha Wakde1, Roshan Punwatkar1, Rupesh Sharma1, Rupal Patil1, Prof. Rohan
Kokate2
1BE Student, Department Computer Science and Engineering, J. D. College of Engineering and Management,

Nagpur, Maharashtra, India


2 Assistant Professor, Department Computer Science and Engineering, J. D. College of Engineering and
Management, Nagpur, Maharashtra, India
ABSTRACT

The Safe System (SS) approach to road safety emphasizes safety-by-design through ensuring safe vehicles, road
networks, and road users. With a strong motivation from the World Health Organization (WHO), this
approach is increasingly adopted worldwide. Considerations in SS, however, are made for the medium-to-long
term. Our interest in this work is to complement the approach with a short-to-medium term dynamic
assessment of road safety. Toward this end, we introduce a novel, cost-effective Internet of Things (IoT)
architecture that facilitates the realization of a robust and dynamic computational core in assessing the safety of
a road network and its elements. In doing so, we discuss about the various existing mechanisms for road safety.
Keywords : Internet of Things, Smartphone, Road Safety, Arduino, G-Sensor, Accelerometer

I. INTRODUCTION star Corporation provides subscription based


communications, in–vehicle security, hand free
Road traffic crashes are one of the world’s largest calling, turn-by-turn navigation, and remote
public health and injury prevention diagnostics systems throughout the united states,
problem .According to the World Health Canada and China turn-by-turn navigation and road
Organisation (WHO), more people die on road in side assistance .The motivation behind the work
India than anywhere else in the world. At least 13 revolution is an attempt to make an embedded
people die every hour in road accidents in our system to bring a positive difference in the field of
country; the latest report of the national crime record road safety and road discipline.
Dureau reveals.
The work tackles some major causes of road accidents
In 2007, 1.14 lakhs people in India lost their lives in such as breaking traffic signal and horning in No
road mishaps improper road infrastructure failure to horning zone. It also has a major objective of
follow the speed limit, and increase in drinking and exercising road discipline such as speed control in
driving habits are among the major factors different areas and horn control in horn prohibited
contributing to deaths from road crashes, WHO said zones .This paper presents vehicle speed control in
in its report on ‘Decade of Action for Road Safety variable zone in this feature ;speed of the vehicle is
2011-2010’.Currently road safety systems are controlled in different areas such as
available in high end luxury cars such as Audi, flyovers ,bridges ,highways ,schools ,hospitals ,cities
Mercedes, Benz, etc., to name a few. Example: On and suburbs .Horn control of vehicle. Horn control of

CSEIT1951131 | Received : 01 Feb 2019 | Accepted : 16 Feb 2019 | January-February -2019 [ 5 (1) : 508-513 ]
508
Chandrashekhar Rathod et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 508-513

vehicles in no honking zone–controlling unwanted cushioned impact at a speed of more than 30 km/h,
disturbances in horn prohibited zone such as with lower speeds resulting in either death or serious
hospitals, public libraries ,courts, schools, etc., injury [1, 2]. The objective of the SS approach is that
the three model elements should be designed and
The Safe System (SS) approach to transport networks monitored to proactively prevent deadly speeds from
originated with the “Safe Road Transport System” happening and allow for a reduced emergency
model developed by the Swedish Transport Agency. response time in the event of an accident.
In its essence, the approach migrates from the view
that accidents are largely and automatically the Elements of the SS approach are as follows.
driver’s fault to a view that identifies and evaluates
the true causes for accidents. Through the (1) Safe Vehicle. Emphasis on vehicle safety is
categorization of safety into the safety of three verified through mandated regulatory testing and
elements (vehicle, road, and road user), SS minimizes rating, as well as technologies such as electronic
fatalities and injuries by controlling speeds and stability control. Beyond this, enforced checks (e.g.,
facilitating prompt emergency response. The model upon license renewals) combined with on the road
has been widely adopted since its introduction and is reporting work to review the status of vehicle safety.
currently motivated by the WHO as a basis for road
safety planning, policy-making, and enforcement. (2) Safe Road. The assessment of road (or road
network) safety is multifaceted. Road inspection
enables clear and direct observation of the state of the
road and assesses the need for repairs or
modifications. The structure of the road network is
amenable to safety assessment through partitioning
into what is called “Traffic Analysis Zones (TAZs)”
[4]. In addition, considerations for crash data and
other supporting data offer further insights into
general safety assessment.

In 2011, the European Road Assessment Programme


(EuroRAP) generated the European Road Safety Atlas
for EU countries [5]. The atlas indicated the safety
level of roads with a star rating based on specially
equipped vehicles for multimedia-based data
aggregation [6]. The EuroRAP efforts continue to
Figure 1 : The Safe-System-Based Safe Road
implement an SS approach across the EU, along with
Transport Systems
several other national programmes within the
International RAP, or iRAP, initiative [7].
An illustration of the model is provided in Figure 1.
A central emphasis is given to speed in the SS
(3) Safe Road User. There are several aspects to road
approach as it is the strongest and most fundamental
user safety, including measures for education and
variable in the outcome of fatality. The fragility of
awareness, travel distance, exposure, licensure,
the human body makes it unlikely to survive a
enforcement, and sober driving [3]. The need for

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


509
Chandrashekhar Rathod et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 508-513

such characterization rises substantially as the Anto bennet et al., [13] had proposed that a review
findings of crash report analysis in cities typically on an Embedded System in Passenger Car for Road
note a critical dependence on either driver behavior Safety using GPS and GSM.GPS Receiver gets the
or driver awareness [8]. A great need is further location information from satellite in the form of
established in these studies for innovative latitude and longitude. The system can be
mechanisms to instill safe driving at the licensing and interconnected with the alcohol detection and alert
post-licensing stages. the owner on his mobile phone.

II. LITERATURE REVIEW Anto bennet et al., [14] had proposed that a review
on Alcohol Detection and Accident Avoidance Using
Pramodshelke et al. [9] had proposed that a review on locking With Tracking. This system will detect
Arduino Based embedded System in Car for Road drunker driver by alcohol sensor through driver
Safety using RFID .In this work when the RFID tag is breathe fitted on steering in front of driver ,the
detected by the RFID reader the speed of the motor message is sent to the police through GSM system
reduces in specified areas and it avoids accidents. and provides GPS base system to track those cars.

Ch . Sonali Shankar et al., [10] had proposed that a III. RELATED WORK
review on Arduino based Embedded System in
Passenger Car for Road Safety .In this work the sound Many countries around the world ban the use of
of the horn stops and speed is also reduced when the hand-held devices while driving, including the
RFID tag is detected by the RFID reader. United States, Canada, Germany, Israel, Greece,
Norway, and more. While smartphones can be a
Deepa et al., [11] had proposed that a review on dangerous distraction, correct use of cell phone apps
Embedded System in Passenger Car for Road can reduce risk and promote safer driving behavior.
Safety.in this work the proposed features are Apps used with hands-free devices can monitor
automatic collision notification that gives notification distractions, protect young drivers, call for help when
to the victim’s relative, red light traffic control makes needed, and prevent speeding. Following are few Top
sure the vehicle doesn’t break signal, speed control Road Safety Apps.
alters speed in different zones, horn control prevents
honking in horn prohibited zone, alcohol detection Kruzr – Driving Assistant for Distracted Driving
detects drunk driving and vehicle security is used to Kruzr Road Safety App Kruzr is a free android-based
prevent theft Suhas Katkar et al., [12] had proposed app that detects when you’re driving and
that a review on An Embedded System In passenger automatically screens incoming calls and messages.
Car For Road Safety. This project is designed to Kruzr tells senders you’re on the road and will
inform about an accident or theft that has occurred to respond later. Don’t worry, you won’t miss critical
the vehicle, to the family members of the travelling calls. Kruzr automatic driving assistant allows urgent
persons and concerned authorities. Anto bennet et al., calls to proceed, while less important calls and
had proposed that a review on An Embedded System messages can be retrieved later.
in Passenger Car for Road Safety using GPS and GSM.
This work is designed to inform about an accident Safely Home – Detects Accidents
that is occurred to a vehicle to the family members of Safety Home Road Safety App Safely Home is a free
the travelling persons. Android app that automatically detects road accidents

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


510
Chandrashekhar Rathod et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 508-513

without any human intervention. If you’re in an such as “Driving right now, will get back to you
accident, the app notifies your emergency contacts. shortly”.
Safely Home includes a “Save a Life” feature to help
injured accident victims get help quickly. The app Fleetsafer
also provides traffic updates, and GPS-guided Fleetsafer is aimed at employers who want to reduce
Roadside Assistance for mechanical problems. the risk of employee liability for accidents on the job.
The app appears on company-issued tablets and
Drive Mode: Safe Driving App smartphones. It detects the driving state and
Drive Mode Road Safety app Drive Mode is an automatically puts the device into safe mode. Only
Android app that includes a voice-control option for emergency calls may be made.
streamlined interfacing with your favorite
applications, as well as simple swipe or tap MamaBear
functionality. With Drive Mode voice control, MamaBear Family Safety is a free app for Android
drivers can easily access navigation, music and and Apple devices. Recognized as the “Best Parenting
messaging apps. App” in 2017 by Mumi Family Awards, MamaBear
Drive Mode can be setup to launch when you start provides parental alerts for speeding, and provides
driving. The app will ignore calls and messages in GPS locations for young drivers. MamaBear also
“Do Not Disturb” mode, send text message auto allows parents to monitor their child’s text messages,
replies, or you can choose to reply to calls and social media, and arrival and departure times at
messages using seamless voice commands, and much locations away from home.
more.
Drive Safely Pro Road Safety App
AT&T DriveMode The DriveSafe.ly app is designed to prevent distracted
ATT Road Safety App driving accidents by providing hands-free access to
AT&T Drive Mode is a free app for Android and emails and text messages. Powered by a text-to-
Apple users and is available on all carriers. This app speech program, Drive Safe.ly Pro converts messages
silences incoming messages when the GPS detects in 27 languages. The app reads text messages and
you are driving 15mph or faster, and turns off when emails aloud in real time, allowing drivers to keep
you drop below 15 mph for two or three minutes. It their hands on the wheel and eyes on the road.
offers you a list of up to five contacts you can call
hands-free. Parents of young drivers can receive In this system, various sensors such as soil moisture,
alerts if the app is turned off. AT&T Drive Mode is DHT11 sensors are connected to the input pins of
also available in Spanish. arduino microcontroller. The sensed values from the
sensors are displayed in LCD. If the sensed value goes
iBolt Dock’n Drive beyond the threshold values set in the program, the
iBolt Road Safety App The iBOLT Dock’n Drive app pump will be automatically switched ON/OFF by the
requires the use of an iBolt docking station. This relay circuit and it is connected to the driver circuit
Android-based app provides automated answering of which helps to switch the voltage. The farmer will be
incoming calls. The “It Can Wait” feature places intimated about the current field condition through
incoming text messages on hold while the phone is GSM module and also updated in the web page. By
docked. Messages are delivered to the phone when using this system, the farmer can access the details
it’s not docked. You can set the auto reply message, about the condition of the field anywhere at any time.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


511
Chandrashekhar Rathod et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 508-513

IV. CONCLUSIONS [7] International Road Assessment Programme


(iRAP), “iRAP,” https://www.irap.org/.
This work illustrates the viability of an economic [8] H. M. Hassan and H. Al-Faleh, “Exploring the
road safety monitoring and assessment solution risk factors associated with the size and severity
through exploiting advances in the Internet of Things of roadway crashes in Riyadh,” Journal of
(IoT) within the context of smart cities. The Safety Research, vol. 47, pp. 67–74, 2013. View
introduced architecture facilitates robust and at Publisher·
dynamic road safety assessment that complements [9] Prof.Pramodshelke, Ram Anjankar,
the Safe System approach motivated by the World SakesheeWanare, Ganesh Surve“A Review on
Health Organization (WHO), which has been Accident Avoidance by Eye Blink Sensor and
increasingly adopted worldwide. An application of Alcohol Sensor Using GSM” E-ISSN:2321-9637,
the dynamic assessment framework for route 09th April 2017
planning is also demonstrated. Future work involves [10] Sonali Shankar Chalwad, Snehal Bhimrao
exploring further applications, especially in the Gaikwad, PrashantA.Chougule “Accident
context of raising driver awareness of the road safety Prevention Using Eye Blink Sensor”ISSN:2393-
conditions during their trips. 2835, Volume-4,Issue-1, January-2017
[11] SuhasKatkar,MaheshManikKumbhar, Priti
V. REFERENCES Navanath Kadam “Accident Prevention System
Using Eye Blink Sensor” Volume:03, Issue:05,
[1] F. Wegman, “The future of road safety: A May-2016
worldwide perspective,” IATSS Research, vol. [12] Pratik S. Danage, Ajay B.Dongare,
40, no. 2, pp. 66–71, 2017. View at Publisher · SangramsinhS.Dongare, Rahul B.Ghogare,
View at Google Scholar · Ganesh N.Khare “Automatic Breaking Sysetm
[2] World Health Organization, “Save LIVES - A Using Eye Blinking Sensor”Vol.2, Issue 2,
road safety technical package,” 2017. pp(166- 170),October2015- March-2016
[3] W. E. Marshall, “Understanding international [13] Deepa K. B, Chaitra M, Ankit Kumar
road safety disparities: Why is Australia so ShrmaSreedar V. S, Prashant Kumar H. R.”
much safer than the United States?” Accident Accident prevention by eye binking sensor and
Analysis & Prevention, vol. 111, pp. 251–265, alcohol detector” Volume-4 , Issue-7 , pp: 351-
2018. View at Publisher · 354, July-2015
[4] X. Wang, X. Wu, M. Abdel-Aty, and P. J. [14] Dr.AntoBennet, M , Sankaranarayanan S,
Tremont, “Investigation of road network Ashokram S ,Dinesh Kumar T R,“Testing of
features and safety performance,” Accident Error Containment Capability in can Network”,
Analysis & Prevention, vol. 56, pp. 22–31, 2013. International Journal of Applied Engineering
View at Publisher · View at Google Scholar · Research, Volume 9, Number 19 (2014) pp.
[5] European Road Assessment Program 6045-6054
(EuroRAP), “European Road Safety Atlas,” [15] Dr.AntoBennet, M, SankarBabu G, Natarajan S,
http://atlas.eurorap.org/. “Reverse Room Techniques for Irreversible
[6] European Road Assessment Programme Data Hiding”, Journal of Chemical and
(EuroRAP), “Star Ratings,” Pharmaceutical Sciences 08(03): 469-475,
http://www.eurorap.org/protocols/star-ratings/. September 2015.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


512
Chandrashekhar Rathod et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 508-513

[16] Dr. AntoBennet, M ,Sankaranarayanan S,


SankarBabu G, “ Performance & Analysis of
Effective Iris Recognition System Using
Independent Component Analysis”, Journal of
Chemical and Pharmaceutical Sciences 08(03):
571-576, August 2015.r.

Cite this article as :

Chandrashekhar Rathod, Sneha Wakde, Roshan


Punwatkar, Rupesh Sharma, Rupal Patil, Prof. Rohan
Kokate, "A Review on Various Road Safety Systems",
International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 508-513, January-February 2019.
Journal URL : http://ijsrcseit.com/CSEIT1951131

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


513
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951129

Review on Various Tools and Techniques for


Plagiarism Detection
Kedar Bhujade1, Tushar Bhandirge1, Rohan Khobragade1, Saurabh Shelare1, Jayprakash Shukla1,
Prof. R. S. Thakur2
BE, Department of Computer Science & Engineering, Dr. Babasaheb College of Engineering & Research,
1

Nagpur, Maharashtra, India


2Assistant Professor, Department of Computer Science & Engineering, Dr. Babasaheb College of Engineering &
Research, Nagpur, Maharashtra, India

ABSTRACT

Being a developing issue, plagiarism is commonly described as literature theft and academic dishonest nature in
the writing, and it must be avoided and adhere to the moral standards. Plagiarism occur in scholastics, paper
publication, music, work of art developing quickly, so the recognizing plagiarism is essential. While the most
recent couple of year’s plagiarism detection tools have been utilized predominantly in research conditions,
refined plagiarism programming and instruments are presently quickly rising. In this paper, we give an outline
of various plagiarism programming and apparatuses to take care of the plagiarism issue. We propose an element
classification conspire that can be utilized to examine plagiarism discovery programming and plagiarism
recognition instruments. This plan depends on the product's general qualities, devices qualities, and apparatuses
property.
Keywords : Plagiarism Detection, Plagiarism Types, Plagiarism Techniques, Plagiarism Algorithms

I. INTRODUCTION content, and reuse thoughts with/without referring


to them and others. Recognition of plagiarized record
The term “plagiarization” is characterized by the fact takes on vital jobs in many applications, for example,
that one takes thoughts, archives, code, etc. from document the executives, copyright assurance, and
some another and passes them without specific plagiarism aversion. Existing conventions expect that
reference as one`s own. Plagiarism is therefore a the substance of records put away on a server is
global issue in many aspects of our lives. There are a straightforwardly open. This presumption constrains
wide variety of types of plagiarism, and plagiarism in progressively down to earth applications, e.g.,
academies can be a deeply deterrent to educators and identifying copied reports between two meetings,
undergraduate study. On the off chance that where entries are confidential [2].
plagiarism isn't tended to adequately, literary thieves
could increase undeserved preferred standpoint, for Plagiarism can be one of the best known types, for
example, more checks for their tasks with less efforts. example, to replicate the whole or some portion of
the record, to rephrase the same substance in
Different types of plagiarism [1] are included: Use different words, to use the thoughts and ideas of
source without valid reference to them, summarize others to refer to incorrect or non-existent source [3].

CSEIT1951129 | Received : 15 Feb 2019 | Accepted : 26 Feb 2019 | January-February -2019 [ 5 (1) : 514-518 ]
514
Kedar Bhujade et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 514-518

Various plagiarism methods include deciphered 1. Web enabled systems


plagiarism, in which substance is interpreted and Web-enabled system are more widely used because
used without specific reference to the first work, they easily and reliably extended their search for
masterful plagiarism, in which distinguishable media, plagiarized resources to the global web. The
including image and recordings, show other people`s following are the two web-enabled screening systems,
work without legitimate reference [3].
✓ EduBirdie
A plagiarized code (also referred to as clone code)
It provides us with a report showing the percentage
which can be described as a reuse of the source code
of content that is unique. One can check any form of
without explicit consent or reference. So a plagiarized
text using this tool, whether it is an essay, academic
program can be described as a program developed
paper, Technical descriptions, case studies, product
with a few routines from some other program
details or white paper.
changes, routine changes, regularly message
substitutions, don't require a definite comprehension
✓ Safe Assign
of the program. Tragically, significant class sizes have
made the plagiarism of programming tasks less
It compares uploaded text document with a set of
demanding. research papers to identify areas of overlapping
between the uploaded document and available work.
In extensive university courses, the plagiarism of PC It is based on a distinctive text matching algorithm
projects can be very fairly normal. A plagiarized that detects the exact and inaccurate matching of a
program with an alternative visual appearance can be paper with the uploaded document. It compares
delivered with a bunch of fundamental manager submissions to several databases such as Inform
activities. This makes it very difficult to manually journal Database.
recognize plagiarized programs in significant classes.
Every one of this plagiarism procedures has a 2. Stand-alone systems
significant impact on the process of education. This software can be installed in the computer system.
Therefore, how can we ensure the management of Some of stand-alone system are, Their paper showed
how the instrument and also the experience of
plagiarism frameworks and the recognition and
using it in four different programming task in the
management of plagiarism? A basic issue requires
last 12 years.
Computer researchers to arrange.
Hermann et al. [7] state that plagiarization is
II. PLAGIARISM IN DOCUMENTS robbing someone else's work of credit. As per the
creators, content plagiarism implies that a creator
Documentary plagiarism is more relevant to the simply duplicates it with giving it the real credit.
academic purpose of the student community, They represent the main attempts to acknowledge
especially the postgraduate who modifies the plagiarized portions of a content using measurable
available documents and presents it as his own. This models of dialect and perplexity. The
should be prevented as it affects the quality of the investigations were carried out on two specific and
ability of the students themselves. It is therefore academic corporation. The first record and
necessary to detect plagiarism in documents first and linguistic form and stemmed adaptation were
for this purpose the following systems can be used, contained in the two specific works. The

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


515
Kedar Bhujade et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 514-518

plagiarism on these reports was distinguished and institutions such as Ohio University, Umass Boston
the results were checked. and Trinity College Dublin.

Jinan et al. [8] concentrated on the instructive setting and ✓ WCopyFind


confronted comparable difficulties. They show the most
competent method for checking cases of plagiarism.
What's more, they intended to fabricate learning networks
This system is plagiaristic between two or more tasks.
of understudies, educators, organization, and personnel
and staff all teaming up and developing solid connections
Many commercial tools for detecting plagiarism are
that give the establishment to understudies to accomplish available. Table 1 presents the tools` comparative
their objectives with more noteworthy achievement. They analysis.
also advanced the sharing of data. In a straightforward,
customizable and reusable way, they gave consistent
Table 1. Comparison software based on their
combination heritage and various different applications.
different features
Learning gateway may give a help device to this
learning framework. In any case, fabricating and
adjusting learning gateway is certifiably not a simple
errand. This paper recognizes the plagiarism of java
understudy assignments in the product.

Plagiarism can be differentiated by an understanding


of sentences and paragraphs from paper, which can
also be found with the help of web indexes. They
pointed this out to create a free software that can be
used to identify plagiarism in their classed by any
teacher or encouraging partner. Nathaniel et al. [9]
characterize plagiarism as a major problem affecting
III. LITERATURE REVIEW
copyrighted records/materials. They state that
plagiarism is expanded nowadays because of the
Allan et al. [5] exhibited a system for identification of
productions in on the web. They suggested a new
plagiarism. The development of the web, with
discovery technique for plagiarism called SimPaD.
bottomless data online, exacerbates the issue even.
The reason for this strategy is to contrast sentence by
The creators have discovered four diverse approaches
sentence to create similarities between two archives.
to approach plagiarism discovery. They continued to
Examinations show that SimPaD increasingly
pursue comprehensive looking and took the center
accurately identifies plagiarized reports and outstrips
ground technique instead of thoroughly or
existing approaches to plagiarism recognition.
accidentally searching for sentences on the web in a
study paper. They found the manifestation of thought
✓ Plagiarisma Checker X
they had acquired.
It is a simple system for undergrads, educators,
content producers, SEO experts and site owners to Francisco et al. [6] state that research facility work
verify that others have copied their work. According assignments are essential for software engineering
to the programmer, its clients include educational learning. The study showed that 400 understudies
duplicate a similar research in illustrating their

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


516
Kedar Bhujade et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 514-518

assignment at the same time during the last 12 years. document is mark as plagiarism document and
This has made the instructors to give careful similarity percentage is calculated.
consideration on finding the plagiarism. In this way,
they constructed a discovery device for plagiarism.
This device had the full range of tools to help
administrator to manage the work of the research
facility. To quantify the similarities between two
assignments, they used four comparability criteria.

IV. PROPOSED SYSTEM

Fig. 1 shows the proposed system architecture. In our


proposed approach user inputs a single document for
plagiarism checking. Initially pre-processing is
performed on document in which unnecessary space
within document, special characters, etc. are
removed and then stopword removal process is
performed in which the keywords such as a, an, the,
numbers in documents & other stopword are
removed. Then stemming processed is performed in
which ing, ed, etc. of each keyword is removed. At
the end only dictionary keywords are remain in input
document.

After getting dictionary keyword from document,


important keywords separated out (keywords having
count greater than threshold k). These top k keyword
set is passed to neural network classifier which
performs classification on previously stored
documents in database in two classes such as
documents containing top k keywords (say class 1)
and documents which don’t contain top k keywords
(say class 0). Then we use documents containing top
k keywords (class 1) for further processing.

Figure 1. System Architecture


After this, the document vector of Input document
and class 1 document is generated. Then TF-IDF of
V. CONCLUSION
all document is generated and finally cosine
similarity is calculated between input document and
In this paper we study plagiarism detection is very
class 1 documents. If similarity is found between
important not only in academics, but also in industry,
input document and any other document then input
music, artworks, etc. In this study, in particular, it

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


517
Kedar Bhujade et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 514-518

was shown how the problem of plagiarism can be Plagiarism in Programming Assignments,” IEEE
addressed using different techniques and tools. In this Transactions on Education, vol. 51, no. 2, pp.
paper, we have seen that different software and tools 174-183, 2008.
are available to check plagiarism. Comparison of [7]. Hermann M., Frank K., and Bilal Z.,
software and tools has shown that they still have no “Plagiarism -A Survey,” Universal Computer
software and tool that can detect that the document Science, vol. 12, no. 8, pp. 1050-1084, 2006.
has been plagiarized at 100 percent, since each [8]. Jinan F., Alkhanjari Z., Mohammed S., and
software and tool has advantages and limitations Alhinai R., “Designing a Portlet for Plagiarism
according to the features and performance described Detections within a Campus Portal,” Journal of
in the table. However, this software has limitations, Science, vol. 1, no. 1, pp. 83-88, 2005.
tools that have a significant impact on the success of [9]. Nathaniel G., Maria P., and Yiu N., “Nowhere
plagiarism detection. We also presented our proposed to Hide: Finding Plagiarized Documents Based
to the detection of plagiarism. on Sentence Similarity,” in Proceedings of
IEEE/WIC/ACM
[10]. International Conference on Web Intelligence
VI. REFERENCES and Intelligent Agent Technology, NSW, pp.
690-696, 2008.
[1]. P. OGR, “What is Plagiarism?”, [On Line]
http://www.plagiarism.org/,Retrieved Nov. 15, Cite this article as :
2010
[2]. C. Lyon, R. Barrett, and J. Malcolm, “Plagiarism Kedar Bhujade, Tushar Bhandirge, Rohan
is Easy, but also easy to detect.” Cross- Khobragade, Saurabh Shelare, Jayprakash Shukla,
Disciplinary Studies in Plagiarism, Fabrication, Prof. R. S. Thakur , "Review on Various Tools and
and Falsification, 2006. Techniques for Plagiarism Detection", International
[3]. L. Romans, G. Vita, and G. Janis, “Computer- Journal of Scientific Research in Computer Science,
based plagiarism detection methods and tools: Engineering and Information Technology
an overview,” the 2007 international (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
conference on Computer systems and 514-518, January-February 2019.
technologies. 2007, ACM: Bulgaria. Journal URL : http://ijsrcseit.com/CSEIT1951129
[4]. S. Mann and Z. Frew, “Similarity and
originality in code: plagiarism and normal
variation in student assignments,” the 8th
Australian conference on computing education,
2006.
[5]. Allan K., Kevin A., and Bruce B., “An
Automated System for Plagiarism Detection
Using the Internet,” in Proceedings of World
Conference on Educational Multimedia,
Hypermedia and Telecommunications,
Chesapeake, pp. 3619-3625, 2004.
[6]. Francisco R., Antonio G., Santiago R., Jose L.,
Pedraza M., and Manuel N., “Detection of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


518
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951116
An Elementary Study on Various Techniques Involved in Face Recognition
Systems : A Review
Payal Maken
School of Computer Science & Engineering, Shri Mata Vaishno Devi University, Kakryal, katra, Jammu &
Kashmir, India
ABSTRACT

Face recognition has now become one of the interesting fields of research and has received a substantial
attention of researchers from all over the world. Face recognition techniques has been mostly used in the
discipline of image analysis, image processing, etc. One of the face recognition techniques is used to develop a
face recognition system to detect a human face in an image. In face recognition system a digital image with a
human face is given as an input which extracts the significant features of face such as (eyes, nose, chin, cheeks,
etc) to recognize a face in a digital image which is an exhausting task. Security of information is very salient
feature and is difficult to achieve. Security cameras are present in offices, universities, banks, ATMs, etc. All
these security cameras are embedded with face recognition systems. There are various algorithms which are
used to solve this problem. This paper provides an overview of various techniques which are often used for this
face recognition in a face recognition system. This paper is divided into five parts, first section concludes
various face detection techniques, second section describes about image processing ,third section have details
about face recognition techniques, fourth section describes various classification methods and last section
concludes all of these sections.
Keywords : Face Recognition, Face Detection, Features Extraction, Pre-Processing

I. INTRODUCTION passwords; person identity numbers (PINs), etc. But


identity cards can be easily stolen or misplaced;
Face recognition techniques is used in wide variety of passwords and PINs can be forgotten. However
face recognition systems. A face recognition person’s biological features can never change,
technique involves identification of faces and then misplaced or stolen. So, this technology can give
comparing it with the images in the database. Human solution to this problem since a face of a person is
beings can easily differentiate the human face on the undeniably connected to a person only except in the
basis of various factors. But it is complicated to case of twins. For the security purpose biometric
develop such a computer system which can imitate techniques gives better results than the identity cards,
human capabilities of recognition and detection of a passwords, etc. Various biometric techniques
face in a digital image. In recent years we can see that comprises of fingerprint impression, iris detection,
researches related to face recognition have been DNA detection and face recognition. Among all Face
escalating, and this is so because of the need of recognition is most challenging and strenuous area in
security in various areas. Person’s identification is this area.
most important for the purpose of security. We can
identify a person through its identity cards,

CSEIT1951116 | Received : 15 Feb 2019 | Accepted : 27 Feb 2019 | January-February -2019 [ 5 (1) : 519-525 ]
519
Payal Maken Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 519-525

Face recognition system basically consists of Each year new techniques are developing for this
following blocks: purpose.

II. RELATED WORK


1. Input image
2. Facial detection A. Face Detection
3. Image processing Face detection is the first step in any face analysis
4. Facial recognition system. Face detection involves finding a face in any
a. Features extraction given input image. So, basically it means to find out
5. Verification/Validation where is the face irrespective of the background. For
example in a figure 2, there is a photo frame over
A. Input Image
there but what it tries to do is extract from an image
A physical or behavioral sample of image is captured just a segment which represents a face.
from any physical image capturing device, which
serves as an input the system.

B. Facial detection
It basically means locating a face in a given image.

C. Image processing
Images are cropped to avoid facial image remains and
Figure 2. Detecting a face in an image
images are converted into black and white and then
There are number of techniques for detecting a face
to gray images.
in an image.
D. Facial recognition
1. Voila Jones face detection
It means to extract the features of a face (eyes, nose,
This technique was given by Paul Viola and Michael
etc) and make a template of it.
Jones. It is also known as haar cascade algorithm, this
E. Verification algorithm detect a face in an image on the basis of
System recognizes a given individual and gives a various features like eyes, nose, upper body, lower
decision. body. A system is designed by giving input some faces
and some non-faces and training a classifier. This
F. Validation
classifier is known as haar cascade classifier. Now a
In this system compares a given individual face with
cascade classifier is something that tells what we have
faces in the database to identify a given face.
to look in an image for detecting a face. For example,
if we are distinguishing whether a person is
overweight or not, then weight is the feature that is
to be considered not their age or skin color [1]. So we
basically train a classifier to detect a face. The data is
Figure1. Block diagram of face recognition system then stored in a file in the system. By referring to
Figure 1 shows the block diagram of face recognition that trained data one can detect a face.
system. Nevertheless ample work has been done on This algorithm includes following:
face recognition problem, but still not any work is up • Haar features
to the mark for the implementation point of view. • Integral image

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


520
Payal Maken Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 519-525

• Adaboost this we understand that this pattern is absorbed in


• Cascading this picture. Viola Jones uses 24*24 window as a base
window size to start evaluating these features in any
a. Haar features
given image. Considering all possible parameters of
A haar cascade is a series of “haar-like-features” that the haar like features we get about 16000+ features in
are combined together to shape a classifier. this window.

b. Integral image
In an integral image the value at pixel (x, y) is the
sum of pixels above and to the left of (x, y).

Figure3. Haar-like features

Figure5. Input image to Integral image


So haar-like-features are on the whole are
rectangular pattern data. These haar-like features c. Adaboost
have some sort of resemblance to facial features or Adaboost eliminates the redundant or irrelevant
characteristics of face (eyes, nose, cheeks, etc). These features and narrow it down to some 1000 features.
features are applied one by one to whole image and So that we have to apply only useful features to an
where the face features matches with it, it gives high image instead of all the features to reduce the
intensity value. The sum of pixels under white complexity and to make it time efficient. Adaboost
rectangular is subtracted from the sum of pixels find the best features among all the 16000+ features.
under black rectangular gives a single value. Figure 3
shows the Haar- like- features for face detection. d. Cascading
In every 24*24 we need to evaluate 2500 features that
we obtained after performing adaboost. But, instead
of calculating 2500 features in every 24* 24 window,
we use cascades. That is out of 2500 features, we
make a set of 10 features in one classifier, then the
next 20 or 30 features in next classifier and then 200
or 300 in next classifier and so on. Thus cascading
results in a strong classifier as a linear combination of
weak classifiers.
Figure4. Usage of haar-like-features

In figure 4 two haar features are applied to an image


and these features are able to extract the nose and
eyes from the image and this is done by applying the
feature all over the face and get the high value only
at the pixels where this pattern matches exactly, from
Figure6. Cascaded classifiers

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


521
Payal Maken Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 519-525

2. Face detection using skin color 1. Background removal/subtraction or cropping


Human skin color is proven to be most efficient 2. Converting RGB into gray image
method of detecting a human face than any other 3. Converting grayscale into binary image
algorithms that uses facial features[2]. Using skin 4. Filtering or smoothing
color for detection has several advantages, that the
1. Background subtraction or cropping
processing is much faster. The image for which face
In background subtraction, the face is detected from
has to be detected must be in color picture not in
the whole image, and this is done by face detection
grayscale. Although, different people have different
method with the help of features extraction (eyes,
skin colors and the difference lies between their
nose, lips, etc).
intensities rather than chrominance [3]. Since the
input color image is in RGB format, so this technique
usually uses the color components in other color
spaces such as YCbCr, because of the RGB
components which may fail to detect the face.
In YCbCr format, luminance information is
contained in Y component and the chrominance
information is in Cr and Cb components. Therefore Figure 7. Background removal or subtraction
luminance information of the skin can be easily de-
2. Converting RGB into gray image
embedded and this will indicates the skin region.
Now the face detection in skin model involves RGB (red, green and blue) image is known as true
detecting the skin areas. Later the skin region is color image, is stored as m*n*3 data array having red,
localized using: green and blue color components for each pixel. The
color of each pixel is decided by the combination of
• Template matching
red, green and blue intensities. Now, Gray images are
• Neural networks
those in which the only colors are the shades of gray.
B. Image preprocessing In ‘gray’ color the red, green and blue components
have equal intensity. The reason for converting RGB
Normally the images that are obtained from image
into grayscale is that less information needs to be
acquisition are not suitable for face recognition.
provided for each pixel. RGB image can be converted
Quality of image plays a crucial role in face
into Grayscale by applying weighted method or
recognition rate. Quality of image is compromised
luminous method:
due to the certain factors such as noisy images, poor
a. Get the RGB value of pixel.
resolution, etc. So to overcome these problems image
b. Use the formula: Grayscale = 0.3*R + 0.59*G +
preprocessing is done before feature extraction (face
0.11*B
recognition) which will increase the face recognition
c. Replace the R, G and B value of the pixel
rate.
with the Grayscale value calculated at
There are various image preprocessing techniques the
previous step.
can be used to improve the quality of input images
such as image normalization, de-noising, image
filtering, background subtraction etc.
Various steps involved in image pre-processing are
[4]:

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


522
Payal Maken Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 519-525

C. Face recognition
Face recognition primarily means extracting the
features from face i.e. eyes, nose, chin, etc. Various
techniques which are used for this purpose are
following:

1. Principal component analysis (PCA) or Eigen faces


Figure 8. RGB to Grayscale image It is mostly used tool for data analysis making a
3. Converting grayscale into binary image predictive model (face recognition system). Face
recognition technique based on PCA generally make
Binary image is a digital image that has only two
use of an Eigenfaces. In face recognition system,
possible values for each pixel (0 for black and 1 for
training set consists of a set of M images, and each of
white). It uses only two colors black and white.
these images is of m*m dimensions, so there will be
Grayscale image can be converted into binary image
m2 pixels. The PCA transforms the set of images into
by calculating its pixel value, if the intensity value/
low dimensional pictures known as Eigen faces. So
pixel value is greater than 127 then change it to 255
basically, PCA is a mathematical method which use
or 1 and if the intensity value is less than 127 change
an orthogonal transformation to convert a set of
it to 0, then the image is converted into binary image.
values of possibly correlated M variables(face images
in the training dataset) into a set of values of K
uncorrelated variables called principal
components(Eigen faces). The numbers of Eigen faces
is always less than or equal to the number of test
images in the training set i.e. K< =M.

Figure 9. Grayscale to binary image

4. Filtering or smoothing
Filtering or smoothing is done to enhance the quality
of the image by the removal of the noise (blurring,
poor resolution, etc) from the image.
Median filter: Filtering set the value of Figure 10. Face images to Eigen faces
the output pixel as the median of all the In Eigenfaces the first principal component shows the
neighborhood pixels that is taken as input. Other most dominant feature of the dataset and each
filters replace the pixels values by mean and succeeding component in turn shows the next
average value: dominant component.
a. Take the median value of pixels in a window in
the neighborhood.
Median= sort all the pixels in an increasing
order, take the middle one.
b. Replace the pixel value with the median value.

Figure 11. Eigen faces

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


523
Payal Maken Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 519-525

To reduce the calculations that are needed for finding separable and produces the good results even under
the principal components, the dimensionality of the varying illumination conditions and images with
original dataset is reduced before they are calculated. different poses and expressions.
Since each preceding components how less features
and more noise, only few first principal components
are selected and rest of the last are discarded.

Figure14. Fisher Faces

Figure12. Selected k useful Eigenfaces 3. Independent Component Analysis


Each image in the original dataset can be represented ICA is an abstract principle of PCA. PCA uses second
as a linear combination (weighted sum) of Eigenfaces. order statistics while ICA uses high order statistics to
form a better set of basis vector of face images. It
represents faces as a linear combination of
independent sources. It treats the images as random
variables and pixels as outcomes. So, it constructs the
face without face class information.

D. Verification/Validation
Verification and validation is the process of
Figure13. Weighted sum of Eigen faces recognizing an individual.
So, PCA reduces the number of values (from M to K) 1. Distance-wise classification
needed to recognize a face.
Once the features are extracted from face recognition
2. Linear Discriminant Analysis or Fisher Faces techniques, it is then to classify whether the test face
LDA is also known as Fisher faces. LDA is derived belongs to the training set or not. It is done by
from Fisher’s Linear Discriminant (FLD) technique. It measuring the distance between the training dataset
is considered as a classical technique in pattern and the image. If the distance is minimum, the test
recognition. It is used to find a linear combination of image is verified and the person is recognized.
features which separate two or more classes into 2. K-Nearest neighbor
objects [5]. Data is assumed to be uniformly
In this method, nearest neighbor is searched, and if
distributed in each class. It is used to convert the
maximum number of neighbors belongs to one class,
training samples into bases vectors. Linear
we can classify test image to that class. With
combination that is obtained using FLD is known as
increases k value, the classification rate decreases.
Fisher faces while in PCA it is known as Eigen faces.
LDA is preferred over PCA because it groups the 3. Neural Networks
classes into projected faces that can be linearly

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


524
Payal Maken Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 519-525

It is mostly used because of its similarities with the [4]. A. Sengupta and P. Mondal, “Estimating The
human brain. Neural networks make the mapping of Age Of Human Face In Image Processing Using
the images with the person’s expressions. Matlab,” vol. 4, no. 3, pp. 10899–10902, 2015.
[5]. M. N. S. Zainudin et al., “Face Recognition
III. CONCLUSION using Principle Component Analysis ( PCA )
and Linear Discriminant Analysis ( LDA ),” no.
The research in face recognition has been an exciting 05, 2012.
area and it will keep attracting many engineers and
scientists for many years in this field. In this paper Cite this article as :
different techniques that can be used in face
recognition are discussed. It also gives the whole Payal Maken, "An Elementary Study on Various
concept of face recognition system. According to this Techniques Involved in Face Recognition Systems : A
literature survey, Viola Jones gives real time Review", International Journal of Scientific Research
performance for face detection. Skin detection in Computer Science, Engineering and Information
technique also works well. Image preprocessing is Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
necessary to reduce the noise from the images and to 5 Issue 1, pp. 519-525, January-February 2019.
give better recognition rate. PCA and LDA works Available at doi :
better for features extraction. Among PCA and ICA, https://doi.org/10.32628/CSEIT1951116
ICA uses high order statistics. Neural networks, k- Journal URL : http://ijsrcseit.com/CSEIT1951116
nearest neighbor techniques and Distance-wise
classification methods are used for further
recognition and classification. Combining all these
techniques, a face recognition application or system
can be developed.

IV. REFERENCES

[1]. D. Saini and R. Kaur, “Assembling of Human


Beings in an Image to Detect a Group on the
basis of Distance using Digital Image
Processing,” International Journal of Current
Engineering and Technology, vol. 6, no. 2, pp.
462–466, 2016.
[2]. H. M. Z. Hewa Majeed Zangana, “A New
Algorithm for Human Face Detection Using
Skin Color Tone,” IOSR Journal of Computer
Engineering, vol. 11, no. 6, pp. 31–38, 2013.
[3]. A. Amine, S. Ghouzali, and M. Rziza, “Face
Detection in Still Color Images Using Skin
Color Information,” Network, vol. 24, no. July
2014, p. 4, 2002.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


525
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951132

E Return Tool
Pooja Aher*, Manish Kirnapure, Shubhangi Nemade, Tejaswita Shrivastava, Lulkaran Chohale
Department of computer science and Engineering, Priyadarshini J.L. College of Engineering, Nagpur,
Maharashtra, India
ABSTRACT

Every organization has need to pay salary to their employees, this work is done every month. Each and every
organization has to maintain the record monthly for future use. Now a day as we know the countries goes
become digit so this work is also performed digitally. This project E RETURN TOOL involves everything
regarding to employee payment and related to salary tax. This also includes delivering of taxes to the
government and also keep track of employee payments. This tool will facilitate to the employers to pre-validate
the ECR and also carry out amendments in the ECR file before uploading on Unified Portal. The return tool has
member wise details of wages and the contributions including basic details for the new and existing members’
return Tools also maintain the record of EPF member. Employer can check monthly or yearly details of every
EPF member. When a company hires its first employee it started its e return tool. In the United States, every
new employee must be reported to the state along with a completed W-4 tax form. The W-4 determines how
many allowances the employee qualifies for when calculating the federal income tax that should be withheld
from each check. Generally, the employee has to pay income tax depending upon their dependents.
Keywords : E Return Tool, ECR, EPF, EPFO, SME

I. INTRODUCTION existing members. E return Tools also maintain the


record of EPF member. Employer can check monthly
E Return Tool an Offline Tool for employers for and yearly details of every EPF member.
validating the ECR (electronic challan cum return)
before uploading it at the unified portal. Now a day, The E RETURN TOOL process typically includes
EPFO has not produced any software for validating calculating employee pay, recording payroll
the ECR or EPF return. Many employers are facing transactions and determining and paying taxes. A
this problem how to validating ECR for uploading at company must have in place of a timekeeping system
the unified portal because unified portal has only that accurately reflects the hours put in by
accept validating ECR "In order to assist the employer nonexcempted employees as well as the regular salary
to create the correct ECR file every month, E Return payments for exempt workers. Typically, employee
Tool provided an offline tool for employers for has to pay income from their earning; at the end of
validating ECR before uploading at Unified Portal," the year, they must report all wages, tips and
this tool will facilitate the employers to pre-validate othercompensation paid. Companies also must
the ECR and also carry out amendments in the ECR withhold the Social Security and a Medicare
file before uploading on Unified Portal. The return contribution from employees' wages and have to pay
has member wise details of the wages and a matching amount.
contributions including basic details for the new and

CSEIT1951132 | Received : 15 Feb 2019 | Accepted : 28 Feb 2019 | January-February -2019 [ 5 (1) : 526-530 ]
526
Pooja Aher et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 526-530

II. METHODS AND MATERIAL Form 5 are mandatory. In case a member has exited,
you can enter the date of exit and in such case the
E Return Tool software does a lot of the heavy lifting reason for exit will become mandatory. There are
of salary management for you. Like outsourced some additional details which also can be filled up.
payroll services, this software requires very little These details may become helpful later when
input from the employer, just employee wage member applies for claim. However, they are not
information and days. All of the tricky calculations mandatory.
and withholdings are calculated by the software,
which receives automatic updates every time tax laws D. Salary Information
are changed. The software can also fill out
This module taking the input from employer like
employment tax forms for you and remind you when
number of days and rate. Basis of that input it
to file them.
calculate the employee deduction and stored their
record in database
Module
E. Summery:
A. Registration:
This screen after the entry of the Return Month takes
Registration is first module that is authentication to
you to the actual data entry/upload screen for the
user means in Registration module software create
monthly return. This screen also displays the
the username and password
summary, account wise and month wise of the
already entered/uploaded wage details. In this
B. Log-in:
complete summery will be recorded in the form of
This is the screen where you have to enter your user A/c 1, A/c 2, A/c 10, A/c 21, A/c 22, Month, Year and
id and password for login. You have come here after total. The screen will ask you to enter in the Return
selecting an added establishment from the earlier Month and Return Year fields.
screen. The screen at top shows the Establishment ID
you have selected. In case you have selected a wrong F. Monthly Return:
establishment from the list, if multiple establishments
In this module EPF details is recorded and the based
have been added, please click BACK instead of login.
on this details EPF return fill in Unified Portal. This
It will take you back to the first screen for selecting
screen is to be used for entry of wages for each
the correct establishment.
member for a month.

The system allows the admin user to create/delete


1. Data can be entered for each member one by one.
new/existing users and to change password. However,
2. Data can be imported through csv file
the admin user cannot be deleted and only his
password can be changed. G. ECR and Report:
a. ECR:
C. Employee Registration
In this screen, thevalidated text file will be generated
This screen is used for upload/entry of the details of on the basis of the wages and other information
the members as required for filing the Return in entered by the user through the various functions of
Form 5, Form 10, for International Workers and for the software andbefore uploading the file and
Disabled Workers. The data that are required for the remitting the amounts in the bank against the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


527
Pooja Aher et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 526-530

uploaded return, the employers should satisfy Flowchart :


themselves with the correctness of the data in the
text file.

b. Report:
i. Form 5:
In this module Form 5 is generated between pf range
or between Joining date. Form 5 is nothing but the
complete basic information of employee.

ii. Form 3A:


Form 3A is monthly details of individual employee
break in service.
iii. Form 6A:
Form 6A is yearly details of individual employees
including break in service.

III. PROBLEM FORMULATION


H. Parameter:
The use of traditional spreadsheet applications such
In this module user can easily change the parameter.
as Excel in the past few years has proved to be an
This is most important for user because some time
inefficient way of keeping faculties’ records,
employer has paid previous challan like Arrears.
calculating faculties’ enumerations and deducting
obligatory contributions from salaries. No option to
I. Back up and Reset:
restore or get the backup of all databases of faculties
This screen is to be used by admin only to reset data by just clicking a button Small and mid-sized
for the particular month. User can easily take backup enterprises (SMEs) are most affected as they
of complete file and store them to another location sometimes tend to use low end applications even as
for future used. they expand in business activities. This manual
method of calculating faculties’ salaries and carrying
out deductions impossible task due to its incapacity to
handle large amount of data without error. No bar
graphs to show the financial profit or loss of colleges.

IV. PROPOSED SYSTEM

This project is an online tool which is used to create


the validating ECR to upload in Unified Portal for
creating the EPF challan, this project is used for
involves everything that has to do with the payment
of employees and the filing of employment taxes.
calculating wages, taxes and other deductions,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


528
Pooja Aher et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 526-530

printing and delivering checks and paying statutory


employment taxes to the government. compliances
This tool will facilitate the employers to pre-validate like PT,
PF,ESIC,
the ECR and also carry out the ECR file before
TDS & IT
uploading on Unified Portal. The return has member
etc.
wise details of the wages and contributions including
basic details for the new and existing members. 3 ECR In this tool In this There is [4
return Tools also maintain the record of EPF TEMP used to tool no record ]
member. Employer can check monthly or yearly LATE create an used to will store
details of every EPF member. 2.0 ECR file in create
text format an If the
for filling Arrears employee
V. LITERATURE SURVEY
the EPF file in is cross 58
return at text year then
S Appli Features Proble Algorith R
unified format it’s not
r cation m m ef portal for created
Stateme filling accurate
N nt
the EPF text file
o
return
1 Mont In this 3A not bubble [1 at
hly E system generat sort and ] unified
return employer e binary portal
Softw can store the search
are EPF details KYC algorithm
of employee file not is used in
and generate create existing
ECR in text system VI. CONCLUSION
format to fill
EPF return Although a conclusion may review the main points of
on EPFO the paper, do not replicate the abstract as the
website conclusion. A conclusion might elaborate on the
importance of the work or suggest applications and
2 Comp A- 1) No Quick [2
extensions. Authors are strongly encouraged not to
uter PAYshould option sort and ]
Based take care of to binary call out multiple figures or tables in the conclusion
Payro Personnel restore search these should be referenced in the body of the paper.
ll Management or get used in
Syste , Payroll, the existing VII. REFERENCES
m Attendance backup system
Management of all [1]. "Automated Payroll System (A-PAY)", by
using databas
Arjun V. Singh , Siddesh V. Chaphekar ,Yogesh
biometric es of
S. Sawant, Volume 03, Issue 02, [February –
machine, faculties
Leave by just 2016] , ISSN (Online):2349–9745; ISSN
Management clicking (Print):2393-8161.
, SMS a button [2]. "Computer Based Payroll System
intimation, Implementation For E-Governance at Punjab

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


529
Pooja Aher et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 526-530

Agricultural University", by Poonamdeep Kaur,


Dr. Dinesh Grover, e-ISSN: 2278-067X, p-ISSN
: 2278-800X, Volume 5, Issue 3 (December
2012), PP. 55-60.
[3]. Taking EPF Parameter from
https://www.epfindia.gov.in/site_docs/PDFs/Mi
scPDFs/ContributionRate.pdf
[4]. ECR TEMPLATE 2.0 and ARREARS
TEMPLATE 2.0 available on
https://drive.google.com/file/d/0BwN9ZAwr4e
nzVVNUbU43S1ljcVk/view

Cite this article as :

Pooja Aher, Manish Kirnapure, Shubhangi Nemade,


Tejaswita Shrivastava, Lulkaran Chohale, "E Return
Tool", International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 526-530, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT1951132
Journal URL : http://ijsrcseit.com/CSEIT1951132

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


530
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951128

IoT Based LPG Gas Leakage Detector


Prof. Pranay Meshram*1, Nancy Shukla2, Stuti Mendhekar3, Renuka Gadge4 , Shivani Kanaskar5
*12345Computer Science &Engineering, Priyadarshini J. L. College of Engineering, Nagpur, Maharashtra, India

ABSTRACT

Gas leakage is a major problem in the industrial sector, in residential locations etc. One of the preventive
methods to stop the incident related to the gas leak is to install a gas leak detection kit at vulnerable locations.
The goal of this paper is to propose a system that can detect, alert and automatically control gas leaks. In
particular, a gas sensor has been used which has a high sensitivity to gases such as propane and butane together
with LPG. There is an alarm that is triggered once the LPG has been detected. The gas leakage system consists
of a Wi-Fi module that alerts the user by sending an SMS message.
Keywords : LPG gas, IoT (Internet of Things)

I. INTRODUCTION introduced so that it can also be used with alternating


current. To support this last case, it has a bridge
In homes and in various commercial areas, LPG is rectifier with a capacitor filter. This is followed by a
used for various purposes, such as cooking, Heating, controller that provides regulated power of + 5V.
Lighting, Cooling, etc. This energy source consists MQ-5 gas is used to detect LPG. This sensor can be
mainly of propane and butane, highly flammable used at + 5V. the sensitivity of this sensor is very high
chemical compounds. LPG loss can occur, albeit and its response time is fast.
rarely, in a home, business or gas vehicle. The loss of
this gas can be dangerous because it increases the risk There are about 30 crore users of LPG in the country
of explosion. An odor such as ethanol is added to the where the majority of 40% of the population. Several
LPG so that the losses can be easily detected by most standards have been put in place for the gas leak
people. However, some people with a reduced sense detection system. The existing system provides an
of smell may not rely on this intrinsic safety alarm system primarily to detect a gas leak in
mechanism. In these cases, a gas leak detector domestic and commercial premises. The main
becomes vital and helps protect people from the objective of this system is to monitor the leakage of
dangers of gas leaks. A number of research papers liquefied petroleum gas (LPG) in order to avoid
have been published on gas leak detection serious fire accidents and facilitate safety measures.
techniques. We have studied a research article about The system detects LPG leaks from the gas sensor and
our topic "LPG GAS WARNING DETECTION AND informs the consumer of the gas leak by sending a
LOSS SYSTEM". This document presents the text message.
detection and alarm system for LPG gases to prevent
fire accidents and ensure the safety of the home. The When the system determines that the LPG
LPG leak detection and reporting system has been concentration in the air reaches the specified level,

CSEIT1951128 | Received : 15 Feb 2019 | Accepted : 28 Feb 2019 | January-February -2019 [ 5 (1) : 531-534 ]
531
Prof. Pranay Meshram et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 531-534

tell the consumer by sending text messages to the Arduino is an open-source electronics platform based
registered mobile phone and informing people in on easy-to-use hardware and software. It
your home by activating the alarm including the has 14 digital input/output pins (of which 6 can be
alarm and the message display the message on the used as PWM outputs), 6 analog inputs, a 16 MHz
LCD screen to take required action. quartz crystal, a USB connection, a power jack, an
ICSP header and a reset button. Arduino boards are
WHAT IS IoT? able to read inputs - light on a sensor and turn it into
an output - turning on an LED, activating buzzer and
From homes and industries to enterprises, Internet
WiFi module for sending SMS. You can tell your
of Things have become commonplace in all walks of
board what to do by sending a set of instructions to
life. Connecting everything and everyone, IoT is
the microcontroller on the board. To do so you use
making the world smarter and better than ever
the Arduino programming language (based
before. However, this is just the beginning.
on Wiring), and the Arduino Software (IDE), based
Unlocking the full potential of Internet of Things
on Processing.
requires that businesses understand the
opportunities for value creation and systematically
Gas Sensor (Mq5) :
address the underlying challenges.
Generally, semiconductor sensors are used to detect
Internet of Things (IoT) is an ecosystem of connected LPG gas. MQ5 semiconductor sensor is used in this
physical objects that are accessible through the system. Sensitive material of MQ-5 gas sensor is SnO2,
internet. The ‘thing’ in IoT could be a person with a which with lower conductivity in clean air. When
heart monitor or an automobile with built-in-sensors, the target combustible gas exist. The sensor
i.e. objects that have been assigned an IP address and conductivity increases along with the rising gas
have the ability to collect and transfer data over a concentration. MQ-5 gas sensor has high sensitivity
network without manual assistance or intervention. to Propane, Butane and LPG, also response to Natural
The embedded technology in the objects helps them gas. The sensor could be used to detect different
to interact with internal states or the external combustible gas, especially Methane; it is with low
environment, which in turn affects the decisions cost and suitable for different application. The MQ-5
taken. Devices and objects with built in sensors are can detect gas concentrations anywhere from 200 to
connected to an Internet of Things platform, which 10000 ppm. The sensor's output is an analog
integrates data from the different devices and applies resistance.
analytics to share the most valuable information with
applications built to address specific needs. WI-FI Module :
The WI-FI technology is radio wireless networking
These powerful IoT platforms can pinpoint exactly technology. It is mainly used for communication
what information is useful and what can safely be purpose. In this project, the WI-FI system is used to
ignored. This information can be used to detect communicate with the user. When the gas leakage
patterns, make recommendations, and detect possible occurs, the microcontroller stops the leakage and
problems before they occur. alerts the surroundings. Then, the information about
the leakage has to be informed to the user. For this
II. METHODS AND MATERIAL purpose, WIFI is used. Using WI-FI, a warning is sent
to the user. The method of communication is
Arduino Kit : asynchronous serial communication. The

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


532
Prof. Pranay Meshram et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 531-534

corresponding code has to be loaded into the is connected to the Arduino kit. As soon as the gas
microcontroller, to which WI-FI is connected. level has exceeded this threshold value, the LED
lights up and the system warns the environment by
Buzzer: sending out a sound that is issued by the buzzer. At
It most commonly consists of a number of switches or the same time, the sound is sent via a WiFi module to
sensors connected to a control unit that determines if the user's mobile phone in order to inform the user of
and which button was pushed or a preset time has the loss. The user can avoid dangerous events such as
lapsed, and usually illuminates a light on the fire with the help of an exhaust fan.
appropriate button or control panel, and sounds a
warning in the form of a continuous or intermittent SYSTEM SENSO INDICATI FEATURE
buzzing or beeping sound. R ON
ARM Based MQ-6 Inform GSM
III. LITERATURE SURVEY Gas through Module is
Monitoring SMS used to
To overcome the issue of LPG leakage, many system System notify the
has been propossed. Gas leak detection is the process user
of identifying potentially hazardous gas leaks by Hazardous MQ- LCD Detects the
means of various sensors [2]. Several designs of LPG gas detection 4,MQ-7 Displays concentrati
detection and alert system have been proposed in the using arduino on
literature. Apeh et al. designed kitchen gas leakage Gas leakage MQ-2 Wireless Sends the
detection and automatic gas shut off system [3]. detector built Module informatio
T.Soundaryaet al. presented the cylinder LPG gas on n through
leakage detection system [4]. Wireless and GSM microcontroll wireless
technology [5] based gas detectors have also been ers module
proposed. This paper presents a LPG leakage IOT Based Electron Through Customer
detection and alert system to avoid fire accidents and real time ic sensor Social can check
to provide house safety. monitoring media the
gas detection previous
IV. PROPOSED SYSTEM system data and
determine
The proposed system uses the Arduino kit with an at which
integrated microcontroller. This Arduino kit forms condition
the core of our proposed system. Arduino is emergency
connected to other components that complement the occurred
system. These components are the MQ5 sensor, the
Gas leakage MQ-5 Through Notify the
LED light, the buzzer and the WiFi module. The
detection and buzzers and user
system uses an MQ5 gas sensor to detect LP gas. A
monitoring alarms through
threshold value (180 ppm for the proposed system) is
using IoT alarms
set in the MQ5 sensor. The sensor continuously
detects the surrounding atmosphere. For this
continuous air evaluation, the 5V power supply is
supplied to the Arduino via an AC adapter. A buzzer

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


533
Prof. Pranay Meshram et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 531-534

in April 2013.
[5] Online Monitoring of Green House Gas
Leakage inIndustries Angel College of
Engineering and Technology, Tirupur, Tamil
Nadu, India. In DecJan, 2014.
[6] Development of Movable Gas Tanker Leakage
Detection Using Embedded System
Mr.S.B.Patil, Dr.A.J.Patil Dec 2012.

Fig 1. System Block Diagram Cite this article as :

V. CONCLUSION Prof. Pranay Meshram, Nancy Shukla, Stuti


Mendhekar, Renuka Gadge, Shivani Kanaskar, "IoT
In this paper, LPG gas leakage is sensed by the MQ5 Based LPG Gas Leakage Detector", International
sensor. As soon as it detects LPG content in air more Journal of Scientific Research in Computer Science,
the threshold value, LED light glows and sound is Engineering and Information Technology
produced. Also, a SMS is send to user’s cell phone (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
and thus notifies the user about the incident if the 531-534, January-February 2019. Available at doi :
user is out of the range of sound. The system can be https://doi.org/10.32628/CSEIT1951128
used to avoid fire accident and provide house safety. Journal URL : http://ijsrcseit.com/CSEIT1951128

VI. REFERENCES

[1] Shruti Bhoir1, Snehanjali Goregaonkar2, Prof.


ssShilpa Satre“ IoT Based Gas Detection
System” International Journal of Engineering
Science and Computing, April 2017. Volume 7
Issue No.4
[2] E. Jebamalar Leavline1, D. Asir Antony Gnana
Singh2 , B. Abinaya3 H. Deepika4 “LPG Gas
Leakage Detection and Alert System”
International Journal of Electronics
Engineering Research. ISSN 0975-6450 Volume
9, Number 7 (2017) pp. 1095-1097
[3] S.RAJITHA1 ,T.SWAPNA2 Electronics and
Communication Engineering, AP, India in
September-October 2012.
[4] Automatic LPG Booking, Leakage Detection
And Real Time Gas Measurement Monitoring
System B. D. Jolhe, P. A. Potdukhe, N. S. Gawai

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


534
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
Machine learning algorithm for Cyber Security - A Review
Mohammad Asif*, Pratap M. Mohite, Prof. P. D. Satya
Computer Science and Engineering, S.Y.C.E.T. Aurangabad, BATU, Lonere, Maharashtra, India

ABSTRACT

The computer networks are exposed to increasingly safety threats. With new kinds of attacks appearing usually,
growing flexible and adaptive protection-oriented strategies is a severe undertaking. In this context, anomaly-
primarily based community intrusion detection techniques are a precious era to guard target structures and
networks in opposition to malicious sports. Threats the internets are posing higher threat on IDS safety of
statistics. The primary concept is to utilize auditing programs to extract an in-depth set of capabilities that
describe each network connection or host session and practice statistics mining applications to learn rules that
correctly capture the behavior of intrusions and normal activities. Now Intrusion Detection has end up the
priority and on the crucial assignment of statistics protection administrators. A device deployed in a network is
at risk of numerous assaults and desires to be blanketed towards assaults. Intrusion detection machine is a
necessity of these days’ information safety area. It performs a vital function in detection of anomalous site
visitors in a community and indicators the network administrators to manage such visitors. The painting
supplied in this thesis is an attempt to locate such visitor’s anomalies in the networks through generating and
reading the site visitors float information.
Keywords : IDS (Intrusion Detection System), HIDS (Host Based Intrusion Detection System), ML (Machine
Learning), NIDS(Network Based Intrusion Detection System)

I. INTRODUCTION accelerated connectivity of laptop systems gives


greater get admission to outsiders and makes it less
This IDS supplied in this thesis implements the k- complicated for intruders to avoid detection. IDS are
means method of data mining for intrusion detection based on the perception that an intruder’s conduct
and the outlier detection technique the usage of can be exceptionally extraordinary from that of a
community outlier factor to locate the anomalies legitimate consumer. In view that many modern IDSs
present within the site visitors glide. The are built by using manual encoding of expert
performance of these two processes is compared expertise, modifications to IDSs are high priced and
through numerous confusion matrix and overall sluggish. in this paper, we describe a records mining
performance metrics like fake superb charge, framework for adaptively building Intrusion
sensitivity, specificity, type charge and precision, and Detection (ID) fashions. This record attempts a
an evaluation is accomplished to find out that which comprehensive compilation and categorization of
one of the tactics is higher for use for intrusion available intrusion detection gadget (IDS) products.
detection using site visitor’s waft. Intrusion detection lengthy studied and prototyped in academic and
is the hassle of identifying unauthorized use, misuse, authority’s circles, IDS have most effective within the
and abuse of computer systems by both machine last few years all started to emerge as a feasible and
insiders and outside penetrators. Namely, the useful commercial alternative. The primary industrial

CSEIT1951141 | Received : 20 Feb 2019 | Accepted : 28 Feb 2019 | January-February -2019 [ 5 (1) : 535-545 ]
535
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

IDS product becomes launched in 1991, with a ✓ It must impose a minimal overhead at the
relative handful emerging within the subsequent half system wherein it is walking; with a view to no
of dozen years. Then, inside the remaining couple of longer intervene with its ordinary operation.
years, the sphere underwent explosive increase. Even ✓ It needs to be capable of being configured in
after the obvious failure of numerous early releases, step with the security policies of the gadget this
there nevertheless remains at the least seventeen is being monitored.
extant products that claim to provide powerful ✓ It ought to be able to adapt to changes in device
intrusion detection in a networked environment. and person conduct over the years (e.g., new
Given this recent growth, and the reported improved applications being mounted, customers
utility of corporate assets to these products [1, 2, and converting from one hobby to another or new
3], the time has come for a complete assessment of assets being to be had those purpose
the subject. modifications in gadget aid utilization patterns).
✓ Anomalies are deviations from everyday
Intrusion detection structures try to discover laptop consumer behavior. Misuses, alternatively, are
misuse. Misuse is the overall performance of an recognized patterns of attack [5]. at the same
motion that isn't always favored by the machine time as misuse styles are frequently easier to
proprietor; one that doesn't comply with the device’s process and locate, it is frequently the paradox
ideal use and/or security coverage. IDSs mechanically patterns with a purpose to help to discover
examine online person hobby for forbidden (i.e., problems. As misuses are recognized styles of
invalid) and anomalous (i.e., strange, inconsistent) assault, the detection machine tends to fail
conduct. they're primarily based on the speculation whilst novel assault techniques are
that monitoring and reading community implemented. Detection of anomaly patterns is
transmissions, machine audit information, software computationally pricey due to the overhead of
audit data, device configuration, facts files, and keeping track of, and possibly updating several
different information can locate misuse. This record machine profile metrics, because it need to be
encompasses widespread portions of information, tailor-made gadget to device, and every now
effective analysis calls for detection1 specialized and and then even used to a person, due to the
constantly honed expertise, and at the least close to reality conduct styles and gadget usage vary
real-time of misuse is frequently important. significantly.
In [4], the subsequent characteristics are identified as ✓ Different IDSs had been designed to do
suited for IDS: dispensed collection and analysis of records. A
✓ It needs to run constantly with minimal human hierarchical device is defined in [6], and [7]
supervision. describes a cooperative gadget without a central
✓ It has to be fault tolerant in the sense that it has authority. these structures clear up maximum
to be capable of recover from gadget crashes, of the problems mentioned except for the
both unintended or because of malicious reconfiguration or adding skills to the IDS,
activity. Upon startup, the IDS have to be which aren't described in either of the 2
capable of recovering its previous country and designs.
resume its operation unaffected.
✓ The boom of the net has added excellent
✓ It ought to face up to subversion. The IDS need
benefits to society at the identical time the
to be able to screen itself and stumble on if it's
growing attacks on the IT Infrastructure are
been modified by way of an attacker.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


536
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

getting an increasing number of serious issues The growth of Internet has brought great benefits to
and wishes to be addressed. Together with the the society at the same time the growing attacks on
growth of the internet, attacks also are growing the IT Infrastructure are becoming an increasingly
in parallel. serious issue and needs to be addressed. Along with
✓ Host-based intrusion detection system pursuits the growth of Internet attacks are also growing in
at detecting the intrusions at the host stage. parallel. In earlier days, the attacker should have a
They function in my opinion at each host of the good knowledge about the target infrastructure and
community. Consequently, they are able to knowledge on the Network, Operating Systems &
operate on distinctive type and quantity of Applications. Whereas today there are lots of open
records on the equal time. The host-based tools available in the Internet which can trigger
machine is dependent on the host operating automated attacks.
system. Any vulnerability inside the host-
primarily based OS can weaken the integrity of
the host-primarily based.

An outsider can exploit this vulnerability to release


an assault on the way to be hard to be detected
through the HIDS. as a result, a robust HIDS have
to be supported with the non-susceptible host OS.
The bodily deployment of HIDS in a community
may be visible the traffic enters from the internet
via the router or switch or firewall to the local
network. It is the responsibility of the HIDS in the
host to detect possible intrusions in the traffic flow.
Figure 2. Growth of Internet in terms of Host Count

WannaCry attack: In might also 2017, the


WannaCry Ransomware spread via the net, the use
of a make the most vector named EternalBlue. The
ransomware attack infected greater than 230,000
computer systems in over a hundred and fifty
countries the usage of 20 specific languages to
demand cash from users the usage of Bitcoin crypto
currency. WannaCry demanded US$300 in line
with pc. [9]

Petya attack: Petya computer virus spread in the


Figure 1 : Host based intrusion detection system course of April 2016, this malware infected the
grasp boot record of the laptop through encrypting
II. RELATED WORK the report tables of the NTFS file system. as soon as
infected on the next boot expects a ransom is paid.
again in the month of June 2017, a modified model

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


537
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

of Petya the usage of Eternal Blue exploit and this consideration as the collection deviation from
was aimed to create disruption alternatively to normal profile sequence. thus, this device works
generate income. [9] offline the use of previously accrued information
and implements the view desk set of rules for
We located that there are many benefits of gaining knowledge of software profiles significantly.
C4.5algorithms for special attacks occurs on your Duan et al. [12] have concentrated on identifying
dataset and C4.5 will detect the R2L and U2R compromised machines which can be recruited to
attacks and the neural network is the use of for hit upon junk mail zombies. An method SPOT is
detecting the DOS and Probe attacks and many proposed to experiment sequentially outgoing
greater. In this, we are the usage of four algorithms. messages by means of enforcing SPRT (Sequential
The primary one is okay-manner clustering and the possibility Ratio check). This method quick
second steps are fuzzy good judgment 1/3 steps are estimates whether a number is compromised or not.
SVM and the remaining and very last step is C4.5. figuring out compromised machines using malware
The blessings of all this set of rules are to detect the contamination device is said via Bot hunter [13].
assaults from the datasets. After that, it will
examine it with the SVM and C4.5 classifiers to This machine has massive no of steps that allow
find out how many attacks are coming about in the intrusion detection alarms correlation caused using
dataset at the same time as transferring the dataset inbound visitors with outgoing message trade
from supply to destination. [8] sample consequences. Bot Sniffer [14] explained in
his paintings approximately compromised device
This center also evolved a subsequent-era traits which might be a uniform temporal-spatial
mechanism which includes audit profiles of conduct for detecting zombies. This approach
person’s and may monitor the modern-day identifies zombies by using combining flows based
reputation of the consumer, if any change takes totally on server connections and searching flows
place with person’s hobby as compared with audit with comparable conduct respectively. Kumar and
profile of person then it will generate an alarm. Goyal [15] have explained implements genetic
algorithms in dataset training to categories the
Haystack [10] later developed a framework to labels that are smurf attacked and achieves a low
estimate an intrusion detection approach based false high-quality ratio of 0.2%. in addition,
totally on person and anomaly strategies. Six paintings were completed through Abdullah [16]
varieties of intrusion have been detected and and co-people elaborated intrusion detection class
people include the masquerade assaults, malicious regulations using genetic algorithms. Intrusion
use, leakage, carrier denial, the unauthorized detection policies the usage of genetic algorithms
consumer’s wreck-ins try, and get admission to becomes also the have a look at made through
control of a protection system. The source fire Ojugo et al. [17]. This approach uses health
developed suggests a community-based intrusion characteristic for estimating the rules. system
detection and prevention mechanism referred to as learning strategies are also applied to hit upon the
snicker machine that is an open supply. Forrest [11] intrusion. current device mastering strategies
in 1996 created an ordinary profile based on (artificial Neural Networks - ANN) for intrusion
studying the call sequences between intrusion detection was described by way of Roshani group
detection and protection in opposition to a human [18]. Gaikwad et al [19] added a technique based on
machine. An assault on this device is taken into fuzzy clustering and ANN approach.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


538
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

which the statistics is destined can offer the


PROBLEM STATEMENT :- A few of the existing attacker the possibility of acting Insertion and
network- and host-based totally IDSs [20, 21] Evasion assaults. those assaults employ
perform records collection and analysis centrally mismatched assumptions inside the community
using a monolithic architecture. by using this, we protocol stacks of different hosts to hide the
mean that the facts is accrued through a single assaults or create a denial of- provider attacks.
host, both from audit trails or via tracking packets different IDSs had been designed to do
in a network, and analyzed by means of a dispensed collection and analysis of facts. A
unmarried module the usage of one of a kind hierarchical device is defined in [25], and [26]
techniques. other IDSs [22, 23] perform dispensed describes a cooperative gadget without a
facts collection (and some preprocessing) with the government. those systems resolve most of the
aid of the use of modules disbursed inside the hosts problems cited besides for the reconfiguration
which are being monitored, however the or including abilities to the IDS, which aren't
accumulated statistics remains shipped to a crucial described in either of the 2 designs.
vicinity in which it is analyzed through a ✓ Make a larger framework to resource
monolithic engine. an awesome overview of unsupervised ML: The feature choice" step
systems that take both methods is presented in within the framework currently requires
[24]. There are a number of issues with those labeled datasets to discover the most
architectures: discriminative talents. Destiny art work will
✓ The imperative analyzer is a single factor of dispose of this framework requirement to
failure. If an outsider can by some means save manual unsupervised ML with unlabeled
you it from operating (for instance, by crashing records. big effort in this thesis went into
or slowing down the host wherein it runs), the ensuring experiments have been representative
complete community is without safety. of actual-worldwide situations. This ensured
✓ Scalability is constrained. Processing all of the our outcomes have been considerable in
statistics at a unmarried host implies a limit on modern-day networks our datasets have been
the dimensions of the network that can be snapshots of community site visitors which
monitored. After that limit the vital analyzer may be analyzed opine in batch mode. For the
turns into not able to preserve up with the detectors to artwork on a live network.
town of statistics. Disbursed data collection can
also motive problems with excessive facts
visitors within the network.
✓ It’s far hard to reconfigure or add abilities to
the IDS. Changes and additions are generally
carried out by way of editing a configuration
report, including an entry to a desk or putting
in a brand new module. The IDS commonly
must be restarted to make the modifications
take impact.
✓ Analysis of network facts can be awed. As
shown in [20], appearing a set of network
information in a number other than the only to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


539
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

supervised scheme implementing basis of K-means


and KDD CUP algorithms.

Logs documents need to be tested to understand any


compromised bills, originating IP’s, and all sources
accessed via using the attacker. All related sports want
to be collected and examined several weeks or even
months in advance than the detected event. capability
areas of future paintings are computerized correlation
and assessment of the log facts from cyber-attacks.
additional machine mastering algorithm.

PROPOSED SYSTEM: -According to base paper the


machine learning deal with two concept i.e.
supervised learning and unsupervised learning. Both
concept have some limitations, to overcome this
limitation we can concern with the concept of semi
supervised approach. In semi-supervised approach
concern two scenarios. In this approach front-end to
be implementing using of K-means Algorithm and
back-end to implementing using of KDDCUP 99. In
k-means algorithm concern to secure to be user
interface and KDD CUP 99 concern as to centralized
data storages on the system or host.
There are numerous procedures and methods utilized
in id. each method has deserved and demerits.
therefore, this paper highlights the similar
Figure 3 : Classification of ml algorithms for cyber distribution of attacks nature with the aid of using ok-
security applications [28] way and also the powerful accuracy of the Random
forest set of rules in detecting intrusions. This paper
Machine learning algorithms fragmented as describes complete sample popularity and machine
supervised learning and unsupervised learning. learning set of rules overall performance for the four
Behave of the scheme as to be defined as, supervised attack categories, which include Denial-of-service
learning which to be deal with more sophistication (DoS) attacks (deny legitimate request to a system),
rather than unsupervised learning but it has too costly Probing attacks (statistics gathering attacks),
and complexity. consumer-to-root (U2R) assaults (unauthorized access
to nearby splendid-consumer), and faraway-to-
But in case of unsupervised learning it takes a huge
neighborhood (R2L) attacks (unauthorized
amount of input but produce low amount of desirable
neighborhood get admission to from a faraway device)
output. The supervised scheme work under the expert
proven inside the KDD Cup 99 intrusion detection
system.
dataset.
To overcome the limitation of this, we can go with
semi supervised learning scheme. This semi

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


540
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

channeled effectively, vastly, it feasible to deal with


size of network facts and heterogeneity [29].

In this project, we use k-means algorithm to cluster


dataset connections. The k-means set of rules is one
of the widely recognized clustering tools.
K-means agencies the information according to
with their characteristic values right into a person-
particular variety of ok wonderful clusters. Facts
categorized into the same cluster have identical
characteristic values. k, the fantastic integer
denoting the number of clusters, desires to be
Figure 4 : - Overview of IDS furnished earlier. the steps involved in a k-means
algorithm are given consequently: [30]
III. METHODOLOGY
This project consists of the communication of the 1 1. K points denoting the data to be clustered are

algorithm of data mining class methods. Those is k- placed into the space. These points denote the
means primary group centroids.
2. The data are assigned to the group that is
K-means Clustering Algorithm: - adjacent to
the centroid.
Clustering, primarily based on distance 3. The positions of all the K centroids are
measurements carried out on items, and classifying recalculated
gadgets (invasions) into clusters. not like type, as soon as all the data are assigned.
classification because there may be no information 4. Repeat steps 2 and 3 until the centroid
about the label of studying statistics is an unattended unchanged.
getting to know system. For anomalous detection, we
will use welding and in-intensity evaluation to guide This consequence within the partition of facts into
the identity model. Dimension of distance or groups. The preprocessed dataset partition is
similarity performs a critical function in collecting achieved the usage of the ok-means set of rules
observations into homogeneous corporations. with k value as 5. due to the fact we have the
dataset that contains regular and 4 assault
Jacquard affinity size, the longest not unusual order categories which include DoS, Probe, U2R, R2L.
scale (LCS), is vital that the occasion is to evoke the
scale to decide if regular or odd. Euclidean distance is
about two vectors X and Y in space Euclidean n-
dimensions, the size of the distance widely used for
vector area. Euclidean distance can be defined
because the rectangular root of the whole difference
of the same vector dimension. Subsequently,
grouping and category algorithms want to be

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


541
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

KDDCup’99 dataset has variations of education


dataset; one is a complete education set having 5
million connections and the opposite is 10% of
this schooling set having 494021 connections.
since the whole dataset is huge, the test has been
completed on its smaller amount of dataset this is
10% of KDD. moreover, the KDDCup’99 dataset
consists of many assault behaviors, categorized
into 4 companies: Probe, Denial of provider (DoS),
user to Root (U2R), and far-flung to local (R2L)
[33]. these may be visible in desk I. normal
connections are created to profile than predicted
in a military network. The detailed information of
the two versions of schooling dataset may be seen
in table 1.

The KDDCUP 99 was simulated in a army


community environment and used for The 0.33
global understanding Discovery and information
Mining equipment opposition, which became held
alongside KDDCUP 99 The 5th international
convention on knowledge Discovery and facts
Mining. The competition challenge changed into
KDD Cup 99 Dataset: - The assessment of any to study a predictive model or a classifier able to
intrusion detection algorithm on actual network distinguishing among valid and illegitimate
data is extraordinarily tough particularly because connections in a computer community. This
of the high fee of acquiring proper labeling of dataset includes one form of normal records and
community connections. due to the actual pattern 24 distinctive sorts of assaults which might be
Table be gotten for intrusion detection, the categorized into four kinds such:
KDDCup’99 datasets are used as the sample to
confirm the overall performance of the misuse
detection model. The KDDCup’99 datasets,
referred by way of Columbia college, became
arranged from intrusions simulated in military
community surroundings on the DARPA in 1998.
It includes network connections obtained from a IV. CONCLUSION AND FUTURE SCOPE
sniffer that facts all network visitors the use of the In this project we can sophisticate more on K-
TCP unload layout. The overall simulated length is means rather than KDDCUP 99, because it
seven weeks. It was carried out in the MIT Lincoln valuable to aspect of semi supervisor concept. This
Labs after which announced at the UCI KDD Cup scheme facility to labeled and unlabeled dataset as
1999 Archive [32]. efficiency.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


542
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

This paper provides a comparative evaluation [6] S. Staniford-Chen, S. Cheung, R. Crawford,M.


hybrid system getting to know approach to detect Dilger, J. Frank, J. Hoagland, K. Levitt,C. Wee,
the Denial of service (DoS) assaults, Probing R. Yip, and D. Zerkle. GrIDS: A graphbased
intrusion detection system for large
(Probe) assaults, consumer-to-Root (U2R) attacks,
networks.In Proceedings of the 19th National
and far off-to-neighborhood (R2L) attacks. we can
InformationSystems Security Conference,
understand the same nature of assault organization volume 1, pages 361-370. National Institute of
by the use of ok-manner algorithm. And then we Standardsand Technology, October 1996.
use to categories ordinary and attack connections. [7] Gregory B. White, Eric A. Fisch, and Udo
The experiments display that KDD Cup 99 dataset W.Pooch. Cooperating security managers: A
can be carried out as an effective benchmark peer basedintrusion detection system. IEEE
dataset to help researchers evaluate distinctive Network, pages 20-23, January/February 1996.
[8] Intrusion Detection System by using K-Means
intrusion detection fashions. future paintings Clustering, C 4.5, FNN, SVM Classifier Akshay
Takke1, Ravikumar Gujjul2, Mikhil Ghag3 ,
consist of analyzing with other records mining Vivek Pawar4, Vivek Pandey5 Page no:-636
algorithms to categories attack categories and how [9] International Journal of Advanced Research in
it could hit upon on different real-time Computer Science REVIEW PAPER Available
environment dataset. Online at www.ijarcs.info© 2015-19, IJARCS
All Rights Reserved 356 ISSN No. 0976-
We implement this project as more feasible as the
5697INTRUSION DETECTION SYSTEMS: A
basis of K-means algorithm.
REVIEW D. Ashok Kumar,S. R. Venugopalan
Page no:- [356-357]
V. REFERENCES
[10] Patcha, A. and Park, J. M. An overview of
anomaly detection techniques: Existing
[1] Adaptive Network Security: Solutions for solutions and latest technological
Managing Risk in an Interconnected World, trends.Computer Networks, 51(12);2007; 3448–
AberdeenGroup, Vol. 11, No. 5, January 1998. 3470.
[2] Hacker Stoppers? -- Companies Bought $65 [11] Forrest, S., Hofmeyr, S. A. , Somayaji, A. and
Million Worth Of Network-Intrusion Tools Longstaff, T. A. A Sense of Self for Unix
Last Year, But Capabilities Still Lag What's Processes, IEEE Symposium onResearch in
Promised, Information Week, April 1998 Security and Privacy, Oakland, CA, USA,
http://www.techweb.com/se/directlink.cgi?IW 1996;120--128.
K19980420S0066 [12] Duan, Z., Chen, P., Sanchez, F., Dong, Y.,
[3] Adaptive Network Security Management: Stephenson, M. and J. M. Barker, M. (2012).
Intrusion Detection and Security Assessment Detecting spam zombies bymonitoring
Come of Age, The Yankee Group Data outgoing messages, IEEE Trans. Dependable
Communications Report, Vol. 13, No, 10, June and Secure Computing, Apr 2012; 9(2):198–210
1998. [13] Gu, G., Porras, P., Yegneswaran V., Fong, M.
[4] Mark Crosbie and Gene Spafford. Active and Lee, W. BotHunter: detecting malware
defense of a computer system using infection through IDS-drivenialog correlation,
autonomous agents. Technical Report 95-008, Proc. of 16th USENIX Security Symp. (SS ’07),
COAST Group, Departmentof Computer Aug. 2007; 12:1–12:16.
Sciences, Purdue University,West Lafayette, IN [14] Gu, G., Zhang, J. and Lee, W. (2008).
47907-1398, Feb 1995. BotSniffer: detecting botnet command and
[5] Kumar S, Spafford EH (1994) An Application
ofPattern Matching in Intrusion Detection. control channels in network traffic, Proc.
TechnicalReport CSD-TR-94-013. Purdue Of15th Ann. Network and Distributed Sytem
University.
Security Symp. (NDSS ’08),

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


543
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

[15] Goyal, A. and Kumar, C. .GA-NIDS: A Genetic for detecting network intrusion and misuse.
Algorithm based Network Intrusion Detection Computers and Security, 12(3):235-248, May
System, Electrical Engineering and Computer 1993.
Science, North West University, Technical [23] S. R. Snapp, J. Brentano, G. V. Dias, T. L. Goan,
Report;2008.Feb. 2008. L. T. Heberlein, C. Ho, K. N. Levitt, B.
[16] Abdullah, B., Abd-algafar I., Salama G. I. and Mukherjee, S. E. Smaha, T. Grance, D. M. Teal,
Abd-alhafez A. Performance Evaluation of a and D. Mansur. DIDS (Distributed Intrusion
Genetic Algorithm BasedApproach to Network Detection System) - Motivation, Architecture,
Intrusion Detection System, Proceedings of and an early Prototype. In Proceedings of the
13th International Conference on Aerospace 14th National Computer Security Conference,
Sciences andAviation Technology (ASAT-13), pages 167-176, October 1991.
Military Technical College, Cairo, Egypt, [24] Biswanath Mukherjee, Todd L. Heberlein, and
2009;1-5. Karl N. Levitt. Network intrusion detection.
[17] Ojugo, A. A., Eboka, A. O., Okanta, O. E., Yora, IEEE Network, 8(3):26-41, May/June 1994.
R. E. and Aghware, F. O.Genetic Algorithm [25] Thomas H. Ptacek and Timothy N. Newsham.
Rule-Based IntrusionDetection System Insertion, evasion, and denial of service:
(GAIDS), Journal of Emerging Trends in Eluding network intrusion detection. Technical
Computing and Information Sciences, report, Secure Networks, Inc., January 1998.
3(8);2012; 1182 – 1194. [26] S. Staniford-Chen, S. Cheung, R. Crawford, M.
[18] Roshani Gaidhane, Vaidya, C. and Dilger, J. Frank, J. Hoagland, K. Levitt, C. Wee,
Raghuwanshi, M. Survey.Learning Techniques R. Yip, and D. Zerkle. GrIDS: A graph based
for Intrusion Detection System intrusion detection system for large networks.
(IDS),International Journal of Advance In Proceedings of the 19th National
Foundation and Research in Computer Information Systems Security Conference,
(IJAFRC) Feb 2014. ISSN 2348 – 4853, volume 1, pages 361-370. National Institute of
2014;1(2). Standards and Technology, October 1996.
[19] Gaikwad, Sonali Jagtap, D.P. Kunal Thakare [27] Gregory B. White, Eric A. Fisch, and Udo W.
and Vaishali Budhawant. Anomaly Based Pooch. Cooperating security managers: A peer
Intrusion Detection System UsingArtificial based intrusion detection system. IEEE
Neural Network and fuzzy clustering., Network, pages 20-23, January/February 1996.
International Journal of Engineering Research [28] Review Paper on Shallow Learning and Deep
& Technology (IJERT), ISSN:2278-0181, Learning Methods for Network security Afzal
November- 2012; 1(9). Ahmad1*, Mohammad Asif2, Shaikh Rohan Ali
[20] L. Heberlein, G. Dias, K. Levitt, B. Mukherjee, 3 (page: -52) 2018
J. Wood, and D. Wolber. A Network [29] Youssef Ahmed and Ahmed Emam, “Network
SecurityMonitor. In Proceedings of the IEEE Intrusion Detection Using Data Mining and
Symposium on Research in Security and Network Behavior Analysis”, International
Privacy, May 1990. Journal of Computer Science & Information
[21] R. Heady, G. Luger, A. Maccabe, and M. Technology (IJCSIT) Vol 3, No 6, Dec 2011.
Servilla. The Architecture of a Network Level [30] X. Wu, V.Kumar, Ross Quinlan, Joydeep
Intrusion Detection System. Technical report, Ghosh, Qiang Yang, Hiroshi Motoda, Geoffrey
University of New Mexico, Department of J. McLachlan, Angus Ng, Bing Liu, Philip S. Yu,
Computer Science, August 1990. Zhi-Hua Zhou, Michael Steinbach, David J.
[22] Judith Hochberg, Kathleen Jackson, Cathy Hand, and Dan Steinberg, “Top 10 algorithms
Stallings, J. F. McClary, David DuBois, and in data mining”, Survey Paper(2008).
Josephine Ford. NADIR: An automated system

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


544
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

[31] Intrusion Detection Based On Clustering Cite this article as :


Algorithm Nadya El MOUSSAID 1, Ahmed
TOUMANARI 2, Maryam ELAZHARI 3 Page- Mohammad Asif, Pratap M. Mohite, Prof. P. D. Satya,
1062 "Machine learning algorithm for Cyber Security - A
[32] P. S. Rath, M. Hohanty, S. Acharya and M. Review", International Journal of Scientific Research
Aich, “Optimization of IDS Algorithms Using
in Computer Science, Engineering and Information
Data Mining Technique”, Proceeding of 53rd
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
IRF International Conference, Pune,
India,ISBN 978-93-86083-01-2, 2016. 5 Issue 1, pp. 535-545, January-February 2019.
[33] L.S. Parihar and A. Tiwari, “Survey on Journal URL : http://ijsrcseit.com/CSEIT1951141
Intrusion Detection Usingn Data Mining
Methods”,IJSART, , Volume-2 Issue-1 ISSN
(online: 2395-1052) January-2016.

Authors Profile

Mohammad Asif received his B.E. degree


from NMU JALGAON INDIA.
Currently pursuing M.Tech.in Computer
Engineering from S.Y.C.E.T. Aurangabad
Affiliated to BATU Lonere.

Prof. Pratap Mohite B.Tech CSE


M.E.(Software Engineerng), Assistant
Professor Department of computer
Science and Engineering Shreeyash
College of Engineering and Technology,
Satara Parisar,Beed Bypass Road
Aurangabad.

Prof. Satya P.D. B.Tech. CSE, M.Tech.


and P.hD.(Pursuing), Assistant Professor
Department of computer Science and
Engineering Shreeyash College of
Engineering and Technology, Satara
Parisar,Beed Bypass Road Aurangabad.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


545
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951100
Analysis and Importance of Deep Learning for Video Aesthetic Assessments
Manish Bendale, Madhura V. Phatak, Dr. Nitin N. Pise
Department of Computer Engineering, MIT, Pune, Maharashtra, India
ABSTRACT

Deep Learning is one of the active analysis topic obtaining a great deal of analysis attention recently. This
increase in analysis interest is driven by several area as that are being worked on like machine-based reality
finding, good over-seeing, sensory activity recognition, online learning, world of advertisement, text analysis
and so on. Videos have specific characteristics that make their method unique. Visual aesthetic typically:
Remember what they see, understand and learn rather than what they hear. This paper principally emphasizes
deep learning on basics of automatic video aesthetic assessments.
Keywords : Video Aesthetics, Preprocessing Techniques, Feature Extraction, Deep Learning, Applications.

I. INTRODUCTION It is to be noted that ''Deep Learning'' does not


mingle with the standard of computers emotions. It
Computer technology are playing a very important would be imperative to first begin with human
and necessary role in our daily life. Mainly brain structures, to get a basic grasp on them and
computer involved already in many tasks of thus to outline the relation with computers. Then
gathering information using calculation, logic, we based on human psychology, the computers state of
can say computers are very well predict complex being are going to be put into use under the
events in the future like the real condition of thought of Deep Learning and its completely
weather, can able to manage complex systems and different points of view.
architecture like nuclear and atonic power plants, II. DEEP LEARNING
but still we do not take into account the state they
2.1 How video aesthetics are related to Deep Neural
are in the form of ''living''. Computers are very
Network?
different from humans from the perspective of
“thinking of mind”. As for the person, feeling
Deep learning is a part of machine learning
behind beauty could be a special kind of having
technology which has ability to learn and mimic the
thoughts; its Target is special outlines not forever
working principal of human brain. To mimic the
in agreement like “Value”. Because the underlying
brain working structure, deep neural network
values become completely different, from our
required large amount of datasets [2] to learn. Deep
psychological thinking amendment. There’s no
learning is a subset and the new research area of
great purpose or quality not like between different
machine learning and its original goal is: Artificial
science like calculation or physics. So, to bridge
Intelligence. Deep learning has ability to learn from
over the limit of computers, the new research
the multiple levels of representation and the
concept of ''Deep Learning'' is introduced by
abstraction, these layers are helps in processing the
Walter Pitts and Warren McCulloch in 1943 [1].
amount of data like, images, videos, text, sound, etc.

CSEIT1951100 | Received : 20 Feb 2019 | Accepted : 28 Feb 2019 | January-February -2019 [ 5 (1) : 546-554 ]
546
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

Deep learning is a part of machine learning Aesthetics assessment is a subjective field [4] because
technology which has ability to learn and mimic the individual preferences differ according to personal
working principal of human brain. To mimic the taste too, thus what may be pleasing to one person
brain working structure the deep learning technology may not be pleasing to the other. Therefore different
required more number of datasets [3]. communities define aesthetics differently, based on
psychological and emotional aspects [8].
There is a different types of media aesthetics visions,
video aesthetics is one of the aesthetics, due to the
increasing in the popularization of high quality smart
phones, the videos and image aesthetics are in great
demands. In the current market the videos are
playing important role in various fields like,
advertisement, video conferencing, industrial
representation etc. Previous works done using
machine learning approaches, also called as
handcrafted method in which, feature extraction 2.3 Comparisons of Deep Learning with previous
were a common ways and methods for aesthetic method about Video Aesthetic Assessment
analysis. CNN is recently used method to solve
critical and complex problems as state-of-arts All the present system is based on the handcrafted
approaches [4]. methodologies as we find in literature. Hand-crafted
feature for images and videos aesthetics is good
feature extractors based on user preferences
2.2 Which are the features that can be identified?
algorithms, to design this features required a
considerable amount of engineering skill and domain
Many other specific photographic rules can be sued expertise. The main limitation is the excessive time
to discover particular aesthetic of videos by directly consumption and the accuracies of aesthetic
measuring physiological facts. This understanding assessments. There are many researches done on the
and captured information is used to processed and image aesthetics using hand crafted as well as deep
extract meaningful information from the gathered learning methods [16].
data. In the earlier research, aesthetics was defined
by Handcrafted and learned features [6]. Video aesthetics using deep learning, there exists a
scope for future research as this field is less explored
[19]. The table (table 1) shows the difference
The aesthetic quality of images and videos are judged
between the hand crafted and deep learning
by their low, middle and high level properties. The
technology:
low level properties of an image are that of color,
texture, edges and intensity [7]. The middle level
Deep Learning
property is the object in the image or video. And the Handcrafted Method
Method
high level properties are the photographic rules,
Handcrafted features Deep learning
mainly comprising of the Rules of Thirds (RoT),
judge by commonly method work on
Visual Balance (VB), Diagonal Dominance (DD),
established highly and accurate
Simplicity and the Depth of Field (DoF) [7].

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


547
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

photographic rules, feature extractions 2.4 Areas of Deep Learning related to video aesthetic
mainly low and middle by using hidden assessments
level features. layer concept. • Detecting and recognizing aesthetic information
This is sequential step- This is parallel Aesthetics is a study that relates to the relationship
by-step method and method using hidden between the mind and emotions in assessing beauty
time consuming to layers concept and [8]. The aim of aesthetics analysis is to define the
extract and classification required less amount science in assessing the aspects of arts and beauty
of features. of time. sought in images and videos. Human brain thinking
This method require This method structure begins with gathering information from
engineering skills and depends on the different environment which take fact about users
human experience and amount of dataset to psychology and understanding with behavior. The
do not need large train the network for fact gathered are of the form of features which are
dataset to train network. accuracy. contains in a video and this features used to perceive
Backpropagation is not Backpropagation is people understanding behind videos beauty. For
allowed to correct the main advantage of example, a camera might take pictures of the any
final results. CNN. object; Then human mind start its working and
Classification done at Allow to use more identify the object position, color contrast between
once only. fully connected layer foreground and background, Depth of Field, and etc.
for classification. • Psychology in machine
Another area within video aesthetic computing is
Table 1. Difference between the handcrafted and that the design of procedure systems to exhibit either
deep learning method. innate robust mind psychological powers or that is in
a position of as definitely acting the part of
There are the many researcher working on the deep aesthetics. An additional helpful move involving
learning to improve more accurate result in different computers based mostly on current technology based
field also. In the today life, the video aesthetic is the powers, is that the simulation of aesthetic in taking
important multimedia application. Many field use the pleasure in communication agents to boost
video for improver the selling quantity of the interactivity between human like and machine like
products. So, the video aesthetic is self-increasing communications. Thinking ability in machines could
area of aesthetic and in the future, it will go on the be connected with define states connected with
great demand. The deep learning provides a good forward development (or exist without of forward
opportunity for the video aesthetic assessment to development) of a pc primarily based learning
improve the aesthetic time and the accuracy of video system supported rules [8].
aesthetic editing. In this method, we deploy the
2.5 Another applications related to video aesthetic
multiple method of photography on the neural
network and extract the good feature using the Consider online-learning applications based on
hidden layer of neural network. computers wherein computing the users feelings can
be used to adjust the conference learning, content
can be more polished and tasteful form may be used
while teaching when a learner is uninterested, or
paused. Human Psychological being healthy,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


548
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

suggestions behind aesthetic, can help to the users, 2.6 Different Algorithms Employed by Researchers
based on current emotion behind video beauty by In the work on regression and classifying video
computing applications when working out a aesthetic, the most frequently used classifier
particular user’s state of feelings [9]. algorithms are Artificial Neural Networks (ANN),
Deep Neural Network, Hidden Markov Models
Visual aesthetic has possible applications in human
(HMMs), Gaussian Mixture Model (GMM), Support
computer interaction, now a day human computer
Vector Machines (SVM), k-nearest neighbor (k-NN),
interaction is varying area in different fields. Many
and Decision Tree Algorithms. Numerous studies
researchers are working on gesture technology, in
have decisively showed that choosing the acceptable
which camera capture the video of moving body
classifier will meaningfully improve the performance
gesture and according to movement of gesture the
of the system.
machine will understand, what action required on
the basis of gesture. Virtual reality is also a greater Below gives a brief description about algorithm:
area in human computer interaction, which shows
• K-NN – Classification prediction by defining the
the 3D video and create environment according to
object in the feature space, and comparing this
human psychology and emotion at the time of
object with “k” nearest neighbors. In which the
interactions.
majority vote predict the classification of
gathered data.
Video aesthetics in Advertisements [10]: Now a day,
• GMM – it is a probabilistic based model used for
the increasing in the population the requirement of
signifying and classification of the presence of
the peoples goes on high demand. Due to renewed
sub part within the whole pats. Every sub part is
and changing consumer demands and the rapidly
outlined exploitation the mixture distribution
developing technological factors, companies need to
which permits for classification of explanations
search of new strategies to make a difference in
into the sub-parts [27].
their products. As competition increased marketers
• SVM – is used to binary linear classification and
started to focus on new approaches for product to
it is also used for regression, which selects in the
attract consumers' perception and attention. One of
two (or more) different classes, each input may
the most effective ways of differentiating is using
fall into this class [28].
video aesthetics.
• ANN – is a mathematical model, which has
In gamming area a mistake that’s unfortunately only ability to mimic brain structure of working
too common within the gaming business, is to lump process, it is enthused by biological neural
video graphics and video aesthetics together as if they networks that can work and extract feature
were the same factor. This, unnecessary to say, is similar to human mind [29].
false. The distinction becomes obvious once one • Decision Tree algorithms – are supported
compares, as an example, a heavily artificial, cartoon succeeding a decision tree throughout that the
game that aims for realism specially else. The leaves of the tree represent the classification
former’s video aesthetic, despite its presumptively result, and branches represent the conjunction
smaller budget and scope, can possibly hold up longer of consecutive features that lead to the
than the latter’s. classification [30].
Video aesthetic is also being applied to the growth • HMMs – An applied mathematics markov model
of communicative technologies for use by people during which the conditions and state
[10]. transitions are unobtainable for the observation.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


549
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

In its place, the series of outputs dependent on recognize and classify by the deep learning method
the states are visible [31]. to predict accurate results.

Other applications are like advertisement [20] and


2.7 Artificial Intelligence and Deep Learning online learning, is great in demands, and video
In ‘Streaming Media magazine by Nadine Krefetz aesthetic has ability to make perfect communication
[33]’, defined Video AI (artificial intelligence) has the between the industries and their client. And for the
capability to resolve variety of time-consuming, child, learning environment creation according to
video-related issues with automation. However that child psychology is great aspects of AL and deep
doesn’t mean its supernatural powers which will learning in video assessment areas.
exclude human brain thinking behind beauty control. III. FUTURE SCOPE
To supply a way of where video AI is in early 2018,
what follows are variety of real-life examples during 3.1 Deep learning and Market of video aesthetic
which AI helps to feature structure to the application
unstructured world of video aesthetic assessment. It is likely that the Deep Learning market will
revenue in the Video Aesthetic Advertising segment
At current years, there are many hype of applied
amounts to US$27,799m in 2018. Revenue is
science and especially within the field of artificial
expected to show an annual growth rate (CAGR
intelligence. Inspiring the machines to find out from
2018-2023) of 13.7%, resulting in a market volume of
the information which are gathered from different
US$52,760m by 2023. The average revenue per
source and build accurate decisions within the areas
Internet user currently amounts to US$9.34 [34].
like in the field of education, in medical and
healthcare industries, finance and etc. There are Deep Learning is about to revolutionize the approach
multiple applications which are the mix of each organizations, particularly across sectors like retail,
artificial intelligence and Deep Learning. healthcare, government & defense, and academe,
gather, organize, collaborate, and convey data. This is
often a large rising within the world IT market and
holds an awfully strong growth potential.

This currently innovative technology permits


devices or machines to assemble, predict, resolve
complexity, scrutinize, and decide future outcomes
and be sort of a sophisticated and intelligent system,
like human brain working structure. The system
captures all dimensions of sensory vertex and
Fig. 6 Surveillance with Deep learning and AI different physiological and psychological changes in
Surveillance and security are the major concern of subjects/people beneath observation. Deep Learning
daily life, now a day, and every field contain video mechanism relies on the component of the human
camera, with the intelligence in which when any brain that plays a large role in higher psychological
unwanted event happened then camera or mental process by examining the behavior of a
automatically operate and collect the large amount of human’s thinking and taking choices and decision
data continuously. Internally this complex data are on the concept of it slow series analysis [22].The
Deep Learning systems incorporate technologies as

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


550
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

well as contact primarily based and non-contact


technologies and captures human brain calculation
through software system ways like neural analytics,
gesture recognition, speech recognition, natural
language processing and different physiological
recognition features, to capture, archive, and discuss
with existing knowledge [25].

3.2 Comparison between The Gartner Hype Cycle

"The Gartner hype Cycle for emerging Skills” as


shown in figure 4 is that the broadest hype Cycle,
which incorporates that includes technologies that Fig 7. Gartner Hype Cycle 2015
are the stress of attention due to significantly nice
levels of interest, and people that Gartner believes
have the potential for important impression. Betsy
Burton distinguished analyst at Gartner says, "We
encourage CIOs and alternative IT front-runners to
devote time and energy focused on innovation,
instead of simply business progression, whereas
additionally gaining inspiration by scanning on the
far side the bounds of their trade."

“Deep Learning” isn’t on the graph for the first time


since 2015”, on that time people do not motivated
about the Deep learning technology(see figure 7). In
the 2018 Garter hype Cycle for rising Technologies Fig 8. Gartner Hype Cycle 2018
(see figure 8) the location of “Deep Learning” within
the transition section is on the “Peak” whereby a lot IV. CONCLUSION
of analysis focus is driving expectations from the
sector on top of ever before and within the next Aesthetic assessment of videos and the images is a
decade it's expected to peak because the technology challenging field. Many researchers are working on
matures and analysis outcomes are reported from the classification techniques and they find that
varied analysis teams [32]. there is no foolproof technique for the classification
and prediction [2]. The field of aesthetic assessment
is subjective [17] and thus deep learning provides a
multi-level learning method to get accurate
classification [22]. Here deep learning provides the
automatic aesthetic assessment for videos and
images [10]. The image features provided for the
video motion feature are extraction, which can
provide to be broadly applicable for video aesthetic

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


551
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

assessments. The frame conversion and the shots Detection Method for Art Images”, IEEE
detection is help to understand the characteristics Photonics Journal ( Volume: 9, Issue: 6, Dec.
of video motion to good shot quality of videos. In 2017
the future, the feature extraction and the [7]. Roberto Gallea, Edoardo Ardizzone, and
classification both using the deep learning to take Roberto Pirrone, “Automatic Aesthetic Photo
more accurate result on aesthetics. Many editor are Composition”, Image Analysis and Processing –
working on the animation work in the videos, using ICIAP 2013: 17th International Conference,
deep learning we can modify and get the better Naples, Italy, September 9, 2013.
result over that videos. The video transmission is [8]. Stephen Groening, “Introduction: The
based on the video size, so we can use the deep Aesthetics of Online Videos”, The Aesthetics of
learning base aesthetic video transmission with less Online Videos, June 2016,
size and high aesthetic assessments. DOI: http://dx.doi.org/10.3998/fc.13761232.004
0.201
[9]. christos tzelepis, eftichia mavridaki, vasileios
V. REFERENCES
mezaris, ioannis patras, “Video aesthetic quality
assessment using kernel support vector
[1]. Dhiraj Joshi, Ritendra Datta, Elena
machine with isotropic gaussian sample
Fedorovskaya, Quang-Tuan Luong, James Z.
uncertainty (ksvm-igsu)”, IEEE Int. Conf. On
Wang, Jia Li, and Jiebo Luo, “Aesthetics and
Image Processing (Icip 2016), Phoenix, Az, Usa.
Emotions in Images”, IEEE Signal Processing
[10]. Jun-Ho Choi, Jong-Seok Lee, “Automated
Magazine, Volume: 28, Issue: 5, Sept. 2011
Video Editing for Aesthetic Quality
[2]. Xin Lu, Zhe Lin, Hailin Jin, Jianchao Yang, and
Improvement”, 23rd ACM international
James. Z. Wang, “Rating Image Aesthetics using
conference on Multimedia, Brisbane, Australia
Deep Learning ”, IEEE Transactions on
— October 26 - 30, 2015
Multimedia, Volume: 17, Issue: 11, Nov. 2015
[11]. Anselm Brachmann, Christoph Redies,
[3]. Yubin Deng, Chen Change Loy, and Xiaoou
“Computational and Experimental Approaches
Tang, “Image Aesthetic Assessment”, IEEE
to Visual Aesthetics”, Front. Comput.
Signal Processing Magazine, Digital Object
Neurosci., 14 November 2017 |
Identifier 10.1109/MSP.2017.2696576, 11 July
https://doi.org/10.3389/fncom.2017.00102.
2017.
[12]. Aarushi Gaur, “Ranking Images Based on
[4]. Yubin Deng, Chen Change Loy, and Xiaoou
Aesthetic Qualities”, Submitted for the Degree
Tang, “Image Aesthetic Assessment: An
of Doctor of Philosophy from the University of
Experimental Survey”, arXiv:1610.00838v2
Surrey, April 2015.
[cs.CV] 20 Apr 2017.
[13]. Loris Nanni, Stefano Ghidoni, Sheryl Brahna,
[5]. Ritendra Datta Dhiraj Joshi Jia Li James Z.
“Handcrafted vs. non-handcrafted features for
Wang, “Studying Aesthetics in Photographic
computer vision classification”, Elsevier Journal
Images Using a Computational Approach”,
on Pattern Recognition 71 (2017) 158–172,
Computer Science, vol 3953. Springer, Berlin,
2017.
Heidelberg, ECCV 2006
[14]. Hsin-Ho Yeh, Chun-Yu Yang, Ming-Sui Lee,
[6]. Chongke Bi ,Ye Yuan, Ronghui Zhang , Yiqing
and Chu-Song Chen, “Video Aesthetic Quality
Xiang,Yuehuan Wang, and Jiawan Zhang, “A
Assessment by Temporal Integration of Photo-
Dynamic Mode Decomposition Based Edge

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


552
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

and Motion-Based Features”, IEEE Transactions classification,” in IEEE Conference on


on Multimedia, Vol. 15, No. 8, December 2013. Computer Vision and Pattern Recognition
[15]. Katharina Schwarz, Patrick Wieschollek, (CVPR), pp. 3642–3649, 2012.
Hendrik P. A. Lensch, “Will People Like Your [23]. Y. Sun, X. Wang, and X. Tang, “Hybrid deep
Image? Learning the Aesthetic Space”, learning for face verification,” in The IEEE
arXiv:1611.05203v2 [cs.CV] 4 Dec 2017 International Conference on Computer Vision
[16]. Yanran Wang, Qi Dai, Rui Feng, Yu-Gang (ICCV), 2013.
Jiang, “Beauty is Here: Evaluating Aesthetics in [24]. Jingwei Xu, Li Song, Rong Xie, "Shot boundary
Videos Using Multimodal Features and Free detection using convolutional neural
Training Data”, ACM international conference networks", Visual Communications and Image
on Multimedia, Pages 369-372, Barcelona, Processing (VCIP) 2016, pp. 1-4, 2016.
Spain — October 21 - 25, 2013 [25]. Simone Bianco, Luigi Celona, Paolo
[17]. Yuzhen Niu, Feng Liu, “What Makes a Napoletano, and Raimondo Schettini,
Professional Video? A Computational “Predicting Image Aesthetics with Deep
Aesthetics Approach”, IEEE Transactions on Learning” ,Springer International Journal on
Circuits and Systems for Video Technology, Advanced Concepts for Intelligent Vision
Volume: 22, Issue: 7, July 2012 Systems, 21 October 2016, volume: 10016.
[18]. Subhabrata Bhattacharya, Behnaz Springer, Cham.
Nojavanasghari, Tao Chen, “Towards a [26]. C. D. Manning and H. Schütze, “Foundations of
Comprehensive Computational Model for Natural Language Processing,” Reading, pp.
Aesthetic Assessment of Videos”, ACM 978-1- 678, 2000.
4503-2404-5, 2013, Barcelona, Spain, [27]. P. K. Abhishek Kumar Chauhan, “Moving
http://dx.doi.org/10.1145/2502081.2508119. Object Tracking using Gaussian Mixture Model
[19]. Chun-Yu Yang, Hsin-Ho Yeh, and Chu-Song and Optical Flow,” vol. 3, no. 4, pp. 243–246,
Chen, “Video Aesthetic Quality Assessment By 2013.
Combining Semantically Independent And [28]. D. K. Srivastava, L. Bhambhu, and B. Cet, “Data
Dependent Features”, Acoustics, Speech and classification using support vector machine,”
Signal Processing (ICASSP), IEEE International Journal of Theoretical and Applied Information
Conference, 2011, Prague, Czech Republic. Technology, 2010.
[20]. Yanxiang Chen, Yuxing Hu, Luming Zhang, [29]. M. S. and W. P, “Research Paper on Basic of
Ping Li, Chao Zhang, “Engineering Deep Artificial Neural Network,” International
Representations for Modeling Aesthetic Journal of Recent Innovation Trends Computer
Perception”, IEEE Transactions on Communication, vol. 2, no. 1, pp. 96–100, 2014.
Cybernetics, Volume: PP, Issue: 99 , 11 [30]. P. Sharma and A. P. R. Bhartiya,
December 2017. “Implementation of Decision Tree Algorithm to
[21]. A. Krizhevsky, I. Sutskever, and G. E. Hinton, Analysis the Performance,” Int. J. Adv. Res.
“Imagenet classification with deep Computer Communication. Eng., vol. 1, no. 10,
convolutional neural networks,” in Advances in pp. 861–864, 2012.
Neural Information Processing Systems (NIPS), [31]. F. Park, “The Hierarchical Hidden Markov
pp. 1106–1114, 2012. Model : Analysis and Applications,” vol. 62, pp.
[22]. D. Ciresan, U. Meier, and J. Schmidhuber, 41–62, 1998.
“Multi-column deep neural networks for image [32]. Gartner Hype Cycle:

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


553
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

[33]. https://medium.com/machine-learning-in-
practice/deep-learnings-permanent-peak-on-
gartner-s-hype-cycle-96157a1736e
[34]. NadineKrefetz , streaming media
https://www.streamingmedia.com/Authors/683
7-Nadine-Krefetz.htm

Cite this article as :

Manish Bendale, Madhura V. Phatak, Dr. Nitin N.


Pise, "Analysis and Importance of Deep Learning for
Video Aesthetic Assessments", International Journal
of Scientific Research in Computer Science,
Engineering and Information Technology
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
546-554, January-February 2019. Available at doi :
https://doi.org/10.32628/CSEIT1951100
Journal URL : http://ijsrcseit.com/CSEIT1951100

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


554
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1838116

Automatic Tool for Prediction of Type of Cancer Risk and


Recommendations
Pallavi Mirajkar1, Dr. G. Prasanna Lakshmi2
1Research Scholar, Faculty of Computer Science, Pacific Academy of Higher Education and Research University,
Udaipur, Rajasthan, India
2(WOS-A) Andhra University, AU North Campus, Visakhapatnam, Andhra Pradesh, India

ABSTRACT

Cancer can begin in any part of the body and can spread to other parts also. It is uncontrollable and it has many
types. In the proposed thesis research paper, a tool for prediction of type of cancer risk with five different
cancer diagnosis and recommendations is presented. For recognizing cancer disease number of tests ought to be
required from the patient. But using data mining techniques these test can be diminished. Indeed, an accurate
prediction of cancer is very difficult task for medical practitioner and it is also high concern to the patients so
that better treatment can be given and it will also increase the survival time of the patients. Our findings
suggested that suitable prediction tool can effectively reduce the several tests for diagnosing cancer and
prediction accuracy thereby increasing the technical possibility of early detection of cancer. The main features
of the tool comprise a balance between the number of necessary inputs and prediction performance, being
portable, and it empowers the automatic development of the cancer risk prediction tool in cancer disease.
Keywords : Prediction Tool, Cancer, Data Mining, Automation, Integration.

I. INTRODUCTION able to consult with proposed prediction tool, we


chose to develop cancer prediction tool as an online
The rising high-performance computing has system.
benefited numerous disciplines in finding realistic
solutions to their issues. Our health services are no Here the scope of the prediction tool is that
special case to this. Data mining tools have been integration of various risk factors that causes cancer,
created for useful investigation of medical with computer-based patient records could reduce
information, to help oncologist in improving medical errors, enhance patient safety, improves the
determination for treatment purposes. In cancer prediction of cancer risk in practice variation, and
disease research, data mining technique have played improve patient survival rate. The application is fed
out a noteworthy role. Cancer disease categorization with different details to help medical practitioner to
contributes the unsafe reason for the treatment of predict risk of type of cancer. The application permits
patients. user to share their health connected issues. It then
The aim of thesis work is to present an easy to use processes user’s particular details to determine for
tool that provides predictions of cancer risk in varied cancer disease that might be related to it. Here
patients or individual. Due to the internet facility we tend to utilize some intelligent data mining
available everywhere and the ease with which one is techniques to figure the correct risk level of cancer

CSEIT1838116 | Received : 28 Dec 2018 | Accepted : 03 Jan 2019 | January-February -2019 [ 5 (1) : 01-08 ]
01
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

that may be identified with patient’s details. Based on paper work is very helped to the cancer research
result, system automatically demonstrates the result center and hospitals to prevent the Non Small Cell
to specific doctors for specific test/ treatment. Lung Cancer.

II. LITERATURE REVIEW Arpit Bansal et, al.(2017) have proposed technique for
a modification in K-Means Clustering Algorithm. In
Fan Wang et.al (2018) worked on predictor to proposed modification, the K-Means clustering will
precisely predict six breast cancer cell types using vanish off the two most important drawbacks of K-
gene expression. They identified 308 predictors of Means clustering that are accuracy level and
breast cancer cell types using machine learning. calculation time consumed in clustering the dataset.
They found that mixture of protein coding genes and Although when they have also used small datasets
long non-coding RNAs are superior predictors these two factors accuracy level and calculation time
compared to when the two sets of transcripts are may not matter much but when they used large
treated independently. They have also shown that datasets that have trillions of records, then little
predictors constrained to a particular cell type dispersion in accuracy level will matter a lot and can
provide as better prognostic markers for the relevant lead to a disastrous situation, if not handled properly,
patient subtype. Author has stated that the breast it can be stated that proposed modification can be
cancer cell type predictors are also related with more extended to achieve the full accuracy level up
patient survivability and therefore have clinical to 100%, with very little time and with more quality
significance. clusters.

Subrata Kumar Mandal (2017) have applied different R.Senkamalavalli and Dr.T.Bhuvaneswari (2017) have
techniques such as data cleaning, feature selection, proposed novel algorithm was experimented on the
feature extraction, data discretization and Breast cancer database. It has been proved that
classification for predicting breast cancer as perfectly. proposed approach achieved a very high accuracy
They stated that Logistic Regression Classifier gives rate than the existing methods. They also
the maximum accuracy with reduced subset of demonstrated a certain level of accuracy in the
features and time complexity of proposed algorithm is classifier, and for finding accurate results there must
least compared to other two classifiers. be sufficient preprocessing of data done. They have
also demonstrated accuracy in diagnosing breast
R.Kaviarasi, Dr.A.Valarmathi (2017) has applied two cancer disease using the K-means classifier, adaboost
types of clustering. The hierarchical clustering is and Support Vector Machines.
produced dendrogram results are produced using
Euclidean distance and Ward.D linkage. The K- Sumalatha.G, Archana.S (2017) has analyzed cancer
Means clustering are produced WSS values against patient data using ZeroR method and J48 algorithm
number of cluster K values. The proposed method of data mining techniques. They stated that proposed
finally validate to the two type of clustering fit values. cancer risk prediction system may give easy and a
The validation measurement result is helped to the cost effective approach for screening cancer and may
distance are measured in two clustering values. The play a significant role in earlier diagnosis process for
results are helped to at the beginning of Non Small different types of cancer. Proposed system can also be
Cell Lung Cancer prevention through human way of used as a source of record with detailed patient
life handled risk factors characteristics. The proposed

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


11
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

history in hospitals as well as help doctors to III. PROPOSED WORK


premeditate on particular therapy for any patient.
Diagnosis of cancer is still challenging for the
Megha Rathi et.al. (2016) have developed a software specialists in the field of medicine. Even now the
tool for the prediction of disease which helps in genuine reason and complete treatment of cancer is
decision making for the treatment method. Proposed not invented. Different tests are available for
tool will be helpful in diagnosing the disease type and predicting cancer, however detecting cancer in
to help out for decision support in medical system. earlier stage is troublesome, yet prior discovery of
They studied hybrid classification technique is used cancer disease is treatable. With the help of data
for classification of medical data sets and is applicable mining we endeavor to predict the risk of cancer in
in healthcare domain. SVM and bootstrap is earlier stage. We build up a system called the cancer
integrated to improve classification accuracy. Author prediction tool which predicts specific cancer risks.
expresses that prediction tool will help doctors or Specifically, Cancer prediction tool estimates the risk
patient to decide in a short time whether the person of the breast, ovarian, oral, stomach, and lung cancers
is suffering from disease and is generic to all types of by analyzing a number of user-provided genetic and
disease. non-genetic factors. The major objective of
prediction tool to provide the earlier warning to the
Prabhakar Chalise et,al.,(2014) have proposed Cluster users, to make a precaution based on their risk status.
analysis aims to highlight meaningful patterns or The proposed prediction tool assists the people to
groups inherent in the data that will be useful in know their chance of cancer risk with ease and
identifying the subtypes of the diseases. Different additionally, it helps the people to take the proper
types of clustering algorithms have been proposed decision based on their cancer risk status. Once the
that use various assays of molecular variation of cells user enters into the cancer prediction system, they
most of which are designed for one type of data at a need to answer the questionnaire, related to cancer
time. These types of methods have been successfully causing factors and symptoms. After that the
implemented in several disease classification studies. prediction system assigns the risk value based on the
user responses. Once the risk value is predicted, the
Rahul Patil et.al (2016) has used various clustering range of the risk can be computed by the prediction
techniques of data mining such as partition based system. We have three levels of risk low level,
method, hierarchical based clustering. They stated medium level, and high level. The result can be given
that hybrid method will help society to analyze and away to the user through data base. The above
understand disease and their health status. mentioned technique can be successfully applied to
the data sets for any cancer (such as breast cancer,
Tanu Minhas and Nancy Sehgal (2017) concluded lung cancer, oral, stomach cancer), as it is effectively
that prediction analysis is an efficient technique for checked on the different types of cancer. Once the
the complex data analysis. They have applied k- risk is calculated the data given by the user is stored
means clustering algorithm with Boltzmann in the data base. The result will be displayed to the
algorithm to increase accuracy of data clustering. user through the database.
They stated that the performance of proposed
algorithm is tested in MATLAB and it has been
analyzed that accuracy is increased up to 20 percent.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


12
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

check their risk of cancer. Collected data also


User Interface contains both patient and non patient data. From the
previous studies total 18 risk factors were considered
for risk level cancer assessment, which includes- age,
User Enter Personal gender, hereditary, occupational hazard, smoking,
Information
chewing tobacco, alcohol, daily exercise, intake of
User Fill Questionnaire vegetables, weight loss, weight gain, soft drink,
contraceptive pills and various symptoms of different
types of cancer.
User Select Organ of
Symptoms 4.3 Data Preprocessing
Sometimes it happens that collected data consists of
repetition of information or more information of
Compare the user input with same person or missing any values of information.
data stored in database Therefore, data preprocessing is vital task which is
necessary in data mining. The main objective of data
pre-processing is making an appropriate analysis of
Match the input value of user collected data. Data preprocessing allows avoiding
with stored data in database the double data and adds the missing values
according to the need of tool. It also helps to reduce
the memory and normalizes the values that are stored
into the database.

Store Computed
Score 4.4 System Architecture
The proposed cancer risk prediction tool is developed
Display Results using various web technologies such as HTML, CSS
and Java Script. It is suitable for several platforms and
Fig. 1 Working of Cancer prediction system is distributable through an application store, just like
other applications. The proposed tool is developed
IV. METHODS AND MATERIAL using Angular JS which wraps the HTML/JavaScript
app into a native container which can access the
4.1 Methodology device functions of multiple platforms. When the
Collected data is prepared to fit into the developed medical practitioner submits the values for multiple
tool appropriately. Here converse about the entire queries, the data is sent over the HTML POST
procedure (collection of data to significant pattern) of method. The data is inserted into a Postgre SQL
prediction of different types of cancer risk level. innodb database and, for each new registration. The
proposed tool generates new XML file for each type
4.2 Data Collection of cancer. For the health care industry, cancer
Collection of data for the developed prediction tool prediction applications yielded new boundaries in
has been done from different diagnostic centers providing better care and treatments to patients.
which contain both male female data of different ages Furthermore, it is building a revolution in the way
and from the various non-patients those who want to data is managed and made available [21]. The

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


13
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

portability of proposed prediction applications can


enhance the productivity of health care professionals.

Sends

Queries

Web Based Cancer


Risk Prediction Tool

Fig.4 Personal Information Tab


Updates

Updates
Queries

Uses

Fig. 2 Architecture of Cancer Risk Prediction


4.5 Use Case Tool

Fig. 5 Lifestyle Tab


Figure 2(a) shows the first screen that appears when
the prediction tool is initiated. User has to login and
then clicking on the Test Fig. 2, all tabs containing
questionnaire available in the application become
visible. User has to answer each question as shown in
fig. 4, fig. 5 fig. 6, fig 7and fig. 8.

Fig.6 Eating Habit Tab

Fig. 3 Home Page

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


14
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

Gender Responses
Male 76
Female 93
Total 169
Table 1

VI. CONCLUSION

The thesis paper in which I developed cancer


prediction system based on the online discourse of
patient’s information. It was shown that relying on
Fig. 7 Symptoms of different types of cancers cancer prediction system will give warning to the
user about cancer risk level. Combining the different
approaches boosts results, as content and interaction
seem to model complementary aspects of patient
discourse. The proposed prediction tool can be
helpful to health industry who wants to diagnose
patient online. The proposed system is web based,
easy to use, versatile and reliable that can be
implemented in remote areas like rural regions in
Primary Health center, to mimic like human
diagnostic expertise for finding the chances of having
cancer. The main aim of the proposed prediction tool
Fig. 8 Display of Result
is to provide the earlier warning to the users, and it
also saves cost and time of the user. It predicts
V. RESULTS AND DISCUSSION
specific cancer risks based on the symptoms found in
the patient. Specifically, Cancer prediction system
CANCER PREDICTION SYSTEM –
POPULATION AND SAMPLE RESULTS estimates the risk of the breast, Oral, Stomach,
Ovarian, and Lung cancers by examining a numerous
▪ To find out the performance of the Data Mining user-provided genetic and non-genetic factors. The
Based Cancer Prediction System among the users proposed prediction system is easily accessible;
based on cancer prediction. people can easily check their risk and take
▪ To identify cancer risk level of the user using appropriate action based on their risk status.
newly developed Data Mining Based Cancer
Prediction System based on various risk factors. Future Scope on the prediction tool can include the
development of different models with high end
To find the effectiveness of data mining based cancer technologies to prevent cancer.
prediction system, the proposed system was
implemented on web based technology. Around 169
responses have been collected during Jun to Dec 2018.
Details of the responses given in the table No. 1

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


15
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

VII. REFERENCES [9]. Ahmad LG, Eshlaghy AT, Poorebrahimi A,


Ebrahimi M, Razavi AR (2013) “Using Three
Machine Learning Techniques for Predicting
[1]. Fan Wang, Zachariah Dohogne , Jin Yang , Yu Breast Cancer Recurrence”. J Health Med
Liu and Benjamin Soibam “Predictors of breast Inform 4: 124. doi:10.4172/2157-7420.1000124
cancer cell types and their prognostic power in [10]. V.Kirubha , S.Manju Priya “Survey on Data
breast cancer patients” BMC Genomics (2018) Mining Algorithms in Disease Prediction”
19:137 https://doi.org/10.1186/s12864-018- International Journal of Computer Trends and
4527-y. Technology (IJCTT) – Volume 38 Number 3 -
[2]. Dave Smith, “Data Mining in the Clinical August 2016.
Research Environment”. Available at. [11]. P. Saranya , B. Satheeskumar “A Survey on
http://www.sas.com/ Feature Selection of Cancer Disease Using Data
[3]. Damtew A., “Designing a predictive model for Mining Techniques” International Journal of
heart disease detection using data mining Computer Science and Mobile Computing,
Techniques” A Thesis Submitted to the School Vol.5 Issue.5, May- 2016, pg. 713-719.
of Graduate Studies of Addis Ababa University, [12]. R.Senkamalavalli and Dr.T.Bhuvaneswari
2011. “Improved Classification Of Breast Cancer Data
[4]. https://www.cancer.gov/ Using Hybrid Techniques” International
[5]. Subrata Kumar Mandal “Performance Analysis Journal of Advanced Research in Computer
Of Data Mining Algorithms For Breast Cancer Science Volume 8, No. 8, September-October
Cell Detection Using Naïve Bayes, Logistic 2017 ISSN No. 0976-5697.
Regression and Decision Tree” International [13]. Sumalatha.G, Archana.S “A Study on Early
Journal Of Engineering And Computer Science Prevention and Detection of Breast Cancer
ISSN: 2319-7242 Volume 6 Issue 2 Feb. 2017, using Data Mining Techniques” International
Page No. 20388-20391. Journal of Innovative Research in Computer
[6]. R. Kaviarasi, Dr. A.Valarmathi “Near The and Communication Engineering (An ISO
Beginning of Non Small Cell Lung Cancer 3297: 2007 Certified Organization) Website:
Avoidance in Human Way of Life Risk Factors www.ijircce.com Vol. 5, Issue 6, June 2017.
Classification Using Clustering Algorithm in [14]. Megha Rathi , Vikas Pareek “Disease prediction
the R Environment” International Journal of tool: an integrated hybrid data mining approach
Advanced Research in Computer Science, 8 (5), for healthcare” IRACST - International Journal
May-June 2017,1023-1026. of Computer Science and Information
[7]. Arpit Bansal, Mayur Sharma “Improved k-mean Technology & Security (IJCSITS), ISSN: 2249-
clustering algorithm for prediction analysis 9555 Vol.6, No.6, Nov-Dec 2016.
using classification technique in data mining”, [15]. Rahul Patil, Pavan Chopade, Abhishek Mishra,
IJCA (0975 – 8887) Vol. 157 – No 6, January Bhushan Sane, Yuvraj Sargar “Disease
2017. Prediction System using Data Mining Hybrid
[8]. Prabhakar Chalise, Devin C.Koestler, Milan Approach” Communications on Applied
Bimali, Qing Yu, Brooke L.Fridley to Electronics (CAE) – ISSN : 2394-4714
“Integrative clustering methods for high – Foundation of Computer Science FCS, New
dimensional molecular data” , Transl Cancer York, USA Volume 4 – No.9, April 2016.
Res 2014.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


16
Pallavi Mirajkar, Dr. G. Prasanna Lakshmi Int J S Res CSE & IT. January-February-2019 ; 5(1) : 01-08

[16]. Tanu Minhas , Nancy Sehgal “Prediction Author’s :


Analysis Technique using Hybrid Clustering
and SVM Classification” International Journal Mrs. Pallavi Mirajkar passed M.Sc.
(Computer Science) from North
of Innovative Research in Science, Engineering
Maharashtra University, Jalgaon.
and Technology (An ISO 3297: 2007 Certified
Currently she is pursuing her Ph.D.
Organization) Website: www.ijirset.com Vol. 6, from Pacific University, Udaipur,
Issue 7, July 2017. Rajasthan, India
[17]. K. Suneetha “Early Prediction and Detection of
Lung Cancer using Data Mining” International Dr. G.Prasanna Lakshmi, Guide.
Journal of Advanced in Management,
Women Scientist (WOS-A) Andhra
Technology and Engineering Sciences Volume
University, Andhra Pradesh, India
7, Issue 12, 2017 ISSN NO : 2249-7455.
[18]. Chih-Jen Tseng, Chi-Jie Lu, Chi-Chang Chang,
Gin-Den Chen, Chalong Cheewakriangkrai
“Integration of data mining classification
techniques and ensemble learning to identify
risk factors and diagnose ovarian cancer Cite this article as :
recurrence” Contents lists available at Science
Direct Artificial Intelligence in Medicine Pallavi Mirajkar, Dr. G. Prasanna Lakshmi,
"Automatic Tool for Prediction of Type of Cancer
journal homepage:
Risk and Recommendations", International Journal of
www.elsevier.com/locate/aiim/ Artificial
Intelligence in Medicine 78 (2017) 47–54. Scientific Research in Computer Science, Engineering
and Information Technology (IJSRCSEIT), ISSN :
[19]. Huang M-W, Chen C-W, Lin W-C, Ke S-W,
2456-3307, Volume 5 Issue 1, pp. 01-08, January-
Tsai C-F (2017) “ SVM and SVM Ensembles in
February 2019.
Breast Cancer Prediction” PLoS ONE
Available at doi
12(1):e0161501.
: https://doi.org/10.32628/CSEIT1838116
doi:10.1371/journal.pone.0161501.
Journal URL : http://ijsrcseit.com/CSEIT1838116
[20]. Florije Ismaili , Luzana Shabani , Bujar Raufi ,
Jaumin Ajdari , Xhemal Zenuni “Enhancing
breast cancer detection using data mining
classification techniques” 2nd World
Conference on Technology, Innovation and
Entrepreneurship May 12- 14, 2017.
[21]. Ana Silva, Tiago Oliveira, Vicente Julian, Jose
Neves, and Paulo Novais “A Mobile and
Evolving Tool to Predict Colorectal Cancer
Survivability” Published by Springer
International Publishing Switzerland 2016. All
Rights Reserved L. Iliadis and I. Maglogiannis
(Eds.): AIAI 2016, IFIP AICT 475, pp. 14–26,
2016.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


17
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19511

Movie Recommended System by Using Collaborative Filtering


Bheema Shireesha1, Navuluri Madhavilatha2, Chunduru Anilkumar3
1,3 Assistant Professor, Department of Computer Science and Engineering, Dr. APJ Abdul Kalam IIIT Ongole,
Andhra Pradesh, India
2Guest Faculty, Department of Computer Science and Engineering, Dr. APJ Abdul Kalam IIIT Ongole, Andhra

Pradesh, India

ABSTRACT

Recommendation system helps people in decision making an item/person. Recommender systems are now
pervasive and seek to make profit out of customers or successfully meet their needs. Companies like Amazon
use their huge amounts of data to give recommendations for users. Based on similarities among items, systems
can give predictions for a new item’s rating. Recommender systems use the user, item, and ratings information
to predict how other users will like a particular item. In this project, we attempt to under- stand the different
kinds of recommendation systems and compare their performance on the Movie Lens dataset. Due to large size
of data, recommendation system suffers from scalability problem. Hadoop is one of the solutions for this
problem.
Keywords : Hadoop, Collaborative filtering, Machine Learning, HDFS, RSSI

I. INTRODUCTION then the user will simply stop using it. This has led to
a high emphasis by tech companies on improving
Recommendations system is a type of information their recommendation systems. However, the
filtering system which attempts to predict the problem is more complex than it seems. For
preferences of a user, and make suggests based on recommendation, our proposed system uses
these preferences. There are a wide variety of collaborative filtering machine learning algorithm.
applications for recommendation systems. These have Collaborative filtering (CF) is a machine learning
become increasingly popular over the last few years algorithm which is widely used for recommendation
and are now utilized in most online platforms that we purpose. Collaborative filtering finds nearest neighbor
use. The content of such platforms varies from movies, based on the similarities. The metric of collaborative
music, books and videos, to friends and stories on filtering is the rating given by the user on a particular
social media platforms, to products on e-commerce item. Different users give different ratings to items.
websites, to people on professional and dating Users, who give almost same rating to items, are the
websites, to search results returned on Google [1]. nearest neighbors. In case of User based collaborative
Due to the advances in recommender systems, users filtering, based on the ratings given by the users,
constantly expect good recommendations. They have nearest neighbors has been find. Item based
a low threshold for services that are not able to make collaborative filtering predicts the similarity among
appropriate suggestions. If a music streaming app is items. To recommend an item, items which are liked
not able to predict and play music that the user likes, by the user in his past have been found. Item which is

CSEIT19511| Received : 01 Jan 2018 | Accepted : 03 Jan 2019 | January-February -2019 [ 5 (1) : 09-15 ]
09
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar Int J S Res CSE & IT. January-February-2019 ; 5(1) : 09-15

similar to those items has been recommended [2]. measurements and analysis were made. The
Internet contains a huge volume of data for remaining part of this report is as follows: Related
recommendation purpose. Due to size of data, if Works, Measurement Setup, Proposed Work,
recommendation computation has been done in single Measurements Analysis, Results, Discussions, Future
system, then performance may degrade, and we Work.
cannot find an efficient solution. Hence we require
distributed environment so that computation can be II. RELATED WORK
increased and performance of recommendation
system gets improve .My goal is to apply a 2.1 Collaborative Filtering
collaborative filtering algorithm in a rating website
that collects users’ information, such as location and In the collaborative filtering algorithm, the system
gender, item’s information, such as category and has a recorded set of items and users and how the
description, as well as ratings for items by users. There users rated those items. Then algorithm is used to
are many algorithms that could be applied on data to predict the rating for a user who has not rated the
predict a user preference. User-based, Item-based, and item yet. A rating for an item can be predicted from
Model-based methods are ways of predicting a user the ratings given to the item by users who are similar
preference. The number of users, items, or clusters in in taste to the given user. The traditional collaborative
each one respectively will determine the function filtering algorithms include User-based, Item-based,
performance [3]. However, the most well-known and and Model based methods. To explain how these
common one is User-based Collaborative Filtering. In methods works we are going to use the following
this algorithm, we predict an item’s rate for a user by notations [6]. “Let U be a set of N users and I a set of
collecting information about this user and similar M items. The traditional collaborative filtering
users. algorithms include User-based, Item-based, and
Model based methods. To explain how these methods
Machine Learning: works we are going to use the following notations.
Applying machine learning in real-time using “Let U be a set of N users and I a set of M items. Vui
Collaborative Filtering. Parsing data retrieved from a denotes the rating of user u U on item i I, and SI
database and predicting user preference. Evaluating stands for the set of items that user u has rated”.
different approaches of recommender systems. What I
was trying to do was to build a system that collects 2.2 User Based Collaborative Filtering
information and then uses the stored data in a In this method, we predict the user behavior against a
machine learning algorithm [4]. Predicting users’ certain item using the weighted sum of deviations
preferences using data may give more accurate results from mean ratings of users that previously rated this
than any algorithm that does not use previous data. item and the user mean rate. The weight that we
Most systems like Amazon, eBay, and others suggest previously mentioned can be calculated using Pearson
things to users based on similarities among users, Correlation according to the following formula:
items, or both. This will make those systems more
personalized and efficient from a user’s perspective.
Commercial and trading systems gain trust and profits
The prediction formula is stated as below:
using such systems if they successfully predict what
users want at what time and where. The datasets were
created and averaged [5]. The corresponding

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


10
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar Int J S Res CSE & IT. January-February-2019 ; 5(1) : 09-15

Often, these systems are able to collect information


about a user’s.

3.2 Definition 2
2.3 Model-Based Collaborative Filtering
In this method, the system will use an unsupervised Collaborative Filtering: Collaborative Filtering
learning method to partition the space and then techniques make recommendations for a user based
classify the users using a similarity metric to a on ratings and preferences data of many users. The
segment or a cluster. To avoid having large clusters, main underlying idea is that if two users have both
systems use different methods to generate more small liked certain common items, then the items that one
practical clusters. This process starts with an initial set user has liked that the other user Search Engine
of clusters that contains only one user, and then Architecture.
repeatedly assigns users to these clusters based on a
similarity metric. Limiting features may become 3.3 Definition 3
necessary to reduce clustering complexity. The system
will create vectors for each segment and match the Content Based Recommendation: Content Based
user to the vector. Nonetheless, the user may be Recommendation algorithm takes into account the
classified as belonging to more than one cluster with a likes and dislikes of the user and generates a User
measure of similarity strength . Profile. For generating a user profile, we take into
account the item profiles (vector describing an item)
and their corresponding user rating. The user profile
is the weighted sum of the item profiles with weights
being the ratings user rated. Once the user profile is
generated, we calculate the similarity of the user
profile with all the items in the dataset, which is
III. PRELIMINARIES
calculated using cosine similarity between the user
profile and item profile. Advantages of Content Based
3.1 Defination1
approach is that data of other users is not required and
the recommender engine can recommend new items
Recommended system: A recommendation system is a
which are not rated currently, but the recommender
type of information filtering system which attempts to
algorithm doesn’t recommend the items outside the
predict the preferences of a user, and make suggests
category of items the user has rated.
based on these preferences. There are a wide variety
of applications for recommendation systems. These
IV. PROPOSED APPROACH
have become increasingly popular over the last few
years and are now utilized in most online platforms
4.1 Proposed System
that we use [7]. The content of such platforms varies
from movies, music, books and videos, to friends and
For recommendation, our proposed system uses
stories on social media platforms, to products on
collaborative filtering machine learning algorithm.
ecommerce websites, to people on professional and
Collaborative filtering (CF) is a machine learning
dating websites, to search results returned on Google.
algorithm which is widely used for recommendation
purpose. Collaborative filtering finds nearest neighbor

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


11
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar Int J S Res CSE & IT. January-February-2019 ; 5(1) : 09-15

based on the similarities. The metric of collaborative implementations, consult the Breese etc. All paper
filtering is the rating given by the user on a particular and/or our code.
item. Different users give different ratings to items.
Users, who give almost same rating to items, are the Item-based collaborative filtering is a model-based
nearest neighbors. In case of User based collaborative algorithm for making recommendations. In the
filtering, based on the ratings given by the users, algorithm, the similarities between different items in
nearest neighbors has been find. Item based the dataset are calculated by using one of a number of
collaborative filtering predicts the similarity among similarity measures, and then these similarity values
items. To recommend an item, items which are liked are used to predict ratings for user-item pairs not
by the user in his past have been found. Item which is present in the dataset. The system will search group
similar to those items has been recommended. for users by using k-means to find the distance
Internet contains a huge volume of data for between users, the group of users and clustering of
recommendation purpose. Due to size of data, if users. The system clustering with the k- means
recommendation computation has been done in single algorithms by measuring the distance of each data
system, then performance may degrade, and we point from the center of the 10 groups by using
cannot find an efficient solution. Euclidean distance and calculated information will be
stored in the database.
Hence we require distributed environment so that
computation can be increased and performance of 4.2 User Based CF
recommendation system gets improve. My goal is to
apply a collaborative filtering algorithm in a rating The dataset is firstly loaded into Hadoop distributed
website that collects users’ information, such as file system (HDFS). Then we perform User-based CF
location and gender, item’s information, such as using Mahout [9]. We take rating matrix, in which
category and description, as well as ratings for items each row represents user and column represents item,
by users. There are many algorithms that could be corresponding row-column value represents rating
applied on data to predict a user preference. User- which is given by a user to an item. Absence of rating
based, Item-based, and Model-based methods are value indicates that user has not rated the item yet.
ways of predicting a user preference [8]. The number There are many similarity measurement methods to
of users, items, or clusters in each one respectively compute nearest neighbors. We have used Pearson
will determine the function performance. However, correlation coefficient to find similarity between two
the most well-known and common one is User-based users [10]. Hadoop is used to calculate the similarity.
Collaborative Filtering. In this algorithm, we predict The output of the Hadoop Map phase i.e. userid and
an item’s rate for a user by collecting information corresponding itemid are passed to reduce phase. In
about this user and similar users. reduce phase, output has been generated and sorted
according to userid. Output again has been stored in
Memory-based algorithms approach the collaborative HDFS.
filtering problem by using the entire database. As
described by Breese et. al , it tries to find users that 4.3 Item Based CF
are similar to the active user (i.e. the users we want to
make predictions for), and uses their preferences to 3.2 Item Based CF Dataset is loaded into HDFS, then
predict ratings for the active user. This page will talk using Mahout we performs Item based CF. Past
about the general ideas; for specific equations and information of the user, i.e. the ratings they gave to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


12
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar Int J S Res CSE & IT. January-February-2019 ; 5(1) : 09-15

items are collected. With the help of this information


the similarities between items are builded and
inserted into item to item matrix. Algorithm selects
items which are most similar to the items rated by the
user in past. In next step, based on top-N
recommendation, target items are selected.

5.1.1 Global average:


V. EXPERIMENTAL ANALYSIS
The global average technique serves as a simple
baseline technique. The average rating for all users
5.1 Baseline methods
across all movies is computed. This global average
serves as a prediction for all the missing entries in the
We try out the following simple baseline methods to
ratings matrix.
give us an idea of the performance to expect from.
The reason behind these inappropriate results is, RSSI
5.1.2 User average:
provides advantages but further research is necessary
All users exhibit varying rating behaviors. Some users
for its sensitive reaction due to environment. RSSI
are lenient in their ratings, whereas some are very
ranging can be inaccurate and inconsistent especially
stringent giving lower ratings to almost all movies.
in indoor environment
This user bias needs to be incorporated into the rating
predictions. We compute the average rating for each
[8]. More over the received signal strength can vary
user. The average rating of the user is then used as the
considerably over small distances and small time
prediction for each missing rating entry for that
scales, due to multipath fading received signals or
particular user. This method can be expected to
path loss can exhibit wide variations even when d
perform slightly better than the global average since it
changes by as little as a few centimeters in the case of
takes into account the rating behavior of the users
802.15.4 radios[8].
into account.

5.1.3 Movie average:


Some movies are rated highly by almost all users
whereas some movies receive poor ratings from
everyone. Another simple baseline which can be
expected to perform slightly better than the global
average is the movie average method. In this
technique, each missing rating entry for a movie j is
assigned the average rating for the movie j.

5.1.4 Adjusted average:


This simple method tries to incorporate some
information about the user i and the movie j when
making a prediction for the entry ri j. We predict a
missing entry for user I and movie j, by assigning it

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


13
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar Int J S Res CSE & IT. January-February-2019 ; 5(1) : 09-15

the global average value adjusted for the user bias and Collaborative filtering is the most successful and
movie bias. popular algorithm in the recommender system’s field.
It helps customers to make a better decision by
The adjusted average rating is given by recommending interesting items. Even though this
ri j = lobalav + (uav(i)ij) + (mav(j)ij) algorithm is the best, it suffers from poor accuracy
and high running time. To solve these problems, this
The user bias is given by the difference between the paper proposed a recommendation approach based on
average user rating and the global average rating. The user clustering by using the Euclidian distance to
movie bias is given by the difference between the calculate two users to cluster dataset. This method
average movie rating and the global average rating. combines clustering and neighbors’ vote to generate
Consider the following example which demonstrates predictions. In the future there may be techniques,
the working of the adjusted average method. fuzzy c-means in the group stages of the first system
to provide a more effective segmentation.
5.2 Dateset:
For experiment, we have used MovieLens dataset of VII. REFERENCES
size 1M. The dataset contains 10000054 ratings and
95580 tags applied to 10681 movies by 71567 users. [1]. A Survey of Collaborative Filtering
There are three files, movies.dat, ratings.dat and Techniques;https://www.hindawi.com/journa
tags.dat. Ratings data file has at least three columns; ls/aai/2009/421425/.,2017, Vol. 13(7).
those are UserId, given by user to movie. [2]. Google News Personalization: Scalable
Online Collaborative Filtering; Das et al;
5.3 Result Analysis: https://www2007.org/papers/paper570.pdf
For movie recommendation, important factor is the [3]. Intro to Recommender Systems:
list of recommended items as soon as possible. Since Collaborative
we are using Hadoop, speedup and efficiency varies as Filtering;http://blog.ethanrosenthal.com/2015
number of nodes varies. To analyze this we have /11/02/intro-to-collaborative -filtering/"
obtained the number of movies which are [4]. Zhan J, Hsieh CL, Wang IC, Hsu TS, Liau CJ,
recommended as threshold changes, Speedup and Wang DW. Privacy-preserving collaborative
efficiency according to number of nodes. recommender systems. IEEE Transactions on
Systems, Man, and Cybernetics, Part C
VI. FUTURE WORK AND CONCLUSION (Applications and Reviews).
[5]. Gong S. A collaborative filtering
Most obvious ideas is to add features to suggest movies recommendation algorithm based on user
with common actors, directors or writers. In addition, clustering and item clustering. Journal of
movies released within the same time period could Software. 2010; 5(7):745-52.
also receive a boost in likelihood for recommendation. [6]. Manvi SS, Nalini N, Bhajantri LB.
Similarly, the movies total gross could be used to Recommender system in ubiquitous
identify a user’s taste in terms of whether he/she commerce. In international conference on
prefers large release blockbusters, or smaller indie electronics computer technology 2011 (pp.
films. However, the above ideas may lead to over 434-8). IEEE.
fitting, given that a user’s taste can be highly varied, [7]. Pu P, Chen L, Hu R. A user-centric
and we only have a guarantee that 20 movies evaluation framework for recommender

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


14
Bheema Shireesha, Navuluri Madhavilatha, Chunduru Anilkumar Int J S Res CSE & IT. January-February-2019 ; 5(1) : 09-15

systems. In proceedings of the fifth ACM


conference on recommender systems 2011
(pp. 157-64). ACM.
[8]. Hu R, Pu P. Acceptance issues of personality-
based recommender systems. In proceedings
of the third ACM conference on
recommender systems 2009 (pp. 221-
4).ACM."
[9]. Pathak B, Garfinkel R, Gopal RD, Venkatesan
R, Yin F. Empirical analysis of the impact of
recommender systems on sales. Journal of
Management Information Systems. 2010;
27(2):159-88. " ,
[10]. Witten IH, Frank E, Hall MA. Data mining:
practical machine leaning toots and
techniques. Morgan Kaufmann Publishers,
Elsevier; 2011."

Cite this article as :

Bheema Shireesha, Navuluri Madhavilatha,


Chunduru Anilkumar, "Movie Recommended System
by Using Collaborative Filtering", International
Journal of Scientific Research in Computer Science,
Engineering and Information Technology
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
09-15, January-February 2019.
Available at
doi : https://doi.org/10.32628/CSEIT19511
Journal URL : http://ijsrcseit.com/CSEIT19511

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


15
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19517

Comparison and an Improved Validation Optimistic Approach


for Concurrency Control
Monika Patel1, Prof. Dr. Dhiren B. Patel2
1MCA Department, S. K. Patel Institute Computer Studies, Gandhinagar, Gujarat, India
2 Department of Computer Science, Gujarat Vidhyapith, Ahmedabad, Gujarat, India

ABSTRACT

Concurrency Control entails the synchronization of accesses to the distributed database, such that the integrity
of the database is maintained. Devising an efficient concurrency control technique is challenging. There is a
need for improvised concurrency control technique to coordinate the simultaneous execution of transactions in
a multi-processing database system. Traditionally, concurrency control techniques are locking, timestamp
ordering and optimistic. These techniques have been evaluated by analytical modeling. In this undertaken work,
the analytical modeling has been investigated and evaluated. The concurrency control in order to improve
optimistic validation method with the traditional techniques, a new technique has been proposed.
Keywords : DBMS, DDBMS, Concurrency, Lock Maintenance, Restart Overhead, Abort Ratio, Deadlock
Possibility

I. INTRODUCTION computer network. The distributed database


management system is defined as the software system
Concurrency Control is a feature that is used to that permits the management of the DDBS to make
coordinate the simultaneous execution of transactions the distribution transparent to the users.
in a multi-processing database system while
preserving data integrity. It is the activity to manage In Concurrency Distributed Transactions, each server
the actions of transactions that operate in parallel manages a set of objects and is responsible for
access, shared data and potentially interfere with one ensuring that they remain consistent when accessed
another. The concurrency control problem is by concurrent transactions. Therefore, each server is
exacerbated in a distributed DBMS (DDBMS) because responsible for applying concurrency control to its
users may access data stored in many different own objects. The members of distributed
computers in a distributed system, and a concurrency transactions' servers are jointly responsible for
control mechanism at one computer cannot ensuring that they are performed in a serially
instantaneously know about interactions at other equivalent manner.
computers.

A distributed database is a collection of multiple,


logically interrelated databases distributed over a

CSEIT19517 | Received : 01 Jan 2019 | Accepted : 10 Jan 2019 | January-February -2019 [ 5 (1) : 16-21 ]
16
Monika Patel, Prof. Dr. Dhiren B. Patel et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 16-21

II. RELETED WORK transaction releases its locks after phase 1 of the
protocol.
Traditionally, concurrency control techniques have
been divided into three categories – locking, A lock manger in different servers set their locks
timestamp ordering and optimistic. Concurrency independently of one another. It is possible that
control techniques have been evaluated by analytical different servers may impose different ordering on
modeling. Kung and Robinson have the first to transactions. Consider the following interleaving of
develop the Optimistic Concurrency Control. transactions T1 and T2 at servers X and Y [8].

a. Problem Analysis and Analytic Modeling c. Timestamp ordering concurrency control


A distributed database means multiple servers or In a single site transaction, the lock manger issues a
multiple site from where data can be read or written. unique timestamp to each transaction when it starts.
When same data accessed and updated by many users, Serial equivalence is enforced by committing the
in the case where the multiple copy of same data can versions of the timestamps of transactions that
exist. In such situation there is a possibility that data accessed them. In distributed transactions, we require
read or written wrong which creates inconsistency in that each manager issue global and unique
the database. timestamps. A global a unique timestamp is issued to
the transaction by the first manager accessed by the
Problems occur when multiple users access the same transaction. The transaction timestamp is passed to
data item. It may happen - data item X can be read by the manager at each site whose data perform an
Transaction T1 and written by Transaction T2 at the operation in the transaction.
same time, creates conflicts in the database. When
same data accessed from many terminals there may When timestamp ordering is used for concurrency
be a chance of concurrency. Concurrency may result control, conflicts are resolved as each operation
in an inconsistent data in the database. To overcome performed. If the decision of a conflict requires the
this problem concurrency can be controlled by transaction to be aborted, the manger will be
concurrency control – means coordination between informed and it will abort the transaction from all
two or more transaction to access the data. the sites.
Commonly known methods of concurrency control
are 2-phase locking, optimistic, and ordering. d. Optimistic Method
The optimistic concurrency control method allows all
b. 2-phase Locking transaction to execute and validate before it will be
In a distributed transaction, the locks on an object are committed. Transaction numbers are assigned at the
held locally. The local lock manager can decide start of the transaction which is executed in the serial
whether to grant a lock or make the requesting manner. A distributed transaction is validated by the
transaction wait. However, it cannot release any several independent sites, each site validate for the
locks until it knows that the transaction has been data used from that site.
committed or aborted at all the servers involved in
the transaction. When locking is used for There are four phases in the transaction execution in
concurrency control, the objects remain locked and the optimistic concurrency control method:
are unavailable for other transactions during the 1. Read: Since reading a value of an object cannot
atomic commit protocol, although an aborted cause a loss of integrity, reads are completely

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


17
Monika Patel, Prof. Dr. Dhiren B. Patel et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 16-21

unrestricted. A transaction reads the values of a set of value updated only when it does not affect the
objects and assigns them to a set of local variables. consistency of the database. If updating of transaction
2. Compute: The transaction calculates a set of values not making database inconsistent, then transaction is
for data objects called the write set. These values are updated to the database and it is recorded in the
assigned to a set of matching local variables. Thus, all successfully completed transaction table, and if
writes after calculation take place on a transaction’s updating of transaction makes database inconsistent
copy of the objects of the database. then updated values removed from the temporary
3. Validate: The transaction’s local read set and write memory and it is recorded in the roll backed
set are validated against a set of committed transaction table. Finally, all the transactions are
transactions. recorded in the transaction log table.
4. Commit and Write: If the transaction succeeds in
validation, it is considered committed in the system In this work, an Improved Optimistic Validation
and is assigned a timestamp, otherwise the Method has been proposed, to improve the optimistic
transaction is rolled back or restarted at either the method. In this method, all the transactions are
compute phase or the read phase. If a transaction executed without restriction and when they need to
succeeds in the validation phase, its write set is made update the database, they have to pass the validation
global and the values of the write set become values test but in the improved method read transaction
of entities in the database at each node. does not go for validation check, so it decreases the
possibilities of restart or abort the transaction.
III. PROPOSED WORK
The major problem with optimistic method was
Model for Concurrent Execution in Distributed when one transaction go for validation, due to some
Databases reason it gets aborted because the resources used by
this transaction is also used by any other transaction
The below Fig 1. Shows the basic model for so there is the possibility of conflict. Once the
concurrent execution in distributed databases. transaction is aborted it gets the same timestamp and
Distributed databases mean there are multiple users starts execution, but it can also happen that the same
who are accessing data from various sites and it may resources used by the other transaction and so that
happen that two or more than two users accessing the transaction again gets aborted.
same data at the same time, at this situation problem
comes. When transaction gets aborted several times and does
not get finished or does not get the chance to execute,
The model shows the basic multiuser environment which means the transaction is in the condition of
with concurrent transactions and it’s after results. starvation. Starvation is the situation where
Assume n numbers of transactions are simultaneously transaction doesn’t get the fair chance to execute.
accessing the same database and executed the
numbers of operations in the one transaction. That
means several updates, read, delete operations are
going to be performed on the single database at time.
As soon as transaction starts its execution database
program identify the table and column in which data
is going to be read from or written into. Transaction

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


18
Monika Patel, Prof. Dr. Dhiren B. Patel et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 16-21

Improved Optimistic Validation Method

In Improved Optimistic Validation Method, it was


observed that, in most applications, the possibilities
of two clients’ transactions accessing the same data
item are low. Transactions are allowed to proceed as
though there were no possibilities of conflict with
other transactions until the client completes its task
and issues a terminatetransaction request. When a
conflict arises, some transaction is generally aborted
and will need to be restarted by the client.

This method was developed as an extension of the


optimistic method proposed by Kung and Robinson
[3]. It was assumed that, the conflicts are very rare in
the database system, and the number of read
transactions are very high than the write transaction.

The Improved Optimistic Validation Method adds


new concept to the previous optimistic method. In
Fig-1 : Model for Concurrent Execution in distributed
this method, there are three phases namely
database
Operational Phase, Validation Phase, Update Phase.
In this method, read transaction does not need to
Comparison of ORDER and Optimistic Method with
perform validation phase, which minimizes the
common criteria
execution time of the transaction. This improved
method has another feature of starvation, which
The below table shows the comparison between the
checks for the aborted transaction, and give it a right
two methods ORDER and Optimistic, in which we
chance to execute.
can analyse that the ORDER method uses locks and
Optimistic method does not, uses locks. But it checks
Validation Algorithm
the validation at time of committing the transaction.
Table shows the basic comparison using the basic
Tn: Transaction
assumptions.
StartT: The highest sequence number
at the start of Tn
Table -1 : Comparison of ORDER and Optimistic
ValidateT: The highest sequence
Method
number at the beginning of its
validation phase
declare valid=true
declare op=read/write
if(op==read)then valid=true
else

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


19
Monika Patel, Prof. Dr. Dhiren B. Patel et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 16-21

for t=StartT+1 to ValidateT do Table -2 : Comparison of ORDER, Optimistic and


if(writeset(t)&readset(Tn)=={})then Improved Optimistic Validation Method
valid=true
else
valid=false
if valid=true then go to update phase

Above algorithm will provide efficient results in


situation where the possibility for read transaction is
more.

IV. RESULTS
The above table shows the comparison between three
The Improved Optimistic Validation Method gives methods with some common criteria. Method 3 is
optimum performance when the system has improved method, developed during this research
minimum conflict ratio and the transactions are read work. If we look into the table, method 3 has been
in most cases. The proposed method does not disturb improved in two parameters – Validation for read
read transaction, increases the execution of transactions and starvation for execution.
transaction. The additional feature has been added of
the starvation. Due to the abort ratio, sometime one V. CONCLUSION
of the transactions is not able to perform and gets
aborted many times, in that case proposed method The Improved Optimistic Validation Method solves
helps the transaction to execute with different above discussed problems in the field of concurrency
conditions. control. The proposed method provides non-locking
algorithm for the transaction execution. It reduces
Comparison of ORDER, Optimistic and Improved the overhead of lock acquisition and lock releases.
Optimistic Validation Method The improved method uses no locks, meets no
deadlock leads to minimum abort ratio.
In this section of chapter, we will compare the
methods which are previously developed and new The assumption was made that, proposed method
one. The comparison criteria have been selected as suits to the environment having minimum conflict.
per the need of the database environment. We will And the most of the transaction has read operation to
compare three methods listed below in the Table 2. perform. The main aim to develop this method is to
improve the efficiency and transaction execution
Method 1: ORDER speed.
Method 2: Optimistic Method
Method 3: Improved Optimistic Validation Method The Improved Optimistic Validation Method
provides the direct access of the data without any
locking. In the method each transaction gets the
timestamp which is the real arrival time of the
transaction. There are three phases of the method:

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


20
Monika Patel, Prof. Dr. Dhiren B. Patel et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 16-21

Operational phase, Validation Phase, and Update [8]. G. Coulouris, J. Dollimore and T. Kindberg T.
phase. Distributed Systems Concepts and Design,
published by Pearson Education Limited 2005.
Limitations and Future Enhancements [9]. Morris R., Wong W., Performance analysis of
locking and optimistic concurrency control
The Improved Optimistic Validation Method is algorithms, May 1985.
efficient to work with the system having minimum [10]. M. Mohamed, M. Badawy, A. El-Sayed., Survey
conflicts, at the same time the proposed method may on Concurrency Control Techniques,
not sufficient in the case where possibility of conflict Communications on Applied Electronics (CAE)
is very high and number of write transactions is high. – ISSN : 2394-4714 Foundation of Computer
A method can be developed, which can control over Science FCS, New York, USA,2016
write transactions using locking concept. A new
method can be derived to enhance proposed method Cite this article as :
the way, read transactions should not be disturbed.
Monika Patel, Dr. Dhiren B Patel, "Comparison and
VI. REFERENCES an Improved Validation Optimistic Approach for
Concurrency Control", International Journal of
[1]. B. Bhargava, Concurrency Control in Database Scientific Research in Computer Science, Engineering
Systems, IEEE Transaction on Knowledge and and Information Technology (IJSRCSEIT), ISSN :
Data Engineering, JANUARY/FEBRUARY 2456-3307, Volume 5 Issue 1, pp. 16-21, January-
1999. February 2019.
[2]. A. Thomasian, Concurrency Control: Methods, Available at doi :
Performance, and Analysis, ACM Computing https://doi.org/10.32628/CSEIT19517
Surveys, March 1998. Journal URL : http://ijsrcseit.com/CSEIT19517
[3]. H. T. Kung, J. T. Robinson, On Optimistic
Methods for Concurrency Control, ACM
Transactions on Database Systems, Vol 6, No. 2,
June 1981.
[4]. P. L. Lehman, AND S. B. Yao, Efficient locking
for concurrent operations on B-trees. Submitted
for publication.
[5]. R. Shrinivasan, Network-Aided Concurrency
Control in Distributed Databases, January 2002.
[6]. S. Ceri, G. Pelagatti, Distributed Databases
Principals and Systems, McGraw-Hill
International Edition, 1985.
[7]. A. Thomasian, Distributed Optimistic
Concurrency Control Methods for High
Performance Transaction Processing, IEEE
Transaction on Knowledge and Data
Engineering, VOL. 10, NO. 1, January/February
1998.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


21
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19514

Airline Data Analysis


Navuluri Madhavilatha1, Bheema Shireesha2, Chunduru Anilkumar3
1Guest Faculty, Department of Computer Science and Engineering, Dr. APJ Abdul Kalam IIIT Ongole, Andhra
Pradesh, India
2, 3 Assistant Professor, Department of Computer Science and Engineering, Dr. APJ Abdul Kalam IIIT Ongole,
Andhra Pradesh, India
ABSTRACT

In the contemporary world, Data analysis is a challenge in the era of varied inters disciplines through there is a
specialization in the respective disciplines. In other words, effective data analytics helps in analyzing the data of
any business system. Flight delays hurt airlines, airports, and passengers. Their prediction is crucial during the
decision-making process for all players of commercial aviation. The goal of our project is to get monthly wise
statistics of airline data and taking particular airport as target we are further analyzing the data to get the
hourly statistics. And also we are finding out the most popular source-destination pairs and calculating the
average delays at every airport. The data for this project comes from the stat-computing.org website. In
particular, in the year 2008 data 70,09,728 titles recorded there which includes information on the Origin,
Destination, Month, Year, DayofWeek, DayofMonth, DepDelay, ArvDelay, DepTime, ArvTime and a few other
less interesting variables. Conveniently, you can export the data directly as a csv file.
Keywords : Machine Learning, Data Mining, Big Data, Statistics, Data Visualization, Data Analytics, ASCII, SRS,
GNU

I. INTRODUCTION Data:
Data is a set of values of qualitative or quantitative
The goal of our project is to get monthly wise variables. It can representation of facts as text,
statistics of airline data and taking particular airport numbers, graphics, images, sound or video. Raw data,
as target we are further analyzing the data to get the also known as primary data, is collected from a source.
hourly statistics. And also we are finding out the If a scientist sets up a computerized thermometer
most popular source-destination pairs and calculating which records the temperature of a chemical mixture
the average delays at every airport [1]. The data for in a test tube every minute, the list of temperature
this project comes from the statcomputing.org readings for every minute, as printed out on a
website in particular, in the year 2008 data 70,09,728 spreadsheet or viewed on a computer screen is "raw
titles recorded there which includes information on data" [2]. Raw data can be input to a computer
the Origin, Destination, Month, Year, DayofWeek, program or used in manual procedures such as
DayofMonth, DepDelay, ArvDelay, DepTime, analyzing statistics from a survey. The term "raw
ArvTime and a few other less interesting variables. data" can refer to the binary data on electronic
Conveniently, you can export the data directly as a storage devices, such as hard disk drives (also referred
csv file. to as "low-level data"). All software is divided into
two general categories: data and programs. Programs

CSEIT19514 | Received : 01 Jan 2019 | Accepted : 10 Jan 2019 | January-February -2019 [ 5 (1) : 22-29 ]
22
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

are collections of instructions for manipulating data. between variables. Once you find these insights, you
Strictly speaking, data is the plural of datum, a single validate the findings by applying the detected
piece of information. In practice, however, people patterns to new subsets of data. The ultimate goal of
use data as both the singular and plural form of the data mining is prediction [5]. Data Science is an
word [3]. The term often used to distinguish binary umbrella that contain many other fields like Machine
machine-readable information from textual human- learning, Data Mining, big Data, statistics, Data
readable information. For example, some applications visualization, data analytics etc.
make a distinction
II. METHODS AND MATERIAL

System Requirement Specification

2.1 Software requirement specifications


It specifies the hardware and software requirements
that are required in order to run the application
properly. The Software Requirement Specification
(SRS) is explained in detail, which includes overview
Between data files (files that contain binary data) and
of this dissertation as well as the functional and non-
text files (files that contain ASCII (American
functional requirement of this dissertation.
Standard Code for Information Interchange) data.
Functional Requirements: R Studio
In database management systems, data files are the
Non-Functional Requirements:Dataset (.csv file)
files that store the database information, whereas
other files, such as index files and data dictionaries,
2.2 Hardware requirements:
store administrative information, known as metadata.
Processor : Any Processor above 500 MHz
The seven V’s sum it up pretty well – Volume,
Ram : 2 GB
Velocity, Variety, Variability, Veracity, Visualization,
Hard Disk : 10 GB
and Value [4]. Science is the intellectual and practical
Input device : Standard Keyboard and Mouse
activity encompassing the systematic study of the
Output device : VGA and High Resolution Monitor.
structure and behavior of the physical and natural
world through observation and experiment Data
III. Technology Description
science is an interdisciplinary field about processes
and systems to extract knowledge or insights from
3.1 Introduction to R
data in various forms, either structured or
unstructured, which is a continuation of some of the
R is a language and environment for statistical
data analysis fields such as statistics, machine
computing and graphics. It is a GNU project which
learning, data mining, and predictive analytics.
is similar to the S language and environment which
was developed at Bell Laboratories (formerly AT&T,
Difference between data mining and data science
now Lucent Technologies) by John Chambers and
colleagues. R can be considered as a different
Data mining refers to the science of collecting all the
implementation of S. There are some important
past data and then searching for patterns in this data.
differences, but much code written for S runs
You look for consistent patterns and / or relationships
unaltered under R.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


23
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

R provides a wide variety of statistical (linear and R is designed around a true computer language, and it
nonlinear modelling, classical statistical tests, time- allows users to add additional functionality by
series analysis, classification, clustering etc) and defining new functions. For computationally-
graphical techniques, and is highly extensible [6]. intensive tasks, C, C++ and FORTRAN code can be
The S language is often the vehicle of choice for linked and called at run time. Advanced users can
research in statistical methodology, and R provides an write C code to manipulate R objects directly. Many
Open Source route to participation in that activity. users think of R as a statistics system. We prefer to
think of it of an environment within which
One of R’s strengths is the ease with which well- statistical techniques are implemented. R can be
designed publication-quality plots can be produced, extended (easily) via packages. There are about eight
including mathematical symbols and formulae packages supplied with the R distribution and many
where needed. Great care has been taken over the more are available through the CRAN family of
defaults for the minor design choices in graphics, but Internet sites covering a very wide range of modern
the user retains full control. statistics. R has its own Latex-like documentation
format, which is used to supply comprehensive
R is available as Free Software under the terms of documentation, both on-line in a number of formats
the Free Software Foundation’s GNU General Public and in hardcopy.
License in source code form. It compiles and runs on
a wide variety of UNIX platforms and similar systems 3.3 R Studio
(including FreeBSD and Linux), Windows and MacOS.
R Studio is an integrated development environment
3.2 The R Environment (IDE) for the R programming language. Some of its
features include: Customizable workbench with all of
R is an integrated suite of software facilities for data the tools required to work with R in one place
manipulation, calculation and graphical display. It (Console, source, plots, workspace, help, history, etc.).
includes an effective data handling and storage Syntax highlighting editor with code completion.
facility, a suite of operators for calculations on arrays,
in particular matrices, a large, coherent, integrated Execute code directly from the source editor (line,
collection of intermediate tools for data analysis, selection, or file). Full support for authoring Sweave
graphical facilities for data analysis and display either and TeX documents.
on-screen or on hardcopy, and a well-developed,
simple and effective programming language which Runs on all major platforms (Windows, Mac, and
includes conditionals, loops, user-defined recursive Linux) and can also be run as a server, enabling
functions and input and output facilities[7]. The term multiple users to access the R Studio IDE using a web
“environment” is intended to characterize it as a fully browser.
planned and coherent system, rather than an
incremental accretion of very specific and inflexible
tools, as is frequently the case with other data
analysis software.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


24
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

IV. System Design graph we can conclude that most number of flights
(627931) are running in July month. And second
most number of flights (616090) are running in
March month.

5.2 Flight statistics of Indy airport


Here we are going to take Indy airport as an Origin
as well as destination calculating monthly wise,
weekly wise, day wise and hourly wise flight
statistics.

5.2.1 Monthly wise statistics of Indy airport

Indy as Origin:-

Fig. 4.1 Analysis of airline data

Data set (.csv file) contains 29 columns as “origin,


destination, month, day of week departure delay,
departure time, arrival time, arrival delay etc”. By
taking data set as an input we generate the graphs
which can be easily understood by the client.

V. Experimental Results

5.1 Monthly wise statistics of entire Airline


data

From the above graph we can conclude that more


number of flights are leaving Indy airport in the
month of July and the less number of flights in the
month of November.

Indy as Destination:
From the below graph we can conclude that more
number of flights are arriving Indy airport in the
month of July and the less number of flights in the
month of November.

The above graph is drawn by taking Months on X-


axis and Number of Flights on Y-axis. From the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


25
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

5.2.3 Day wise statistics of Indy airport

5.2.2 Weekly wise statistics of Indy airport Indy as Origin:-


Indy as Origin:-
From the below graph we can conclude that more
From the below graph we can conclude that more number of flights are leaving the Indy airport in the
number of flights are leaving Indy airport on duration of 6am – 7am and the less number of flights
Wednesday and the less number of flights on in the duration of 12am – 4 am in a day.
Saturday in a week.

Indy as Destination:- Indy as Destination:-


From the below graph we can conclude that more From the below graph we can conclude that more
number of flights are arriving Indy airport on number of flights are arriving the Indy airport in the
Wednesday and the less number of flights on duration of 10am – 11am and the less number of
Saturday in a week. flights in the duration of 2am – 6 am in a day.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


26
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

5.2.4 Hourly wise statistics of Indy airport 5.3 Source to Destination airport pairs
Indy as Origin:- Here we are going to take a look at the most and
From the below graph we can conclude that more least popular airport-pairs based on the flights
number of flights are leaving the Indy airport in the operating in between them (Means one as Origin
duration of 6am – 7am and the less number of flights and another as Destination).
in the duration of 12am –
4 am in a day. 5.3.1 Most Popular

Indy as Destination:- These are the top 6 airport pairs with max number
From the below graph we can conclude that more of flights running between them. In these SFO to
number of flights are arriving the Indy airport in the LAX stands first with 13788 running.
duration of 10am – 11am and the less number of
flights in the duration of 2am. 5.3.2 Least Popular
These are the least 6 airport pairs with minimum
number of flights running between them.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


27
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

5.5.1 Departure delays for all airports

5.4 Average delays of most popular airports


Here we are comparing the average flight delays
among the top 10 most popular airports.

In the above graphs we are representing the


largest average and smallest average departure
Based on the graph we can conclude that OGG delays of the first six (head) airports.
airport running flights with least amount of arrival
and departure delays. And ORD airport running 5.5.2 Arrival delays for all airports
flights with high Amount of departure delay. LGA In the below graphs we are representing the
airport running flights with high Amount of arrival largest average and smallest average arrival delays
delay. of the first six (head) airports.

5.5 Average delays for all airports


Here, we are calculating the average delays (both
arrival and departure) of all the airports.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


28
Navuluri Madhavilatha et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 22-29

airports we come to know the busy hours and days


in which the passenger flow will be high. Those
days we can arrange the alternatives in order to
reduce the problems faced by the passengers as
well as by the authorities like increasing the
security, food and water facilities, transport
facilities outside the airport ,runway clearance etc.

VIII. REFERENCES

[1]. Abdelghany, K. F., Abdelghany, A. F., and Raina S.,


(2004) A model for projecting flight delays during
irregular operation conditions, Journal of Air
Transport Management, Volume 10, Issue 6, Pages
385-394.
[2]. American Statistical Association(stat-computing.org)
[3]. Thearling, K., Data Mining and Analytic
Technologies, www.thearling .com, 2004.
[4]. Tutorials point (http://www.tutorilaspoint.com/r/)
[5]. Aisling, R., and J.B. Kenneth, (1999) An assessment of
the capacity and congestion levels at European
airports, ERSA conference papers ersa 99, pages 241,
European Regional Science Association
[6]. Bureau of Transportation Statistics, Airline On-Time
Statistic. U.S. Department of Transportation.
VI. CONCLUSION Washington, D.C.
http://www.bts.gov/programs/airline information
Not all arrivals can occur when they are scheduled [7]. Hansen, M., and C. Y. Hsiao (2005), Going South? An
to, then airport congestion happens. The delay Econometric Analysis of US Airline Flight Delays
distribution of the airport can make it easier to from 2000 to 2004, Presented at the 84rd Annual
understand the airport delay. We also found the Meeting of the Transportation Research Board (TRB),
Washington D.C., 2005.
most popular origin to destination pairs according
to the number of flights running. We compared
the average arrival and departure delays among Cite this article as :
Navuluri Madhavilatha, Bheema Shireesha, Chunduru
them as well. And also we have calculated the
Anilkumar, "Airline Data Analysis", International
average delays to all the airports.
Journal of Scientific Research in Computer Science,
Engineering and Information Technology (IJSRCSEIT),
VII. FUTURE SCOPE ISSN : 2456-3307, Volume 5 Issue 1, pp. 22-29, January-
February 2019.
Further we can implement this project by Available at doi :
calculating the monthly wise, weekly wise, day https://doi.org/10.32628/CSEIT19514
wise and hourly wise flight statistics at popular Journal URL : http://ijsrcseit.com/CSEIT19514

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


29
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1838108
Li-Fi the Next Generation of Wireless Communication through Visible
Light Communication (VLC) Technology
Imran Siddique, Muhammad Zubair Awan, Muhammad Yousaf Khan, Azhar Mazhar
Department of Computer Science and Information Technology, Afro Asian Institute, Lahore Affiliated with
Government College University Faisalabad, Pakistan

ABSTRACT

Li-Fi stands for Light-Fidelity. This technology is very new and was proposed by the German physicist Harald
Haas in 2011. Light based communication system is the backbone of the future of the communication system.
Li-Fi is a wireless technology that uses light emitting diodes (LEDs) for transmission of data. The development
of the wireless communication leads to advance research in LiFi technology. The term Li-Fi states to visible
light communication (VLC) technology that uses as medium to deliver high-speed communication in a fashion
similar to Wi-Fi. Li-Fi comprises a wide range of frequencies and wavelengths, from the Infrared through
visible and down to the Ultraviolet spectrum. The immense use of Li-Fi may solve some bottleneck of data
transmission in Wi-Fi technology. With the innovation in technology and the number of users, the existing
radio-wave spectrum fails to accommodate this need. To resolve the issues of scalability, availability and
security, we have come up with the concept of transmitting data wirelessly through light using visible light
communication (VLC) technology. This paper objective is to study and describe the LiFi technology. The
improvement of the wireless communication leads to advance research in LiFi technology through Visible
Light Communications (VLC) Technology.
Keywords : Visible Light Communications (VLC), Light Emitting Diodes (LEDs), Radio Frequency (RF),
Infrared (IR), Remotely Operated Vehicles (ROVs), Line of Sight (LOS).

I. INTRODUCTION carrier signal instead of traditional Radio Frequency as


in Wi-Fi. Li-Fi technology communicates with the
With incessantly growing needs of telecommunication help of Visible Light Communication (VLC) spectrum
world, there is an increased thrust for higher and has no side effect as we know the light is very
bandwidth that facilitates faster and secure data much part of our life and so much faster. Moreover Li-
transmission. Present telecommunication industry Fi makes possible to have a wireless Internet in
relies on radio waves of electromagnetic spectrum for specific environments, where Wi-Fi is not allowed
data transmission. Inappropriately, the radio wave due to interferences or security considerations.
spectrum has certain key limitations: Capacity,
Efficiency, Availability and Security. So using the II. BACKGROUND
Light Emitting Diodes that offers in the transmission
of data much faster and reliable than the data that can The history of wireless communication based on
be transmitted through Wi-Fi. Li-Fi is a wireless electromagnetic waves dates back to the Photophone,
communication system in which light is used as a invented by Alexander Graham Bell in the late 19th

CSEIT1838108 | Received : 01 Jan 2019 | Accepted : 10 Jan 2019 | January-February -2019 [ 5 (1) : 30-37 ]
30
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

century. It was the first device used to communicate


without wires and Bell’s invention used the light
produced by the sun to carry the information [6].
There are around 1.4 million cellular staff radio
waves base stations installed, with over 5 billion
mobile phones.

transmission speed. Li-Fi is the new technology has


Mobile phones transmit over 600TB of data. With the
used to label the fast and cheap wireless
advancement in technology and the number of users,
communication system, which is the next optical
the existing radio wave spectrum fails to satisfy to
version of Wi-Fi [4].With the innovation in
this need [1].A new generation of wireless
technology and the number of users, the existing
communications is considered as the fourth stage of
radio-wave spectrum fails to accommodate this need.
the revolution mobile trade. This enables
To resolve the issues of scalability, availability and
connectivity between different objects and machines
security, we have come up with the concept of
all over the world [2].Currently internet have
transmitting data wirelessly through light using
become a major demand; people are in search of Wi-
visible light communication (VLC) technology.
Fi hot spots. This is the advanced scenario in today
communication system which enables the need of
B. VISIBLE LIGHT COMMUNICATION (VLC)
LEDs, Light Emitting
SYSTEMS

Diodes that provides in the transmission of data


Visible Light Communication (VLC) has enlarged
much faster and reliable than the data that can be
great interest in the last decade due to the quick
transmitted through Wi Fi [3].
developments in Light Emitting Diodes (LEDs)
fabrication. Efficiency, durability and long life span
III. METHODS AND MATERIAL
of LEDs make them auspicious residential lighting
equipment as well as alternative cheap and fast data
A. ISSUES IN WIRELESS COMMUNICATION
transfer equipment [5]. One of the ideas laid forward
for wireless optical communication is the visible light
Transfer of data from one place to another is one of
communication method. The signals in the 380-780
the most important daily activities. When the
nm wavelength interval of the electromagnetic
multiple devices are connected to the existing
spectrum are the light signals that can be detected by
wireless networks that connect us to the internet are
the human eye. It is possible to attain illumination
very slow. As the number of devices increasing the
and data transfer simultaneously by means of LEDs
internet access, the fixed bandwidth which is
that is the prominent lighting equipment lately. By
available makes it more and more difficult to utilize
this way, both internal lighting of a room and data
high data transfer rates and connect to a network.
transfer will be attained without the need of an
But, radio waves are just a small part of the spectrum
additional communication system. This technology is
available for data transfer. A solution to this problem
given the name of Visual Light Communication [6].
is by the use of the proposed system LiFi. Li-Fi is
transmission of data through light by sending data
Fundamental entries in a VLC system are the
through an LED bulb that varies in intensity faster
transmitter (LEDs), receivers (photo detectors),
than the human eye can follow and the faster data

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


31
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

modulation of data to optics and the optical technological communication at ultra-high speed
communication channel as shown in Fig -3: greater that of current wireless systems One of the
goals of researchers is to allow 100 megabits of data
transference per second (Mbps) in offices and homes
by modulation of light from promoted lighting
systems. If it is developed correctly, the possibility
exists that many of the problems associated with
present day infrared, radio wave and microwave
communications systems could be at least partially
Fig -1: Visible Light Communication (VLC) Systems resolved, and a more biologically friendly system
[6]
made existing to industries and the general public.

An auxiliary advantage is that VLC systems can


transmit data more securely over short distances than
radiofrequency/microwave communications devices
whose signals can be easily detected outside the
rooms and buildings they invent in [7].

Fig -2: Basic VLC configuration [5]


Fig -4 : Layered architecture of VLC

D. WHY USE VISIBLE LIGHT COMMUNICATION


(VLC)

Fig -3: Block diagram of a VLC architecture [5] The Gamma rays cannot be used as they could
be dangerous [10].
C. LED (LIGHT EMITTING DIODE) VLC X-rays has related health issues.
TECHNOLOGY Ultraviolet light is good for place without
people, but otherwise dangerous for the human
LED (Light Emitting Diode) Visible Light body.
Communications (VLC) systems are familiar as Infrared, due to eye safety regulation, can only
creating a possible valuable addition to future be with low power [10].
generations of technology, which have the potential Radio waves penetrate through the walls so they
to utilize light for the purposes of advanced arises security issues.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


32
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

Hence we left with the only the visible- “Light ✓ A good silicon photodiode, as a receiving
Spectrum”. element.

By using LiFi, it is possible to get the high speed of


100Mbps and more with the help of various
multiplexing techniques. The LiFi emitter system
consists of 4 main subassemblies:
Fig -5: The Electromagnetic Spectrum o Bulb
o RF power amplifier circuit (PA)
E. WHAT IS LI-FI? o Printed Circuit Board (PCB)

Li-Fi is a Visible Light Communications (VLC) system. H. WORKING STYLE OF LIFI


Like Wi-Fi it’s a bidirectional wireless communication
technology. But, where Wi-Fi uses radio waves to It is applied by using a light bulb at the downlink
transmit data, Li-Fi uses visible light from LED light transmitter. Normally the light bulb glows at a
bulbs fitted with a special chip. It’s a little like Wi-Fi, constant current supply however fast and elusive
but has reached speeds 100 times faster in testing, variations in current can be made to produce the
making it far more suited to the demands of the future optical outputs since it just uses the light, hence can
of data [8]. Li-Fi and Wi-Fi are quite similar as both be easily applied to any such area where radio
transmit data electromagnetically. frequency communication is often challenging [11].
The source of light continuously appears because the
However, Wi-Fi uses radio waves while Li-Fi runs on LED can be turned ON and OFF quickly and the
visible light. As we now know, Li-Fi is a Visible Light amount is rapidly modulated that human eye cannot
Communications (VLC) system. This means that it fastening even though it is flickering. The LEDs
accommodates a photo-detector to receive light activity of ON-OFF which invisible and enables the
signals and a signal processing element to convert the transmission of data using the binary codes i.e. 0s and
data into 'stream able' content [9]. 1s, for switch ON LED it is a logical ‘1’, and for
switch OFF LED it is a logical ‘0’. This method of
using rapid pulses of light to transmit information
F. LI-FI IS A WIRELESS TECHNOLOGY? wirelessly is technically referred to as visible light
communication (VLC) [12].
Li-Fi is a wireless technology as it uses visible light
communication or infrared or even near ultraviolet
spectrum. It does not use the radio frequency like the
Wi-Fi technology we use today [10].

G. LIFI SYSTEM CONSTRUCTION [11]

LiFi uses pulses of light to transmit data over wireless


networks. The main mechanisms of LiFi systems are:
Fig -6: Innovative Architecture for Future Li-Fi
✓ High brightness white LED, which acts as the Wireless Technology
transmission source.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


33
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

❖ TRANSMITTER Li-Fi makes use of a free band that does not


prerequisite any licensing.[15]
LED is used as the transmitter or sender of signal Li-Fi Can be used securely in aircrafts
here. The wavering in the LED is basically the signal without affecting airline signals unlike Wi-Fi.
that is to be sent. The quick ON (transmits 1) and The issues of the scarcity of radio frequency
OFF (transmits 0) of the LED is used to encode a bandwidth can be sorted out by Li-Fi.[15]
string of data signal. Now this technique can be Li-Fi also known as Green information
upgraded by using multiple LEDs or LEDs of different technology as it does not effect on the birds,
colours like red, green, blue for more complex data human body’s etc. corresponding to the radio
communications. Tentatively, the speed of data waves or any other communication waves.
communication through Li-Fi can be up to 10Gbps Li-Fi does not have any side effect on any
[13]. living thing. [17]

B. DISADVANTAGES OF LI-FI

High Power Required.


Low Range.
Fig -7: Transceiver LiFi based on VLC Reflection of light takes place. [14]
Cannot penetrate into walls.
Reliability and Network coverage are also
❖ RECEIVER major issues.
Any interference from the exterior light
Since Li-Fi uses visible light to transmit data, it is sources like normal bulbs, opaque materials
important to convert them into digital signals with in the route of transmission will cause
the help of modulation. It includes silicon photo disturbance in communication.[16]
diode which shows good response to visible
wavelength, optical concentrator and filter and an C. APPLICATIONS AREAS OF LI-FI
amplifier. The optical concentrator is used to
compensate for high spatial attenuation due to the A. Educational Systems
beam divergence from the LEDs to illuminate large
area [13]. Li-Fi is the latest technology that can provide fastest
speed for Internet access. So, it can be a good
IV. RESULTS AND DISCUSSION alternative of Wi-Fi at educational institutions and at
companies so that the people there can make use of
A. ADVANTAGES OF LI-FI Li-Fi with the high speed.
High speed.
High Data Transfer Rate. B. Traffic Management
High Connectivity.
Highly Secured and Reliable.[14] In traffic signals Li-Fi can be used to communicate
Low Cost. with passing vehicles (through the LED lights of the
Use Light as career. Vehicles) which can help in managing the traffic in a
Very high operating frequency. better manner resulting into smooth flow of traffic

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


34
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

and controlling the traffic struck and increasing Difference between LIFI and WIFI
accident ratios in developing countries. The LED car
lights can alert drivers when other vehicles are too Parameters Li-Fi Wi-Fi
close via LiFi technology. Transmitter LED Antenna
Frequency band 1000 times 2.4 GHz
C. Better Replacement for Other Technologies
of THz
Standard IEEE 802.15.xx IEEE 802.11xx
Li-Fi uses Visible Light Communication Technology No of users Under The Depend on
(VLC) Instead of Radio Waves. So, it can be easily All over Lamp.(LEDs) access Point.
Data Bits Radio waves
installed at those places where Bluetooth, Infrared,
Transmission
Wi-Fi can’t work properly.
Topology Point to Point Point to
Multipoint
D. Underwater Applications Communicat Based on Visible Based on Radio
ion Light Frequency
Li-Fi can even work underwater where Wi-Fi fails Communication Communication
completely, if the cables of the Remotely Operated (VLC).
Vehicles (ROVs) which operate at a long distance Availability Anywhere, Limited
from the surface are replaced with light (LEDs), available in
which is water-logged then it is easier to access the Airplanes and
internet [16]. Underwater
Power Less More
Consumptio
E. Exclusive Medical Applications
n
Environment Low Medium
Wi-Fi cannot be installed at Hospitals/Operation Impact
Theatres/Medical Laborites as these radio waves can Cost Low High
interfere with the Cellphones/Tablets PC which can Bandwidth Unlimited Limited
block the signals for monitoring equipments thus
proving hazardous to the patient’s health. Usage of G. Industrial Automation
Li-Fi provides more advanced monitoring techniques
and operation strategies. Anywhere in industrial areas data has to be
transmitted, Li-Fi is capable of interchanging slip
F. Cheaper Internet in Aircrafts rings, sliding contacts and short cables, such as
Industrial Ethernet. Due to the real time capability of
The passengers travelling in aircrafts have to Li-Fi, it is also an alternative to common industrial
prescribes to put their electronic devices in the Flight Wireless standards.
Mode or No Internet Mode, because in an aircraft as
these signals may interfere with the navigation H. Safety Environments
signals. So, Wi-Fi is practically impossible in this
scenario.Li-Fi can easily provide high speed Internet In explosion risk environments, the use of electrical
via every light source such as overhead reading LEDs equipment, including mobile phones, is generally
present inside the airplane. greatly restricted. The use of Li-Fi to pass data will
simplify the configuration of data networks in such

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


35
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

environments, and can permit new systems to [3] R.Mallika@Pandeeswari; “Survey of (LI-FI) Light
enhance security in these environments [19]. Fidelity” ,2nd Year-IT, Department of Information
Technology, Mepco Schlenk Engineering College
V. CONCLUSION [International Conference on Emerging Trends in
Applications of Computing (ICETAC 2K17)
Li-Fi is the trend of today and near future. It is one of http://www.ijetjournal.org ISSN: 2395-1303]
[4] Mr. Korde S. K; “Review Paper on Li-Fi (Light
the cheapest and efficient modes of data transfer. It
Fidelity)”
surely can replace the traditional Wi-Fi networks and
[Vol-2 Issue-3 2016 IJARIIE-ISSN (O)-2395-4396]
can further be extended to different platforms to
[5] Taner Cevik and Serdar Yilmaz; “An Overview Of
make it easily accessible and portable which will thus
Visible Light Communication Systems”
cater to the growing demands of the increasing
[International Journal of Computer Networks &
population. One of limitation however is that it only
Communications (IJCNC) Vol.7, No.6, November
works in direct line of sight (LOS) but this would
2015]
solve the problems like the shortage of radio- [6] Adam Böcker, Viktor Eklind Daniel Hansson, Philip
frequency bandwidth and allow internet (WI-FI) to Holgersson, Jakob Nolkrantz & Albin Severinson;
be used in places and situations where it cannot be “An implementation of a visible light
used currently like the hospitals or aircrafts. There communication system based on LEDs”
are billions of light bulbs worldwide, they just need [Department of Signals and Systems CHALMERS
to be replaced with LEDs ones that transmit data. UNIVERSITY OF TECHNOLOGY Gothenburg,
Visible Light Communication (VLC) is very cheaper Sweden 2015]
to Wi-Fi because it uses light rather than radio- [7] Roma Jain, Pallavi Kale, Vidya kandekar, Pratiksha
frequency signals. Kadam; “Wireless Data Communication Using Li-Fi
Technology” [DOI 10.4010/2016.846 ISSN 2321
VI. ACKNOWLEDGEMENTS 3361 © 2016 IJESC]
[8] Jon Mundy and Sacha Kavanagh; “What is Li-Fi?”
The authors would like to thank “Afro Asian Institute, [https://5g.co.uk]
Lahore” Affiliated with Government College [9] Mr. Prashant T Rayrikar; “Wi-Fi Technology Vs Li-
University Faisalabad (GCUF) for the support Fi Technology” [IRJET e-ISSN: 2395 -0056 Volume:
03 Issue: 05 | May-2016 p-ISSN: 2395-0072]
provided to Carry the research work presented in this
[10] Ekta, Ranjeet Kaur; “Light Fidelity (LI-FI)-A
paper.
Comprehensive Study” [International Journal of
VII.REFERENCES
Computer Science and Mobile Computing, Vol.3
Issue.4, April- 2014]
[11] Abhishek Gupta Swayanshu Shanti Pragnya; “The
[1] Roma Jain, Pallavi Kale, Vidya kandekar, Pratiksha
Future Technology LiFi” [IJSRD - International
Kadam; “Wireless Data Communication Using Li-Fi
Journal for Scientific Research & Development| Vol.
Technology” [ISSN 2321 3361 © 2016 IJESC]
4, Issue 08, 2016, ISSN (online): 2321-0613]
[2] Imran Siddique & Muhammad Zubair Awan; “An
[12] Dinesh Khandal, Sakshi Jain; “Li-Fi (Light Fidelity):
innovation in Mobile Communication Technology”
The Future Technology in Wireless
[International Journal of Research e-ISSN: 2348-
Communication” [International Journal of
6848, p-ISSN: 2348-795X Volume 05 Issue 23,
Information & Computation Technology. ISSN
December 2018]
0974-2239 Volume 4, Number 16 (2014), pp. ©

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


36
Imran Siddique et al. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 30-37

International Research Publications House (Electronics & Communication) Engineering from


http://www. irphouse.com] UCEST, Lahore, Pakistan.
[13] Anwesha Chakraborty, Trina Dutta, Sushmita
Mondal, Dr. Asoke Nath; “Latest advancement in Muhammad Zubair Awan
Light Fidelity (Li-Fi) Technology” [International Muhammad Zubair Awan received
Journal of Advance Research in Computer Science his Master Degree (MCS) in
and Management Studies ISSN: 2321-7782 (Online) Computer Science from University
e-ISJN: A4372-3114]
of Lahore, Pakistan and pursuing his
[14] Harsh Deep Ahlawat, Gautham Menon, Girish
MS in Computer Science from
Chhabra; “Advanced Light Fidelity (Li-Fi) System”
COMSATS University, Lahore, Pakistan
[International Research Journal of Engineering and
Technology (IRJET) e-ISSN: 2395 -0056 Volume: 03
Muhammad Yousaf Khan
Issue: 04 | Apr-2016 p-ISSN: 2395-0072]
Muhammad Yousaf Khan
[15] Sowbhagya M P, P Vikas Krishna, Darshan S, Nikhil
A R; “Evolution of Gi-Fi and Li-Fi in Wireless
received his B.Sc. degree in
Networks” [International Journal of Computer Computer System Engineering
Sciences and Engineering ,Volume-4, Special Issue- from Bahauddin Zakariya
3, May 2016 ISSN: 2347-2693] University Multan, Pakistan and
[16] Nikhita Adidam; “Li-Fi - The Future of Internet” completed his M.Sc. in Computer System
[International Journal of Computer Science & Engineering from Islamia University Bahawalpur,
Engineering Technology (IJCSET) ISSN : 2229-3345 Pakistan.
Vol. 6 No. 01 Jan 2015]
[17] Ashmita Shetty; “A Comparative Study and Azhar Mazhar
Analysis on Li-Fi and Wi-Fi” [International Journal Azhar Mazhar pursuing his B.Sc.
of Computer Applications (0975 – 8887) Volume Degree in Electrical Engineering
150 – No.6, September 2016] (2nd year) from Khwaja Fareed
[18] E Ramadhani , G P Mahardika; “ The Technology of University of Engineering &
LiFi: A Brief Introduction” Department of Information Technology (KFUEIT),
Informatics, Universitas Islam Indonesia, Rahim Yar Khan, Pakistan
Yogyakarta, Indonesia [IOP Conference Series:
Materials Science and Engineering]
Cite this article as :
[19] www.lifi-centre.com

About Authors:

Imran Siddique
Imran Siddique received his B.Sc.
degree in Computer Engineering
from Bahauddin Zakariya
University Multan, Pakistan and
pursuing his MS in Electrical

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


37
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19515

Demand Supply Oriented Taxi Suggestion System for


Vehicular Social Networks with Fuel Charging Mechanism
Selvi C1, Keerthana D2
1
Computer Science and Engineering, Velalar College of Engineering and Technology, Erode, Tamil Nadu, India
2
ME-CSE , Velalar College of Engineering and Technology, Erode, Tamil Nadu, India
ABSTRACT

Data mining depends on large-scale taxi traces is an important research concepts. A vital direction for analyzing
taxi GPS dataset is to suggest cruising areas for taxi drivers. The project first investigates the real-time demand-
supply level for taxis, and then makes an adaptive tradeoff between the utilities of drivers and passengers for
different hotspots. This project constructs a recommendation system by jointly considering the profits of both
drivers and passengers. At last, the qualified candidates are suggested to drivers based on analysis. The project
also provides a real-time charging station recommendation system for EV taxis via large-scale GPS data mining.
By combining each EV taxi‟s historical recharging actions and real-time GPS trajectories, the present
operational state of each taxi is predicted. Based on this information, for an EV taxi requesting a
recommendation, recommend a charging station that leads to the minimal total time before its recharging starts.
Keywords : Vehicular Social Networks, Hotspot location, Trajectory data mining, Supply-demand level,
Electronic Vehicle.

I. INTRODUCTION messaging. Social networking sites enable users to


share ideas, pictures, posts, activities, events and
A social networking service (SNS) may be a platform interests with folks in their network. The web
to create social networks or social relations among community services area unit generally thought of a
people that share similar interests, activities, social network service, although in an exceedingly
backgrounds or real-life connections. A social broader sense, social network service sometimes
network service consists of an illustration of every suggests that associate degree individual-centered
user usually a profile, his or her social links, and a service whereas on-line community services area unit
spread of extra services. Social network sites area unit group-centered.
web-based services that enable people to make a
public profile produce a listing of users with whom to The social media sites not solely remained a platform
share connections, and think about and cross the to initiate informal dialogues and a supporter of live
connections among the system. Social network sites messages, however became associate integral a part of
area unit varied and that they incorporate new info promoting ways of the many a business homes. Social
and communication tools like mobile property, photo, Networking shortly became how for whole
video sharing. The Most social network services area promoting and promotion on social sphere, whereby,
unit web-based and supply suggests that for users to the enterprises started victimization these on-line
act over the net, like e-mail and instant electronic communities or websites for developing contacts and

CSEIT19515 | Received : 01 Jan 2019 | Accepted : 10 Jan 2019 | January-February -2019 [ 5 (1) : 38-44 ]
38
Selvi C, Keerthana D Int J S Res CSE & IT. January-February-2019 ; 5(1) : 38-44

driving traffic to their several websites. The appliance However, various vary of novel applications, coming
of those sites has unfold to business homes that together transport networks, exploiting quality
started victimization the Social Networking sites as a pattern, socially aware recommendation systems on
platform to market their services and build whole the roads area unit a number of the factors towards
awareness. These social networking websites kind the whom the analysis community has shown concrete
most tool of social media promoting. The foremost interest. Finally, they bestowed some open analysis
ordinarily used websites Twitter and Face book. The issue for future direction.
main objectives of the Taxi Recommendation square
measure B. Vehicular Social Networks: Enabling Smart
Mobility
 To specialize in a way to maximize drivers‟ profits In this article stressed the importance of high-
whereas high the profit of passengers. potency and reliable transmissions in VSNs for
 To value 2 completely different levels of Demand sensible cities. Significantly, we tend to study a case
offer that square measure appropriate for busy on traffic anomaly detection for VSNs by mechanical
(peak) days and traditional operating days. phenomenon knowledge analysis. They believed that
 To provides a time period charging station VSNs can draw in depth attentions and analysis
recommendation system for taxis. efforts within the close to future because the
 To calculate waiting time beside the gap for the integrations of data technology and social network
recharging stations. services become additional compacted. Though VSNs
II. RELATED WORKS will be considered the mixing of social networks and
IoVs to enhance the standard of life for voters, the
A. Vehicular Social Networks: A Survey avenues of VSN studies aren't flat, and plenty of open
problems are still ahead.
Considered social networking in an exceedingly
transport environment; the authors investigated the C. Energy-Latency Trade-Off For Energy-Aware
possible applications of VSNs and communication Offloading In Mobile Edge Computing Networks
design. VSNs enjoy the social behaviors and quality
of nodes to develop novel recommendation systems In this paper, single and multi-cell MEC network
and route coming up with. Further, they gave an eventualities are thought-about at an equivalent time..
outline of various recommendation systems and path In terms of the mixed number nonlinear drawback
coming up with protocols supported crowd sourcing (MINLP) for computation offloading and resource
and cloud-computing with future analysis directions. allocation, we tend to propose AN reiterative search
They bestowed a progressive literature review on rule combining interior penalty operate with D.C.
socially-aware applications of VSNs, information (the distinction of 2 convex functions/sets)
dissemination, and quality modeling. Further, they programming (IPDC) to search out the optimum
mentioned the various communication protocols resolution. The residual energy of good devices‟
style and information dissemination techniques to battery is introduced into the definition of the
deal with the present gap between VSNs and ancient coefficient issue of energy consumption and latency.
ad-hoc networks that is that the terribly 1st issue to Numerical results show that the projected rule will
be thought-about by the analysis community to get lower total price (i.e., the weighted add of energy
understand the construct of VSNs in public accepted. consumption and execution latency) examination
with the baseline algorithms and also the energy-

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


39
Selvi C, Keerthana D Int J S Res CSE & IT. January-February-2019 ; 5(1) : 38-44

aware coefficient issue is of nice significance to keep a tradeoff between the driver‟s and passengers‟
up the life of good mobile devices. utilities.

III. METHODOLOGY 1) Demand Hotspots Scanning By Clustering


A. Existing Work
By clustering the pick-up points, information from
This work constructs an adaptive recommendation taxi trajectory can be extracted to identify candidate
system based on the supply-demand level, by which a demand hotspots. Traditional DBSCAN algorithm is a
tradeoff is made between the utilities of drivers and kind of density-based clustering methods, which can
passengers. Then the hotspot with the highest score is discover arbitrary clusters and deal with noise or
recommended to available taxis. It considers a outliers effectively First, the distance distribution
passenger‟s utility with the waiting time for vacant matrix is calculated, denoted by Dist nxn.
taxis, which is predicted by mining the pick-up
events.
Input: The pick-up points dataset to be clustered P
Output: The final set of clusters C
1: for pi,pj in P do
2: Dist[i][j] ← getManhattandis(pi,pj);
3: end for
4: Sort Dist in an ascending order line by line;
5: for the i-th column vector in Dist do
6: get average value as epsi;
7: end for
8: DBSCAN (epsi , fixed MinPts) ;
9: Select optimal Eps by the number of cluster and noise;
10: N ← 0;
11: for p in P do
12: N + getEpsNeighbourNum(p);
13: end for
14: Perform DBSCAN with optimal Eps and MinPts;
16: return clustering results C.

where n is the number of pick-up points we extract,


and dist(i,j) is the Manhattan distance between GPS
Fig 1 : Taxi Recommendation System point pi and pj. When the value of i increases, the
First pick-up points for each time segment from the number of clusters and noise both decrease.
taxi trajectory are extracted. Then an adaptive
2) Passenger‟s Waiting Time Prediction
Density-based Spatial Clustering of Applications with
Noise algorithm (I-DBSCAN) for clustering is utilized.
The arrival times of passenger for a particular vehicle
Passengers‟ expected waiting time is predicted based
and actual vehicle arrival time is taken. Then the
on the information of different hotspots. For the
average values of waiting times are calculated and
online part, we retrieve hotspots within certain limits
thus the passenger waiting time is predicted. The
for the correct time segment according to the time
following algorithm is used to predict the waiting
and location of available taxis. The essential
time. With the input of pick up events time stamp
knowledge of each hotspot is calculated for online
sequences, the waiting time is calculated.
recommendation. After evaluating the real-time
demand-supply level of the whole area, we can make

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


40
Selvi C, Keerthana D Int J S Res CSE & IT. January-February-2019 ; 5(1) : 38-44

Input: The pick-up events timestamp sequence Input: Available taxi’s current time curtime and location curloc,
Tp = {pe1,pe2,··· ,pen} candidate hotspots set H
Output: The estimated waiting time w for the hotspot Output: The recommended hotspot
1: λ ← n−1 pen−pe1; 1: MaxScore ← 0,MaxId ← 0;
2: Initiaize the passenger arrival events timetamp sequence 2: U ← ø,w ← ø;
Ta = {ae1,ae2,··· ,aen}; 3: Tracing trajectory and computing the driver’s recent spent
3: for i = 1 to n do time on each hotspots ST = {st1,st2,··· ,stn};
4: if i = 1 then 4: for hi in H do
5: aei = random(0,pei); 5: d ← getManhattandis(curloc,hi.core);
6: // Uniform distribution 6: V ← (hi.revenue−βhi.searchingtime−γd) ;
7: else 7: ε ← sti ∑k stk ;
8: aei = aei−1 + random(0,pei −aei−1); 8: U ∪(V + ε);
9: // Truncated exponential distribution with λ 9: w∪hi.waitingtime;
10: end if 10: end for
11: end for 11: Evaluate real-time demand-supply level α based on curtime
12: w ← 0; using previousAlgorithm
13: for i = 1 to n do 12: for U∗ i , w∗ i corresponding to each hotspot do
14: w + (pei −aei); 14: score ← (1−α)U∗ i + αw∗ i ;
15: end for expected waiting time w. 15: if score > MaxScore then
16: MaxScore ← score;
17: MaxId ← i;
3) Demand-Supply Level Evaluation 18: end if
19: end for
The following algorithm is used for demand supply 20: return recommended hotspot hMaxId.

level evaluation. Total time intervals among the


B. Proposed Work
trajectories and total free/busy counts are calculated
and α value is found out.
1) State Inference Algorithm
Input: Record of trajectory points for the taxi R =
{r1,r2,··· ,rn}
Output: The real-time demand-supply level α
1: S ← ø;
This module is used to update the state inference of
2: for each R do the particular vehicle. It includes vehicle id, entry
3: if r.location in area and r.state was FREE then
4: get r.timestamp as ta; time, time interval, x and y position. This module
5: while FREE IN THIS AREA do
6: get next record; also updates vehicle state inference details every
7: end while
8: get r.timestamp as tb; particular second. These details are stored in „State
9: get r.state as m;
10: ∆t ← (tb −ta + 1−m);
Inference‟ table and viewed by using data grid view
11: S ∪(∆t,m); control.
12: end if
13: end for
14: sum1 ← 0,sum2 ← 0;
15: for (∆ti,mi) in S do Find State Inference (Algorithm)
16: sum1 + ∆ti;
17: sum2 + mi;
18: end for
19: α ← sum2/(sum1 + sum2);
This module is used to find the state inference of the
20: return real-time demand-supply level α. particular vehicle by using state inference algorithm.
When the vehicle id is selected it will automatically
4) Adaptive Recommendation
display the time interval. This algorithm has
threshold value based on this value state inference of
The following Algorithm is carried out in which
a particular vehicle such as recharging, operating or
Input is Available taxi‟s current time curtime and
to station can be calculated. These details are updated
location curloc, candidate hotspots set H and Output
every particular second.
is the recommended hotspot. Real-time demand-
supply level α is taken from previous algorithm. Max
Score is found out based on revenue in various pick
up points. Hotspot with max score is recommended.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


41
Selvi C, Keerthana D Int J S Res CSE & IT. January-February-2019 ; 5(1) : 38-44

Input: an EV taxi v, time t, time interval Ik correspond to


rechargeable time and entry time. It also includes
t, v‟s state distribution {Pk(st),Pk(sr),Pk(so)} in Ik, travel vehicle state details such as recharging vehicle count,
distance d, v‟s real-time trajectory trj = p1,p2,...,pm waiting vehicle count and arriving vehicle count. By
Output: predict v‟s state at t. using these details waiting time will be calculated.
1: if Occupied field of pm is 1 then
2: v‟s state at t is operating;
3: else 3) Vechile Optimization
4: if ∃sp ⊂ trj and pm = ˆpm then Particle Swarm Optimization technique is used to
5: v‟s state at t is recharging; achieve the result. Instead of arithmetic operators, set
6: else {∃sp ⊂ trj and pm = ˆpm}
theory based approach is used here. The velocity and
7: v‟s state at t is operating;
8: else
position update rules in S-PSO also have the same
9: if Pk(st)=max{Pk(st),Pk(sr),Pk(so)} and d<τd then format as that in original PSO algorithms but, the
10: v‟s state at t is operating; particle‟s position in S-PSO is represented by a hard
11: else {Pk(st)=max{Pk(st),Pk(sr),Pk(so)} and d>τd} division, i.e., X ⊆ S, instead of the numerical digits in
12: v‟s state at t is to-station;
unique PSO algorithms, and the velocity is
13: else {Pk(so) = max{Pk(st),Pk(sr),Pk(so)}}
14: v‟s state at t is operating; represented by a set with possibilities. And this
15: else {Pk(sr) = max{Pk(st),Pk(sr),Pk(so)}} module are not in existing system , so we used it as
16: v‟s state at t is recharging; proposed technique. The algorithm is a very powerful
17: end if
optimization technique.
18: end if
19: end if 1.Initialize particles
2.Repeat for each particle do
3.Calculate the fitness value if fitness value is better than the best
fitness value (pbest ) in hist
4.Set current value as the new pbest
2) Waiting Time for Fuel Charging Calculation 5.end if
6.end for
Input: A charging station s, the number of charging piles Ns, 7.Choose the particle with the best fitness value of all the particles
current time t, the number of EVs being recharging and waiting at a for each particle do
s is l and n, respectively, the number of EVs arriving at s earlier 8.Calculate particle velocity according to velocity update equation
than v0 is k; 9.Update particle position according to position update equation
Output: Wv 0 (s), i.e., v0’s waiting time at s;. 10.end for
1: if n = 0 then 11.until maximum iterations or minimum error criteria is attained
2: if l + k < Ns then
3: W v 0 (s) = 0;
4: else {l + k > Ns}
5: M = l + k
6: when an EV taxi finishes recharging, M = M − 1 4) Node Failure Scenario
7: denote the time when Ns − M = 1 as t1;
8: W v 0 (s) = t1 − t;
9: end if Delivering the goods in road failure as well as vehicle
10: else {n > 0}
11: when an EV taxi finishes recharging, n = n − 1 failure scenario is also carried out. During the vehicle
12: denote the time when n = 0 as t2;
13: M = l + k routing if any of the node (vertex) in the graph is
14: after t2, when an EV taxi finishes recharging, M =M − 1
15: denote the time when Ns − M = 1 as t3; found to be failure, then alternate path is also found
16: W v 0 (s) = t3 − t;
17: end if
out from one hop before that failure node (vertex).
Node failure scenario is taken in to account. Any
vehicle if found to be failed in intermediate stage,
This module is used to calculate the waiting time of a then the solution works from the remaining available
particular station. It contains information such as vehicles.
station id, number of outlet when the station id is
selected it will automatically display the number of The solution that N-PSO finds for the optimization
outlets, minimum rechargeable time, maximum problem is thus the best position found by the swarm,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


42
Selvi C, Keerthana D Int J S Res CSE & IT. January-February-2019 ; 5(1) : 38-44

represented as a set of elements. In every iteration, The finding seat available between two location and
each element is updated by following two "greatest" average of seat counting Time details as followed:
values. The first one is the greatest solution (fitness)
AVERAGE TIME CALCULATION PSO-
it has achieved so far. (The fitness value is and stored.) NPSO
This value is call pBest. Another" greatest" value that
100
is tracked by the particle swarm optimizer is the
80
greatest value, obtained so far by any element in the

AVG Time
60 PSO
population. This greatest value is a overall best and 40 NPSO
called gBest. 20
0
v[ ] = v[ ] + c1 * rand() * (pbest[ ] - present[ ]) + c2 * rand() * 1 2 3 4 5
(gbest[ ] - present[ ]) + (( c3 * rand()) + a[t])) + (( c4 * rand( )) - Vehicle
s[t])) - (a)present[ ] = present[ ] + v [ ] - (b)
where, a[t]=v[ ]+ pbest+ gbest
a[t]=v[ ]- pbest- gbest
Fig 2. Average Time Calculation PSO- N-PSO
v[ ] is the element velocity, present[ ] is the current element
(solution),
rand ( ) is a random number between (0,1); c1, c2 are learning Example:
factors usually c1 = c2 = 2. Number of Seat Count, C
Number of Distances, D
IV. ARCHITECURE Total Number of Vehicle TV
Total Number of distances TD
N number of Iteration N

AVG= [(C+D) * (TV+TD)/ N] * 1000

VI. CONCLUSION

The project constructs an adaptive recommendation


system by jointly considering the benefits of drivers
and passengers. First, a spatio-temporal clustering
method named I-DBSCAN is leveraged to group pick-
up locations into different clusters. Second, to
improve the drivers‟ utility, kinds of metrics
including expected revenue, driving distance,
searching time and preference are taken into
consideration. Recent advances in designing to find
the shortest path for minimum traveling costs and
V. EXPERIMENTAL RESULTS
number of vehicles without violating the constraints
and loading capacity of vehicle with Time Windows.
Experimental results illustrate the effectiveness and
To optimize the objective of the work, SBPSO
efficiency of the proposed algorithm, for it provides
algorithms have been found to have a global search
better results than the existing results. Experimental
ability, fast convergence speed, strong robustness, etc.,
result for comparison of PSO and N-PSO algorithm.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


43
Selvi C, Keerthana D Int J S Res CSE & IT. January-February-2019 ; 5(1) : 38-44

VI. REFERENCES Transportation Systems, vol. 17, no. 11, pp.


3318–3324, 2016.
[1] Z. Ning, F. Xia, N. Ullah, X. Kong, and X. Hu,
[9] H. Zhou, P. Wang, and H. Li, “Research on
“Vehicular social networks: Enabling smart
adaptive parameters determination in DBSCAN
mobility,” IEEE Communications Magazine, vol.
algorithm,” Journal of Xian University of
55, no. 5, pp. 16–55, 2017.
Technology, vol. 9, no. 7, pp. 1967–1973, 2012.
[2] A. Rahim, X. Kong, F. Xia, Z. Ning, N. Ullah, J.
[10] G. Qi, G. Pan, S. Li, Z. Wu, D. Zhang, L. Sun,
Wang, and S. K. Das, “Vehicular social
and L. T. Yang, “How long a passenger waits for
networks: A survey,” Pervasive & Mobile
a vacant taxi – large-scale taxi trace mining for
Computing, DOI: 10.1016/j.pmcj.2017.12.004,
smart cities,” in Proc. IEEE Green Computing
vol. 43, 2017
and Communications, pp. 1029–1036, 2013.
[3] W. Hou, Z. Ning, L. Guo, and X. Zhang,
[11] D. Shao, W. Wu, S. Xiang, and Y. Lu,
“Temporal, functional and spatial big data
“Estimating taxi demand-supply level using taxi
computing framework for large-scale smart
trajectory data stream,” in Proc. IEEE ICDMW,
grid,” IEEE Transactions on Emerging Topics in
pp. 407– 413, 2015.
Computing, DOI: 10.1109/TETC.2017.2681113,
[12] Z. Ning, F. Xia, X. Hu, Z. Chen, and M. S.
2017.
Obaidat, “Social-oriented adaptive transmission
[4] J. Zhang, X. Hu, Z. Ning, C. H. Ngai, L. Zhou, J.
in opportunistic Internet of smartphones,” IEEE
Wei, J. Cheng, and B. Hu, “Energy-latency
Transactions on Industrial Informatics, vol. 13,
trade-off for energy-aware offloading in mobile
no. 2, pp. 810–820, 2017.
edge computing networks,” IEEE Internet of
[13] X. Wang, Z. Ning, and L. Wang, “Offloading in
Things Journal, DOI:
Internet of vehicles: A fog-enabled real-time
10.1109/JIOT.2017.2786343, 2017
traffic management system,” IEEE Transactions
[5] X. Kong, F. Xia, Z. Ning, A. Rahim, Y. Cai, Z.
on Industrial Informatics, DOI:
Gao, and J. Ma, “Mobility dataset generation for
10.1109/TII.2018.2816590, 2018.
vehicular social networks based on floating car
[14] K. Zhang, Z. Feng, S. Chen, K. Huang, and G.
data,” IEEE Transactions on Vehicular
Wang, “A framework for passengers demand
Technology, vol. 67, no. 5, pp. 3874–3886, 2018
prediction and recommendation,” in Proc. IEEE
[6] Z. Ning, X. Wang, and J. Huang, “Vehicular fog
SCC, pp. 340–347, 2016.
computing: Enabling real-time traffic
management for smart cities,” IEEE Wireless
Cite this article as :
Communications, 2018
Selvi C, Keerthana D, "System for Vehicular Social
[7] X. Kong, X. Song, F. Xia, H. Guo, J. Wang, and
Networks with Fuel Charging Mechanism",
A. Tolba, “LoTAD: long-term traffic anomaly
International Journal of Scientific Research in
detection based on crowdsourced bus trajectory
Computer Science, Engineering and Information
data,” World Wide Web, vol. 21, no. 3, pp.
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
825–847, 2017.
5 Issue 1, pp. 38-44, January-February 2019.
[8] J. Tang, H. Jiang, Z. Li, M. Li, F. Liu, and Y.
Available at doi :
Wang, “A two-layer model for taxi customer
https://doi.org/10.32628/CSEIT19515
searching behaviors using GPS trajectory data,”
Journal URL : http://ijsrcseit.com/CSEIT19515
IEEE Transactions on Intelligent

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


44
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT183812

Aspect Ranking Technique for Efficient Opinion Mining using


Sentiment Analysis : Review
Prof. Sonali D. Borase#1, Prof. Prasad P. Mahale*2
Assistant Professor, Department of Computer Engineering, NMIMS Mukesh patel school of technology,
#

Shirpur, Maharashtra, India


* Assistant Professor, Department of Computer Engineering, SES’s R.C Patel institute of technology, Shirpur,
Maharashtra, India

ABSTRACT

Opinion mining, also called sentiment analysis, is the field of study that analyses people’s opinions, sentiments,
evaluations, appraisals, attitudes, and emotions towards entities such as products, services, organizations,
individuals, issues, events, topics, and their attributes. Even though facts still play a very important role when
information is sought on a topic, opinions have become increasingly important as well. Opinions expressed in
blogs and social networks are playing an important role influencing everything from the products people buy to
the presidential candidate they support. Thus, there is a need for a new type of search engine which will not
only retrieve facts, but will also enable the retrieval of opinions. Such a search engine can be used in a number
of diverse applications like product reviews to aggregating opinions on a political candidate or issue. This paper
consist review works have been designed for opinion mining by using classification and ranking techniques.
Keywords : Sentiment Analysis, Opinion Mining, POS, Ranking Algorithm, Feature Selection Method,
Semantic Orientation.

I. INTRODUCTION

Sentiment analysis, also called opinion mining, is the reach opinions/sentiments regarding a particular
process of extracting subjective information, such as topic, product, event or individual. Sentiment
opinions, sentiments and attitudes in the source analysis process can be modelled as a classification
materials towards an entity. It is an interdisciplinary problem. Sentiment analysis can be conducted at
research field that combines tools and techniques different levels of detail. Based on the levels,
from natural language processing, text mining and sentiment analysis can be broadly divided into three
computational linguistics [1]. Opinions are important main levels: document level, sentence-level and
factors and influencers of the decision-making aspect-level sentiment analysis [3]. Document-level
process. The determination of people’s opinions sentiment classification aims to determine the overall
toward a particular event can be extremely important sentiment orientation of an entire document, such as
in several fields, such as management sciences, a review text, assuming that each document contains
political science, economics and other disciplines of information regarding a single entity. Sentence-level
social sciences [2]. The Web provides a rich and sentiment analysis aims to identify subjective and
progressively expanding source of information to objective sentences. In sentence-level sentiment

CSEIT183812 | Received : 01 Jan 2019 | Accepted : 10 Jan 2019 | January-February -2019 [ 5 (1) : 45-49 ]
45
Prof. Sonali D. Borase, Prof. Prasad P. Mahale,. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 45-49

analysis, the sentiment orientation of subjective accordingly. Each file form target domain is assigned
sentences is also identified.[6] The classification of score on which basis it is classified as positive or
review documents at document or sentence level of negative. Each word has weight assigned to it.
granularity does not fully reveal the opinions Summation of weights of words in each sentence
regarding different features of a particular entity. gives score to sentence. Then addition of all sentence
Hence, aspect-level sentiment analysis concerns the score is nothing but score of file. On the basis of this,
classification of sentiments by focusing on the file is classified. Clustering helps in reducing
particular features/aspects of entities. mismatch between domain specific words of source
and target domains. Two sets of lexicons are extracted
II. LITERATURE SURVEY as an output with polarity which is compared with
SentiWordNet..[1]
Blenty proposed algorithm works on word level. POS
tagged words are extracted from pre-processing steps
are used. Each word acts as feature. Algorithm is Pre-processing of dataset
executed for total number of features provided in
input dataset. According to this weight words are
classified into two categories. Classified words are
having POS tag, polarity tag, and weight value. From Apply classifier on dataset.
this list, common and uncommon words are picked
and used for bipartite graph clustering.

Probability distribution of class c is calculated based Applying clustering on dataset.


on term frequency. Classification process predicts the
polarity of the target domains lexicon from source Fig 1 : major phases in entropy based classifier
domain. The clustering algorithm is applied on
classified word lists and documents until it reaches Nyaung proposed system in which review crawler
convergence. All extracted words from source crawl the related web pages and retrieves review
domain are tagged, and weight is calculated for each comments only. The filtered review comments will
word using mutual information available for words. be proceeding for more processing steps. Subjective
Target words are extracted and compared with the sentences express the reviewer’s sentiment about the
source. If they match then they will be categorized as product and objective sentences do not have any
domain independent, otherwise domain-specific. direct or support of that sentiment.
Domain-independent words are from both source Subjectivity/objectivity determination works in two
and target domains; whereas, domain-specific are phases – training and classification. For training
from target domain only. A graph is constructed phase, manually labelled sentences are used as trained
between domain dependent and domain-independent data, which is later used to identify subjective
words. Co-occurrence relationship between these unigrams for new dataset. In the second phase, the
words represents edge. Occurrence of domain- classification is centred on the probability of
specific word along with domain-independent word unigrams from test dataset using the training data. In
means that both a related to each other and assign the pre-processing, only subjective sentences are
edge. Using domain-independent words weight is submitted to a pipeline for Parts-Of-Speech (POS)
assigned to domain-specific words and classified tags. POS tagging is used for sentence splitting and to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


46
Prof. Sonali D. Borase, Prof. Prasad P. Mahale,. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 45-49

assign lexical categories to the words in text. The feature candidates are identified by POS tags and
feature and opinion pairs, the parser from previous only the features which are stored in the domain
step are analysed and generate all possible ontology are valid. Maximum Entropy modelling for
information components from them. The extracted prediction of feature and opinion pairs in a different
pairs from dependency relation can contain which way it has been used so far, using weights for both to
are relevant or not. Therefore, maximum entropy emphasize the importance of each one of them in the
model is used to predict opinion-relevant product relation of feature and opinion pairing task.
feature or not by choosing the class with the highest Maximum entropy model is used to predict which
conditional probability p according to this model. feature word should be related with the opinion
Domain ontology to get the domain related features word with maximum probability.[3]
and to define the synonym set for features. Product

Table 1: Summery

Year Author Method Limitation

2018 Blenty et Modified Naïve Byes Algorithm It is not support


al unstructured keyword.
2018 Deshmukh Extracts and classifies opinion words from one domain called It is not consider non-
et al source domain and predicts opinion words of another domain word feature
called target domain using a semi-supervised approach,
Which combines modified maximum entropy and bipartite graph
clustering.
2017 Aytug et The presented framework obtains individual feature rankings with
al information gain, chi-square, gain ratio, symmetrical uncertainty,
Pearson correlation coefficient, ReliefF algorithm and
probabilistic significance measure-based feature selection
methods. Then, these individual lists are amalgamated into a
single ranking list via a genetic algorithm.
2015 Dim En Polarity of extracted opinion word calculated with Comparative sentence
Nyaung et sentiword net. not included
al
2014 Zhang et al Probabilistic aspect ranking algorithm. It used only document
level sentiment
analysis.

Deshmukh et al proposed method collected data from identifying polarity of each aspect n finally the
site that provides with reviews on electronic gadgets product itself. Modify Naivè Bayes model to create
and gives the respective user reviews. Pre-processing an aspect ranking algorithm which eventually will
phase on data through Natural Language processing is rank the products where this step calculates the
done and then feature/aspect list is displayed. weight of features from reviews. [2]
Extracting nouns as features and adjectives as
opinions. Next step is Enhanced Aspect Identification Zhang et al proposed system in which given the
and Opinion Word Extraction In this technique High consumer reviews of a particular product, we first
frequency aspects are ranked and the corresponding identify aspects in the reviews by a shallow
opinion words for the aspects are extracted for dependency parser and then analyse consumer

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


47
Prof. Sonali D. Borase, Prof. Prasad P. Mahale,. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 45-49

opinions on these aspects via a sentiment classifier. measures like Information gain-based feature ranking
System develop a probabilistic aspect ranking is a filter-based feature selection method, which is
algorithm, which effectively exploits the aspect widely utilized in text mining domain. The second is
frequency as well as the influence of consumers’ Chi-squared feature ranking evaluates the merit of
opinions given to each aspect over their overall each feature individually with the chi-squared
opinions on the product in a unified probabilistic statistical measure. Next is Information gain is a
model. The overall opinion in a review is generated biased measure towards features with high values.
based on a weighted aggregation of the opinions on Similar to gain ratio, the symmetrical uncertainty
specific aspects, where the weights essentially coefficient aims to eliminate the bias of information
measure the degree of importance of these aspects. A gain towards the attributes with higher values.
probabilistic regression algorithm is developed to Pearson correlation coefficient is used to measure the
infer the importance weights by incorporating aspect correlation between two attributes. Relief algorithm
frequency and the associations between the overall is a filter-based method that uses a feature relevance
opinion and the opinions on specific aspects. In order criterion to rank the features. Proposed genetic rank
to evaluate the proposed product aspect ranking aggregation-based feature selection model uses Filter-
framework. A product aspect ranking framework to based feature selection methods can obtain different
automatically identify the important aspects of rankings for the same data set. The aggregation of
products from numerous consumer reviews. A several different feature rankings may be beneficial
probabilistic aspect ranking algorithm to infer the to obtain an enhanced ranking of the features. The
importance of various aspects by simultaneously proposed method models the feature selection as a
exploiting aspect frequency and the influence of rank aggregation problem and the feature rankings
consumers’ opinions given to each aspect over their obtained by different filter-based methods are
overall opinions on the product.[4]. combined. Given a set of rankings of the same
candidate sets, the rank aggregation seeks to find a
Aytug at al evaluated Feature selection is the process single better ranking from these multiple rankings.
of obtaining an appropriate feature subset from the The rank aggregation problem can be modelled as an
data set so that the classification optimization problem where the objective is to
obtain a final ranking that is the closest to the all
Algorithms can deal efficiently with high- individual ranking lists. [5]
dimensional feature spaces. Feature selection
methods aim to eliminate irrelevant or redundant III. CONCLUSION
features and to reduce the training time required to
build a classification model Feature selection method This paper give brief information on Opinion Mining
can be broadly divided into two groups: filter-based and the recent literature review putting light on the
and Wrapper-based feature selection methods. steps involved in Opinion Mining. Opinion Mining is
the field of study that tracks the mood or attitude of
Filter-based methods evaluate the usefulness of people towards a particular product or topic. It helps
features based on heuristics/evaluation metrics, to find the product with best quality, which is
whereas wrapper-based methods select the features predicted based on the user reviews. It is also used for
based on the performance of a machine learning recommendation system and business intelligence.
algorithm to optimize the predictive performance. There are many challenges and future developments
System evaluated the different individual filter-based possible in Opinion Mining approach like short

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


48
Prof. Sonali D. Borase, Prof. Prasad P. Mahale,. Int J S Res CSE & IT. January-February-2019 ; 5(1) : 45-49

length and irregular structure of the content such as Cite this article as :
named entity recognition, parsing. Opinion mining is
to find the opinion of a person from sentences and Prof. Sonali D. Borase, Prof. Prasad P. Mahale,
then classify them on the basis of polarity. Opinion "Aspect Ranking Technique for Efficient Opinion
mining helps people to know the semantic Mining using Sentiment Analysis : Review ",
orientation for a product by classifying it as positive, International Journal of Scientific Research in
negative or neutral. Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
IV. REFERENCES 5 Issue 1, pp. 45-49, January-February 2019.
Available at doi :
[1]. J. Deshmukh and A. Tripathy, "Entropy based https://doi.org/10.32628/CSEIT183812
classifier for cross-domain opinion mining", Journal URL : http://ijsrcseit.com/CSEIT183812
Applied Computing and Informatics, vol. 14,
no. 1, pp. 55-64, 2018.
[2]. B. Alengadan and S. Khan, "Modified
Aspect/Feature Based Opinion Mining for a
Product Ranking System", 2018 IEEE
International Conference on Current Trends in
Advanced Computing (ICCTAC), pp. 1-5, 2018.
[3]. D. Nyaung and T. Lai Thein, "Feature-Based
Summarizing and Ranking from Customer
Reviews", International Journal of Computer
and Information Engineering, vol. 9, no. 3, pp.
734-739, 2015.
[4]. Zheng-Jun Zha, Jianxing Yu, Jinhui Tang,
Meng Wang and Tat-Seng Chua, "Product
Aspect Ranking and Its Applications", IEEE
Transactions on Knowledge and Data
Engineering, vol. 26, no. 5, pp. 1211-1224,
2014.
[5]. A. Onan and S. Korukoglu, "A feature selection
model based on genetic rank aggregation for
text sentiment classification", Journal of
Information Science, vol. 43, no. 1, pp. 25-38,
2016.
[6]. B. Liu, "Sentiment Analysis and Opinion
Mining," Synthesis Lectures on Human
Language Technologies, vol. 5, no. 1, pp. 1-167,
May 2012

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


49
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19516

Big Data Anonymization in Cloud using k-Anonymity Algorithm


using Map Reduce Framework
Anushree Raj1, Rio G L D’Souza2
1Department of M.Sc. Big Data Analytics, St Agnes Autonomous College, Mangalore, Karnataka, India
2Department of Computer Science and Engineering, St Joseph Engineering College, Mangalore, Karnataka,
India

ABSTRACT

Anonymization techniques are enforced to provide privacy protection for the data published on cloud. These
techniques include various algorithms to generalize or suppress the data. Top Down Specification in k
anonymity is the best generalization algorithm for data anonymization. As the data increases on cloud, data
analysis becomes very tedious. Map reduce framework can be adapted to process on these huge amount of Big
Data. We implement generalized method using Map phase and Reduce Phase for data anonymization on cloud
in two different phases of Top Down Specification.
Keywords: Anonymization, Big Data in cloud, k-Anonymity, Map Reduce, Privacy Preserving

I. INTRODUCTION algorithm [8]. Using data anonymization key pieces


of confidential data are obscured in a way that
Security and privacy are a significant obstacle that is maintains data privacy. Data can still be processed to
preventing the extensive adaptation of public cloud gain useful information. Anonymization data can be
[1]. Cloud Computing is a revolutionary computing stored in cloud and even processed without concern
approach, which provides massive storage and that other individuals may capture the data [9]. Later,
computational capability [2]. It provides user to the results can be collected and mapped to the
implement applications cost effectively without original data in a secure area.
investment and infrastructure. It enables users to
have flexible, on demand access to computing In this paper we propose an anonymization model
resources via Internet [3]. These advantages are the using the Top Down Specification techniques on
causes of security and privacy problems, which huge amount of data along with Map reduce
emerge because data owned by different users are framework. The Top Down Specification has two
stored in cloud servers and users are losing their own phases which includes the map phase which reduces
control on data. the huge amount of data into small chunks of data
and the reduce phase combines the intermediate
Data anonymization is the process of hiding identity results to give back the k anonymous data [10].
or sensitive data from original data. Various Anonymization takes place at both the phases.
algorithms are in use for implementation [4], [5], [6], Generalization method is used on Data that usually
[7]. The most efficient algorithm for data contain sensitive information and this proves the
anonymization is the Top Down Specialization importance of employing anonymity approaches.

CSEIT19516 | Received : 01 Jan 2019 | Accepted : 10 Jan 2019 | January-February -2019 [ 5 (1) : 50-56 ]
50
Anushree Raj, Rio G L D'Souza Int J S Res CSE & IT. January-February-2019 ; 5(1) : 50-56

Generalization is the most common method to make Sensitive Attribute: Attribute or data that an
the data anonymous to provide the privacy. In this individual is sensitive about revealing.
paper we discuss on data anonymization technique in
section II and then show the implementation using There are two main ways through which k-
Map Reduce in section III. Anonymity can be achieved:
Bottom Up Generalization: In Bottom-Up
II. DATA ANONYMIZATION Generalization strategy, the data is initialized to its
current state and generalizations are carried out over
Data anonymizaton is a deep study, broadly adopted attribute values until k-Anonymity is not violated
for privacy preserving in data publishing. [16].
Anonymization techniques highly focus on data
security in public cloud [11]. The main feature of Top Down Specialization: In Top-Down
anonymization techniques is hiding few key data to Specialization strategy, all the attribute values are
maintain data privacy. Data can further be processed initialized to the root value of the hierarchy tree [17].
to achieve necessary information by analysing it. The The specialization is carried out iteratively over the
simple way of preserving privacy is by removing all attribute values, until the k-Anonymity is violated.
those information, which will directly link to data of Top Down Specification
an individual. This process s referred to
anonymization [12]. There are many approaches to Let D be a data set to which the anonymization
gain data anoymization, such as k-Anonymity, l- technique is applied. R denotes the records in the
Diversity and t-Closeness which helps to make the data set. Let A be the attributes, which can identify N
data more private and anonymous, to protect data individuals through Ai where 1 ≤ i ≤ N, k is the
access from any third party users. anonymity parameter used in the anonymization
levels. Anonymization levels are the single
k- Anonymity: mapreduce job execution level from multiple
k-Anonymity is used to hide some key information of mapreduce jobs. The domain values are represented
the users in a database of k users. Then the attackers by the taxonomy tree [18],[21]. Q determines the
can make sure there are only k users but cannot quasi identifier. A record can contain any N number
interpret the information of an individual user in of Quasi Identifiers, Qi ϵ R, 1 ≤ i ≤ N. the sensitive
particular [13], [14]. This method is used to values are denoted by S. In Top Down Specification, a
anonymize the quasi-identifiers to provide privacy to data set is anonymized through specialization
the data. The main aim of k-Anonymity is to make operations. The domain values are replaced with all
each record indistinguishable from at least k-1 other its sub domain values.
records. k-Anonymity assigns properties in specific
ways [15]. The k anonymity data attributes are of In Top Down Specialization, a hierarchy tree is used
three types to represent the attribute values which is initiated by
a root value. The specialization performed iteratively
Key Attribute: Attribute which can identify n over the attribute values, till the k - anonymization is
individual directly. satisfied. The specialization is performed by
Quasi - identifier: Attribute which can be linked with substitution of the parent attribute value by its child
external information to identify an individual. value in taxonomy tree. For example, the patient
record of a hospital is anonymized and published for

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


51
Anushree Raj, Rio G L D'Souza Int J S Res CSE & IT. January-February-2019 ; 5(1) : 50-56

data analysis using Top Down Specialization which is


a generalization algorithm then Table I and II
represent the anonymization for the quasi identifiers.
Here Q = {AGE, SEX}

Table 1. Original Data

NAME AGE SEX EDUCATION PROFESSION Figure 3. Taxonomy tree for Sex
SAM 34 M MASTER BUSINESS
FUNG 45 M DOCTRATE DOCTOR Table 2. Anonymized data
SONA 53 F MASTER PRINCIPAL
RAFI 25 M BACHELORS HR NAM SE PROFESSI
ZAMA 37 F MASTERS MANAGER AGE EDUCATION
E X ON
ALICE 28 F BATCHELORS BUSINESS
[1- AN GRADUATE
BOB 52 M DOCTRATE PROFESSOR * BUSINESS
34] Y SCHOOL
[35- AN GRADUATE
The specialization is performed by replacing the * DOCTOR
49] Y SCHOOL
parent attribute value by its child value in Taxonomy
[50- AN GRADUATE PRINCIPA
Tree *
63] Y SCHOOL L
[1- AN
TAXONAMY TREES AND QUASI IDENTIFIERS * BACHELORS HR
34] Y
[35- AN GRADUATE MANAGE
*
49] Y SCHOOL R
[1- AN
* BACHELORS BUSINESS
34] Y
[50- AN GRADUATE PROFESS
*
63] Y SCHOOL OR

The above table describes the anonymization using


generalization algorithms like TDS for the selected
Figure 1. Taxonomy tree for Education Quasi Identifiers with the taxonomy tree hierarchy.
Algorithm 1 TDS

1) Identify every value in HT to the top most value of


the hierarchy tree.
2) Let ListAttributei initially include the top most
value.
3) While some ca(categorical attribute) in
ListAttributei is valid and useful, do step 4, 5, 6
Figure 2. Taxonomy tree for Age 4) Find the Best specialization from ListAttributei
5) Perform Best on HT and update ListAttributei

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


52
Anushree Raj, Rio G L D'Souza Int J S Res CSE & IT. January-February-2019 ; 5(1) : 50-56

6) Update InfoGain(ca) and AnonymityLossValue(ca), denoted by AfterSpecification(Q).


and validity for ca ∈ ListAttributei BeforSpecfication(Q) - AfterSpecification(Q) is the
7) End while loss of anonymity.
8) return Generalized HT and ListAttributei AnonymityLossValue(v) =
AVERAGEloss(BeforSpecfication(Q) -
Initially ListAttributei contains only the topmost AfterSpecification(Q))
value ListAttributei, in which case it returns the
masked table together with ListAttributei. In order to III. MAP REDUCE FRAMEWORK
maintain the data utility we have to find the best
specialization, otherwise the data utility will decrease Map Reduce is a programming model, which
considerably and can leave data useless for mining computes large volume of data sets with parallel and
purpose. distributed algorithm on cluster. A Map Reduce
program is composed of two user defined functions
The metrics used to find the best specialization are, [22]. A Map function which performs sorting of data
the Information Gain (InfoGain) and Anonymity loss and Reduce function that performs summary
value (AnonymityLossValue). For each specialization operation. Map Reduce is a framework for processing
SP:P child (P) is calculated. parallelizable problems across large datasets using
The Information Gain (InfoGain) and Anonymity loss large number of computers.
value (AnonymityLossValue) values are calculated to
generalize the record R ϵ D using the SC(R). A typical MR-MPI program makes at least three calls
to the MapReduce -MPI library, to perform map(),
SC(𝑅) collate(), and reduce() operations on a MapReduce
𝐼𝑛𝑓𝑜𝐺𝑎𝑖𝑛(𝑅)
if 𝐴𝑛𝑜𝑛𝑦𝑚𝑖𝑡𝑦𝐿𝑜𝑠𝑠𝑉𝑎𝑙𝑢𝑒(𝑅) ≠ 0 object it creates. In a map() key-value pairs are
𝐴𝑛𝑜𝑛𝑦𝑚𝑖𝑡𝑦𝐿𝑜𝑠𝑠𝑉𝑎𝑙𝑢𝑒(𝑅)
=
else generated by each processor. The key-value pairs
{ 𝐼𝑛𝑓𝑜𝐺𝑎𝑖𝑛(𝑅)
produced are stored locally by each processor; a map()
………………. (1)
thus requires no inter-processor communication.
1) Information gain,
Users call the library with a count of tasks to perform
IG(R) is only positive values
and a pointer to a user function; the MapReduce -
|Rc| MPI map() operation invokes the user function
IG(R) = 𝐼(𝑅𝑐) − ∑ ( ) I(Rc)
|Rr| multiple times as a callback. Depending on which
c variant of map() is called, the user function may be
………………… (2)
passed a file name, a chunk of bytes from a large file,
I(Rx) is the entropy measure of Rx where Rx is the
a task ID, or a v pair. Options for assigning map tasks
number of records. (IRx, Sl )is the records with
to processors are specified by the user and include
sensitive value and entropy is computed by,
assigning consecutive chunks of tasks to processors,
|Rca, S| |Rca, S| striding the tasks across processors, or using a master-
𝐼(𝑅𝑥) = − ∑ ( ) X log2 ( ) slave model that is useful when tasks have widely
|Rr| |Rr|
c varying workloads [19].
…………………… (3)
2) Anonymity Loss Value (AnonymityLossValue)
The basic data stored and operated on by any
Here the anonymity before specialization is denoted
MapReduce framework are key-value pairs [20]. In
by BeforSpecfication(Q) and after specialization is
the MapReduce -MPI library, individual keys or

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


53
Anushree Raj, Rio G L D'Souza Int J S Res CSE & IT. January-February-2019 ; 5(1) : 50-56

values can be of any data type or length, or gathered by enterprises and government agencies.
combinations of multiple types (one integer, a string Such data is subjected to, be abused by internal and
of characters, two integers and a double, etc); they external adversaries. Many anonymization
are simply treated as byte strings by the library. A techniques came into existence. They are k-
key-value pair always has a key; its value may be Anonymity, l-Diversity, t-Closeness and m-Privacy.
NULL. Key-value pairs are stored within a These privacy preserving data mining algorithms
MapReduce object. A user program may create one or cannot work for big data analytics, as such data is
more MapReduce objects to implement an algorithm computed using MapReduce programming paradigm
[21]. Various MapReduce operations (map, reduce, in cloud computing environment. MapReduce is used
etc) are invoked on an object and its key-value pairs; to control the parallel processing power of cloud
key-value pairs can also be passed between and computing infrastructure. In this paper, we proposed
combined across objects. algorithms which are used to parallelize k-
Anonymity using MapReduce programming
Conceptually the map and reduce functions supplied paradigm with Hadoop. First, we studied k-
by the user have associated types. Anonymity model and its usefulness in anonymizing
map (k1,v1) → list(k2,v2) data. We implemented the Top Down Specialization,
reduce (k2,list(v2)) → list(v2) a generalization method on k-anonymity so as to
anonymize the original data. We then implemented
Map Reduce approach involves the following four algorithms that could parallelize k-Anonymity using
steps: Map Reduce framework. The proposed algorithm is
capable of anonymizing big data in order to support
1. Map processors, assigns the K1 input Key value, privacy preserving. In future, we can propose an
and provides the processor with all the input data algorithm using MapReduce programming towards
associated with that key value. efficiently anonymizing big data for preserving
2. Run the user provided Map() exactly once for privacy.
each key value, generating output organized by
key values K2. V. REFERENCES
3. Reduce processors, assigns the K2 key value with
all the Map generated data associated with that [1] L. Hsiao-Ying and W.G. Tzeng, “A Secure
key value. Erasure Code-Based Cloud Storage System with
4. Run the user provided Reduce () exactly once for Secure Data Forwarding,” IEEE Trans. Parallel
each K2 key value produced by the Map and Distributed Systems, vol. 23, no. 6, pp.
5. Produce the final output –The Map Reduce 9951003, 2012.
system collects all the reduce output and sorts it [2] D. Zissis and D. Lekkas, "Addressing Cloud
by K2 to produce final outcome Computing Security Issues," Future Generation
Computer Systems, vol. 28, no. 3,pp. 583-
IV. CONCLUSION 592,2011.
[3] RuilinLiu, Hui Wang, ”Privacy –Preserving
Privacy is very important to protect the sensitive data Data Publishing “ IEEE ,2010.
from the attacker. To provide data privacy the [4] B.C.M. Fung, K. Wang, and P.S. Yu,
anonymization methods may be used. The data "Anonymizing Classification Data for Privacy
mining activities utilize huge amount of data Preservation," IEEE Trans. Knowledge and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


54
Anushree Raj, Rio G L D'Souza Int J S Res CSE & IT. January-February-2019 ; 5(1) : 50-56

Data Eng., vol. 19, no. 5, pp. 711-725, May Systems, Volume 10 Issue 5, October 2002,
2007. Pages 571 - 588
[5] X. Xiao and Y. Tao, "Anatomy: Simple and [15] Meyerson A and Williams R, On the
Effective Privacy Preservation." Proc. 32nd complexity of optimal k-anonymity. In
Int'l Conf. Very Large Data Bases (VLDB'06), PODS04: Proceedings of the twenty fourth
pp. 139-150,2006. ACM SIGMOD-SIGACT-SIGART symposium
[6] K. LeFevre, D.J. DeWitt, and R. Ramakrishnan, on Principles of database systems, pages 223–
"Incognito: Efficient Full-Domain K- 228, 2004
Anonymity," Proc. ACM SIGMOD Infl Conf. [16] Ke Wang, Philip S. Yu, Sourav Chakraborty,
Management of Data (SIGMOD '05), pp. 49-60, "Bottom-Up Generalization: A Data Mining
2005. Solution to Privacy Protection", Fourth IEEE
[7] K. LeFevre, D. l DeWitt, and R. Ramakrishnan, International Conference on Data Mining,
"Mondrian Multidimensional K-Anonymity," 2004. ICDM '04 .. Pages 249 – 256
Proc. 22nd Int'! Conf. Data Eng. (ICDE '06), [17] Xuyun Z, Laurence T Yang ,”A scalable Two
2006. phase Top Down Specialization Approach for
[8] 1 Xu, W. Wang, 1 Pei, X. Wang, B. Shi, and A. Data Anonymization using Map Reduce on
W. Fu. , "Utility-based anonymization using cloud“, IEEE Transaction on Parallel and
local recoding ", In ACM SIGKDD, 2006. Distributed Systems ,TPDSSI-2012.
[9] S.Yu, ”Anonymizing Classification Data for [18] Zhang X, Yang LT, Liu C, Chen J. A scalable
Privacy Preservation“. IEEE Transactions on two‑phase top‑down specialization approach
Knowledge and Data Engineering ,vol 19 no 5 for data anonymization using MapReduce on
,2007 cloud. IEEE Trans Parallel Distrib Syst.
[10] Benjamin C.M Fung, Ke Wang, Philip S.Yu 2014;25(2):363–73
“Top Down Specialization for Information and [19] Jefrey Dean and Sanjay Ghernawat “Map-
Privacy Preservation”. Reduce: Simplified Data Processing on Large
[11] Bayardo R and Agrawal R, Data privacy Clusters” Google,Inc.2004
through optimal k-anonymization. In ICDE05: [20] J. Dean and S. Ghemawat, "Mapreduce:
The 21st International Conference on Data Simplified Data Processing on Large Clusters,"
Engineering, pages 217– 228, 2005. Comm. ACM, vol. 51, no. 1, pp. 107-113,2008
[12] HIPAA 2012 “k-anonymity : A model for [21] Al‑Zobbi M, Shahrestani S, Ruan C.
Protecting Privacy “ International Journal Sensitivity‑based anonymization of big data. In:
Uncertain Fuzz .vol 10 ,no,5 pp 557-570 ,2002. Local computer networks work‑ shops (LCN
[13] Ciriani, S. De Capitani di Vimercati, S. Foresti, workshops), 2016 IEEE 41st Conference on.
and P. Samarati, "k-Anonymity Data Mining: A IEEE; 2016. p. 58–64
Survey", Springer US, Advances in Information [22] Al‑Zobbi M, Shahrestani S, Ruan C.
Security (2007) Implementing a framework for big data
[14] Latanya Sweeney, School of Computer Science, anonymity and analytics access control. In:
Carnegie Mellon University, Pittsburgh, Trustcom/BigDataSE/ICESS, 2017 IEEE. IEEE;
Pennsylvania, "Achieving k-anonymity privacy 2017. p. 873–80.
protection using generalization and [23] Al‑Zobbi M, Shahrestani S, Ruan C.
suppression", International Journal of Multi‑dimensional sensitivity‑based
Uncertainty, Fuzziness and Knowledge-Based anonymization method for big data. In: Elk‑

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


55
Anushree Raj, Rio G L D'Souza Int J S Res CSE & IT. January-February-2019 ; 5(1) : 50-56

hodr M, Shahrestani S, Hassan Q, editors.


Networks of the future: architectures,
technologies, and implementations. Boca
Raton: Chapman and Hall/CRC Computer and
Information Science Series, Taylor & Francis;
2017. p. 448.

Cite this article as :

Anushree Raj, Rio G L D'Souza, "Big Data


Anonymization in Cloud using k-Anonymity
Algorithm using Map Reduce Framework",
International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 50-56, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT19516
Journal URL : http://ijsrcseit.com/CSEIT19516

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


56
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1| ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19518

Improvement of Privacy and Security in Hybrid Cloud with


Attribute Group Based Access Control
Kayalvili S1, Sowmitha V2
1Computer Science and Engineering, Velalar College of Engineering and Technology,Erode, Tamilnadu, India
2 ME-CSE, Velalar College of Engineering and Technology, Erode, Tamilnadu, India
ABSTRACT

Cloud computing enables users to accumulate their sensitive data into cloud service providers to achieve
scalable services on-demand. Outstanding security requirements arising from this means of data storage and
management include data security and privacy. Attribute-based Encryption (ABE) is an efficient encryption
system with fine-grained access control for encrypting out-sourced data in cloud computing. Since data
outsourcing systems require flexible access control approach Problems arises when sharing confidential
corporate data in cloud computing. User-Identity needs to be managed globally and access policies can be
defined by several authorities. Data is dual encrypted for more security and to maintain De-Centralization in
Multi-Authority environment.
Keywords:ABE-Attribute Based Encryption, Authority, Fine Grained, Outsourcing, Dual Encryption.

I. INTRODUCTION Cloud computing provides the tools and technologies


to create data/compute intensive parallel applications
Cloud computing is internet-based computing during with far morereasonablecosts compared to ancient
whichmassiveteams of remote serverssquare measure parallel computing techniques. In Single-authority
networked to permit sharing of data-processing tasks, ABE, every user’s keys area unit generated
centralized information storage, and on-line access to victimizationtotally different random and on the QT
pc services or resources. Clouds are often classified as shared values specified keys generated for various
public, personal or hybrid. Cloud computing that users can't be combined, that prevents collusion
depends on sharing computing resources instead of attacks.
having native servers or personal devices to handle
applications. The mostsanctioning technology for The main objective of this paper isn't to use a central
cloud computing is virtualization. In basic ABE authority to manage users and keys, and
systems, the data shared is oftenat intervals one solelystraightforward trust relations have to be
domain or organization. However, in reality, compelled to be fashioned by sharing the general
infolikes drivers’ licenses andregistration info in public key between every attribute authority (AA).
universities area unit organized by totally different User identities square measuredistinctive by
government departments. The management of combining a user’s identity with the identity of the
attributes and key distributions can't be undertaken AA wherever the user is found. Once a key request
by an equivalent attribute authority. Moreover, has to be created to associate authority outside the
access methodscould also be distributed supported domain, the request has to be performed by the
attributes of various authorities. authority within the current domain instead of by

CSEIT19518 | Received : 01 Jan 2019 | Accepted : 10Jan2019 | January-February-2019 [ 5 (1) : 57-62 ]


57
Kayalvili S, Sowmitha V.Int J S Res CSE & IT. January-February-2019 ;5(1):57-62

the users, so, user identities staypersonal to the AA B. Persona: An Online Social Networkwith User-
outside the domain, which is able to enhance privacy Defined Privacy
and security. Here, once the user ought to requests
associate attribute secret key, if the attributes square These network help users share in order with their
measuresituated outside the domain, the request by friends. However, users entrust thesocial network
the supply AA within the domain to the target AA is provider with such personal information as sexual
employedinstead of by requests by users themselves. preferences, supporting and religious views, phone
numbers, occupations, identities of friends, and
II. RELATED WORKS photographs. Although sites over privacy controls
that let users restrict how their data is viewed by
A. AData Outsourcing Architecture Combining other users, sites provide inadequate controls to
Cryptography and Access Control restrict data sharing with shared affiants or
application developers.
The recent adoption and diffusion of the information
out-sourcing paradigm, whereverknowledge house C. Fuzzy Identity-Based Encryption
owners store their knowledge on external servers,
there are increasing general demands and issues for In this paper, discussed the use fullness of using
knowledge confidentiality. Besides well-known risks biometric s in Identity-Based and then discuss their
of confidentiality and privacy breaks, threats to out- contributions. Using biometrics in Identity- Based
sourced knowledgeembody improper use of Encryption in many situations, using biometric–based
information: the server may use substantial individuality in an IBE system has a number of
components of a setof knowledge gathered and important advantages over “standard” IBE. In
arranged by the information owner, doubtless standard Identity-Based Encryption scheme a user
harming the information owner's marketplace for with a positive identity, for example, “Bob Smith”,
any professional duct or service that comes with that will need to go to an authority to obtain the private
assortmentof data. The projected novel access key corresponding to the identity. In that process the
management model and design that eliminates the user will need to “prove” to the authority that he is
necessity for a reference monitor and depends on indeed entitled to this identity. That will typically
cryptography to make sure confidentiality of involve presenting supplementary documents or
knowledgehold on a server. Knowledgearea unit credentials. Typically, there exists a trade-off
encrypted because theknowledge owner stores them between a system that is expensive in this step and
on associate in nursing external server. one that is less reliable.
Authorizations and codingarea
unitunitedthereforepermitting access D. Mediated Cipher text-Policy Attribute-Based
managementsocial control to be outsourced besidethe Encryption and Its Application
information. The nice advantage is that the
information owner, whereas specifying the policy, Distributed data systems needversatile access
wants not b econcerned in its social control. management models thattranscend discretionary,
necessary and role-based access management. On the
opposite hand, the present trend of service-based data
systems and storage outsourcing needenhanced
protection of knowledgetogether with access

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


58
Kayalvili S, Sowmitha V.Int J S Res CSE & IT. January-February-2019 ;5(1):57-62

managementways that area unit cryptographically control can be achieved by dual encryption method
implemented. The conception of Attribute-Based which takes advantage of the attribute-based
coding (ABE) fulfills the samenecessities. It provides encryption and discerning group key distribution in
a chicmanner of encrypting knowledgesuch the each attribute group. This work demonstrates how to
encryptor defines the attribute set that the apply the proposed mechanism to securely manage
decryptorhas topossesso as to decrypt the cipher- the outsourced data. The analysis results point to that
text. . A secret key holder willdecipher the cipher the proposed system is efficient and protected in the
text if the attributes related to his secret key satisfy data outsourcing systems.
the access policy related to the cipher text. In these
schemes, a ciphertext is related toassociate in nursing The concept of attribute-based encryption (ABE) is a
access policy and therefore the user secret promising approach that fulfills these requirements.
keysrelated toa collection of attributes. ABE features a device that enables an entrance
control over encrypted data using access policies and
III. METHODOLOGY qualified attributes among private keys and cipher
texts. Especially, cipher text-policy ABE (CP-ABE)
1. EXISTING WORK provides a scalable way of encrypting data such that
the encryption defines the attribute set that the
When the system is set up, the public key of each AA decryptor needs to have classify to decrypt the cipher
and the basic parameters are distributed, which text.
simplifies the process of trust establishment. The key
issuing protocol for privacy only needs to use the 2. PROPOSED WORK
public key of AA to realize the trust between AA.
Present data outsourcing systems require supple  Attribute Group - Based Encryption is managed.
access control approaches. In many cases, it is It adapts a dual encryption approach to overcome
attractive to provide differentiate access services such the user access control problem.
that data access policy are defined over user  Multiple attribute groups are included and data is
attributes or roles. Some of the most difficult issues in distributed among them. User privileges may be
data outsourcing scenario are the enforcement of varying for data maintained by different attribute
union policies and the support of policy updates. groups.
Cipher text-policy attribute-based encryption is a  The data owner maintains all the membership
hopeful cryptographic solution to these issues for lists to enable the direct user management in
enforces access control policies defined by a data more than one place.
owner on outsourced information.  Keys are assigned randomly and independently
from each other. But group key mechanism is
But, the problem of applying the attribute-based common for users in single group
encryption in an outsourced architecture introduces  All the data is maintained by more authority
several challenges with regard to the attribute and groups.
user revocation. So the project also proposes an access
control mechanism using cipher text-policy attribute-
based encryption to enforce right to use control
policies with well-organized element and user
revocation capability. The fine-grained way in

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


59
Kayalvili S, Sowmitha V.Int J S Res CSE & IT. January-February-2019 ;5(1):57-62

command button is user to terminate the current


modules.
d) Attribute Key Generation

This module is used to process the key generation


process and the access structure modules is used to
create the access specification for each and every user
for specifying the details with the rights of select,
insert, update, delete operation those process is
selected by the check box control, attribute identity
number and user identity numbers are selected by
user from the Combo Box control.
Fig 1.1. Privacy and Security Improvement

e) Encryption
a) Trusted Key Pair
Encrypt the text using public key for the purpose of
The trusted key pair created in the application for other users not knows the given message. So, the
further process, following key generate the public unrestricted key is extract using get key grasp button
key and master key, for the purpose of encrypt and displayed in the label control, the message is
the message and group key is created. These entered in the textbox manage then the given
information are generate by form command encrypted communication is display in the label
button event and showed in the multiline text control. The encrypted message saved in the
mode. These key saved in the relevance using application using creates cipher text and save
save authority key event. command button event.

b) Attribute Creation f) Decrypt Cipher Text

It contains details such as attribute id, attribute It is used to retrieve the plain data in the application.
names are entered by user in the textbox controls, The specified cipher text is entered the data is
and saved by the save command button. The showed to the user. In these module user
delete button is used is used to delete the specific characteristics number and nobody texts are selected
record and close command button is user to from the combo box control, group identity is display
terminate the current modules. in the label controls. The significance is decrypted in
the cipher text grid view control using the decrypt
c) User Creation authority button occurrence.

Used to create the client facts for accessing the The following Table describes experimental result for
attribute with privilege level. The client id, user proposed system secure attribute selection analysis.
first name and passwords are enter by user in the The table contains number of time slot interval and
textbox control these details are saved by the save given attribute to calculate average numbers of
command button event. The delete button is used selection attribute details are shown.
is used to delete the specific record and close

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


60
Kayalvili S, Sowmitha V.Int J S Res CSE & IT. January-February-2019 ;5(1):57-62

Member Node Existing System Proposed RATIO OF SELECTION ATTRIBUTE SELECTION


(%) System (%) (AS)- [Proposed System]

8 72.54 78.62 1.2


1
12 76.13 78.11 RATIO OF

Ratio [%]
0.8
SELECTION
0.6
16 82.42 83.13 ATTRIBUTE
0.4 SELECTION (AS)
24 86.66 84.67 0.2
0
30 88.13 89.78 1 3 5 7 9 11
Number of Attribute
32 80.44 82.66
38 78.33 80.21 Fig 1.2 Attribute selection- Ratio Analysis
42 87.22 89.76
46 79.22 80.65 VII. CONCLUSION

 Authorization policies and the support of policy


IV. CIPHER TEXT-POLICY ATTRIBUTE-BASED updates are studied.
ENCRYPTION  Proposes a cryptographic approach for
STEP 1: communication between users and attribute
The setup algorithm is executed which is a authorities.
randomized algorithm that takes no input other than  Allows a data owner to define the access
the implicit security parameter of the public key PK control policy and enforce it on his outsourced
and a master key MK. data.
 Efficient and scalable to securely manage the
STEP 2: outsourced data.
The attribute key generation algorithm is
executed which takes input the master key MK, a set of
VIII. FUTURE WORK
attributes  L, and a set of user indices U u as
parameters. It outputs a set of private attribute keys SK
for each user in U that identifies with the attributes set.  It allows a data possessor to define the access
control guidelines and enforce it on his
STEP 3: outsourced data.
The key encrypting key (KEK) generation  It also features a method that enables more
algorithm is executed in this module, which takes a set fine-grained access control with efficient
of user indices U u as input, and outputs KEKs for attribute and user revocation capability. The
each user in U, which will be used to encrypt attribute
proposed scheme will be efficient and scalable
group keys Ki for each Gi G.
to firmly manage the outsourced data.
.
 The innovative system is designed such that
that enhancement can be integrated with
current modules easily with less integration
work.

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


61
Kayalvili S, Sowmitha V.Int J S Res CSE & IT. January-February-2019 ;5(1):57-62

IX. REFERENCES ACM TOCS , 1(3):239{248, August 1983.


Germany, Septemb er 2007.
[11]. J. Crampton, K. Martin, and P. Wild. On key
[1]. S. Vimercati, S. Foresti, S. Jajodia, S. Paraboschi,
assignment for hierarchical access control. In
and P. Samarati, "A Data Outsourcing
Proc. of the 19th IEEE CSFW'06 , Venice, Italy,
Architecture Combining Cryptography and
July 2006.
Access Control," Proc. ACM Workshop
[12]. G. Miklau and D. Suciu. Controlling access to
Computer Security Architec-ture (CSAW ’07),
published data using cryptography. In Proc. of
Nov. 2007.
the 29th VLDB Conference , Berlin, Germany,
[2]. L. Ibraimi, M. Petkovic, S. Nikova, P. Hartel,
Septemb er 2003.
and W. Jonker, "Mediated Ciphertext-Policy
[13]. H. Hacig Aum Aus, B. Iyer, and S. Mehrotra.
Attribute-Based Encryption and Its
Providing database as a service. In Proc. of 18th
Application," Proc. Int’l Workshop Information
ICDE , San Jose, CA, USA, February 2002.
Security Applications (WISA ’09), pp. 309-323,
2009.
[3]. R. Baden, A. Bender, N. Spring, B.
Cite this article as :
Bhattacharjee, and D. Starin,"Persona: An
Online Social Network with User-Defined
Kayalvili S, Sowmitha V, "Improvement of Privacy
Privacy,"Proc. ACM SIGCOMM ’09, Aug. 2009.
and Security in Hybrid Cloud with Attribute Group
[4]. A. Sahai and B. Waters, "Fuzzy Identity-Based
Based Access Control", International Journal of
Encryption," Proc.Eurocrypt ’05, pp. 457-473,
Scientific Research in Computer Science, Engineering
2005.
and Information Technology (IJSRCSEIT), ISSN :
[5]. V. Goyal, O. Pandey, A. Sahai, and B. Waters,
2456-3307, Volume 5 Issue 1, pp. 57-61, January-
"Attribute-BasedEncryption for Fine-Grained
February 2019.
Access Control of Encrypted Data," Proc. ACM
Available at doi :
Conf. Computer and Comm. Security, pp. 89-
https://doi.org/10.32628/CSEIT19518
98, 2006.
Journal URL :http://ijsrcseit.com/CSEIT19518
[6]. J. Anderson. Computer security planning
study.Technical Rep ort 73-51, Air Force
Electronic System Division, 1972.
[7]. J. Saltzer and M. Schro eder. The protection of
information in computer systems.
Communications of the ACM , 17(7), July 1974.
[8]. N. Provos . Encrypting virtual memory. In
Proc. of the 9th USENIX Security Symposium ,
Denver, Colorado, USA, August 2000.
[9]. A. Harrington and C. Jensen. Cryptographic
access control in a distributed ¯le system. In
Proc. of the 8th SACMAT , Como, Italy, June
2003.
[10]. S. Akl and P. Taylor. Cryptographic solution to
a problem of access control in a hierarchy.

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


62
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195113

Survey on Factors Determining the Share Price Movement in


Indian Stock Market - A Big Data Approach
Nitha K. P.*1, Dr. S. Sivakumari2
1Research Scholar, Department of Computer Science and Engineering, Faculty of Engineering, Avinashilingam
Institute for Home Science and Higher Education for Women Coimbatore, Tamil Nadu, India
2 Professor and Head, Department of Computer Science and Engineering, Faculty of Engineering,
Avinashilingam Institute for Home Science and Higher Education for Women Coimbatore, Tamil Nadu, India

ABSTRACT

The aim of investors of any stock is to earn reasonable returns. This paper presents a framework on share price
movement and investigates the factors determining and predicting the share prices in Indian stock market through
big data approach. This study is also trying to focus on the fundamental factors, technical factors and sentimental
factors that stimulate the share price movement. The methodology used mainly focuses on the implications of the
theory of financial markets and features affecting the share price movement. Stock market is one of the burning
areas where data is growing day by day. Big data Analytics is a very promising area and buzzword for the next
generation information technologies. Because of the heterogeneity and other complexities of data, big data
architecture and design, is needed which specifically deals with the stock market data and analyze the heterogeneous
data for the future prediction of the market. Variety and large Volume of factors exists. Moreover, the velocity with
which data generates and the effect of them in the share price movement emphasizes, role of big data in predicting
the share prices in Indian Stock Market.
Keywords : Stock Market, Big Data, Stock valuation model, Technical, Fundamental and Sentimental Factors

I. INTRODUCTION ratio, dividend yield, and dividend payout. This study


also examines the importance of technical factors and
For an investor, predicting the share price is sentimental factors, which highly influences the
significant and critical. Data required for the stock share price movement of Indian companies.
market prediction is placed and generated at distinct
places of the world. Industry specific factors such as A. FUNDAMENTAL FACTORS
EPS, book value, P/E ratio and dividend yield has 1) Book Value (BV) : It is also known as net asset
influenced the share price movement of Indian value per share because it measures the amount
companies a lot. The present study has been of assets, which the corporation has on behalf of
undertaken with an objective to review the existing each equity share. BV shows the net investment
literature by examining the relationship between per share made in the business by the
stock prices and company specific intrinsic factors or shareholder. It is the value at which an asset is
fundamental factors such as book value per share, carried on a balance sheet.
dividend per share, earning per share, price earnings 2) Earning Per Share (EPS): The ratio of the profit

CSEIT1838116 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 63-69 ]
63
Nitha K. P., Dr. S. Sivakumari Int J S Res CSE & IT. January-February-2019 ; 5(1) : 63-69

after tax of the company for any financial year associated with the matching price chart of the
after payment of preference dividend. The equity security. Technical indicators offer many uses such
shareholders are the sole appellants to the net as confirm the trends, generate buy/ sell signals,
earnings of the corporation after making forecast the trend of future prices. Technical
payment of dividend to the preference analysis is a technique, which gives an idea about
shareholders. The significance of this ratio flows future share prices of selected companies in which
from the fact that higher the earnings per share we invest. Based on the knowledge of technical
the more is the scope for a higher rate of analysis one can predict the perfect investment
dividend and of retained earnings, to build up decision of the stock market. By using the nominal
the internal power of the company. pointers, the future market of securities will know
3) Dividend per share (DPS): Dividend is the in which stock to invest. The more accurate
portion of the profit after tax, which is prediction of stock prices of selected companies the
distributed to the shareholders for their investor to carry out fundamental analysis of stock
investment bearing risk in the company. It has a prices, they can predict of future trend of stock
significant influence on the market price of the prices.
share. DPS shows how much the company has
paid out as dividend. It refers to the actual 1) Moving Averages: Most visual aids shows a lot of
amount of dividend (gross) declared per share. discrepancy in price movement. This can make it
The net profit after taxes belong to shareholders difficult for traders to get an idea of a security’s
but the income that they really receive is the overall trend. One simple method traders use to
amount of earnings distributed and paid as cash combat this is to apply moving averages. A
dividend. moving average is the average price of a security
4) Price Earnings Ratio (P/E ratio): This ratio over a set amount of time. They smooth a data
enables an investor to make appropriate series and made it easier to spot tendencies,
calculation of the time required to cover the something that is especially helpful in volatile
investment in a company’s stock. P/E ratio markets.
expresses the relationship between the market 2) Simple Moving Average (SMA): A simple moving
price of a company’s share and its earnings per average is formed by computing the average
share. It indicates the extent to which the (mean) price of a security over a specified
earnings of each share are covered by its price number of periods. For example: a10 day moving
average(SMA) is taken as the closing prices for
B. TECHNICAL FACTORS the last 10 days added together and dividing the
total by 10.
Technical indicators are mathematical formulas 3) Exponential Moving Average (EMA): Exponential
that, when applied to security prices clearly flash moving average is the moving average that is
either buy or sell indications. Price data includes formed by applying weight to the recent price
any combination of the open, high, low, or close changes.
over a period and most of the indicators use only the
closing prices, technical indicators are usually EM A(c) = ((price(c) EM A (p) ∗ K) + EM A (p))(1)
shown in a graphical form above or below a
security’s price chart for the analysis. Once shown Where c=current, p=previous
in graphical form, an indicator can then be

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


64
Nitha K. P., Dr. S. Sivakumari Int J S Res CSE & IT. January-February-2019 ; 5(1) : 63-69

The formula for the smoothing constant K is: polarity” sentiment classification looks, for
instance, at emotional states such as “angry,”
K = 2/ (1 + N) (2) ”sad,” and “happy.”

Where, N = number of periods for EMA Existing methods applied to extract the sentiment
4) Relative Strength Index (RSI) : The relative analysis, which grouped mainly into four
strength index (RSI) is an extremely useful categories: keyword spotting, lexical affinity,
oscillating momentum indicator that was statistical methods, and concept-level techniques.

developed by J. Welles Wilder and is one of the Keyword spotting categorizes text based on the
presence of unambiguous affect words such as
most widely used indicators in technical analysis.
happy, sad, afraid, and bored. Lexical affinity
RSI oscillates between zero and 100. The most
detects an affect words and assigns a probable
popular is the 14 days RSI where the RSI is
affinity to particular emotions. Statistical
calculated based on 14 days values. Traditionally methods has control on elements from machine
the stock is considered to be over bought when learning such as latent semantic analysis, support
RSI is above 70 and over sold when RSI is below vector machines, “bag of words” and Semantic
30. Signals can also be generated by looking for Orientation Point-wise Mutual Information.
divergences and central line crossovers. The
basic formula is: Concept-level approaches leverage on elements
from knowl- edge representation such as
RSI = 100[100/(1 + RS)] (3) ontologies and semantic networks and, hence, are
also able to detect semantics that are expressed in
Where, RS = average of upward price change over a a subtle manner, e.g., through the analysis of
select number of days / average of downward price concepts that do not explicitly convey relevant
change over the same number of days information, but which are implicitly linked to
other concepts that do so.The main data sources
C. SENTIMENTAL FACTORS for the sentimental analysis are taken from
Twitter as they provide an API to extract the
Sentiment analysis extracts subjective tweet data from the site. The following features
information in source materials with the help of capture useful characteristic of Twitter and
natural language processing, text analysis and writers for opinion retrieval.
computational linguistics. Sentiment analysis
aims to determine the attitude of a speaker or a II. METHODS FOR STOCK MOVEMENT
writer with respect to some topic or the overall PREDICTION
contextual polarity of a document. The
elementary job in sentiment analysis is to The Support Vector Machine (SVM) been
categorize the feelings of a given text in the recognized as being able to efficiently handle
document, sentence, and the feature/aspect level high dimensional data and has shown to perform
of the expressed opinion in a document, or a well on many classification and clustering
sentence or an entity feature/aspect is positive, techniques
negative, or neutral. Advanced, “beyond 1) Price only: In this method, only past data or
prices predicts the stock movement. The

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


65
Nitha K. P., Dr. S. Sivakumari Int J S Res CSE & IT. January-February-2019 ; 5(1) : 63-69

purpose of this method is to investigate topics. Latent Dirichlet Allocation (LDA)


whether there are patterns in the history of (Blei, Ng, & Jordan, 2003) is a generative
the stock or not. In addition, this model was probabilistic model of a corpus. The basic idea
used as a standard to estimate whether is that documents are represented as random
combination of the sentiments is effective for mixtures over latent topics, where each topic
the approach by comparing with other is characterized by a distribution over words.
sentiment models. Features used for the Therefore, the LDA chosen as a simple topic
training of SVM are price t1 and price t2 model to discover these hidden topics.
which are the price movements (up, down) at 5) JST-based method: The method refers to
the transaction dates t 1 and t 2, opinion often expressed on a topic or aspect.
respectively. When people post the message on the social
2) Human sentiment: In addition to historical media to express their opinion for a given
prices, this model integrates the sentiments stock, they tend to talk their opinions for a
annotated by human into the prediction certain topic such as profit and dividend.
model. According to the Message Board Based on pairs of topic-sentiment, they would
dataset, the users explicitly attach a sentiment think that the future price of that stock
label with their posts. The sentiment tags like would go up or down. From that intuition, a
strong buy, buy, hold, sell and strong sell are new feature topic-sentiment for the stock
usually seen labels attached with the posts. prediction model results.
Instead of using all the messages, only the 6) Aspect-based sentiment: As an alternative of
messages with annotated sentiments by the considering, the blends of hidden topics and
users need to be considered, and all the other sentiments as in the previous model, in this
irrelevant messages can be discarded. From model the combinations not hidden. Each
these messages, only the explicit sentiment message represented as a list of topics and
will be taken and remove other information their corresponding sentiment values are
such as message content. The perseverance of considered. The topic selected as the
this method is that how attitude interpreted consecutive nouns in the sentence will
by human can be used to predict the stock. provide tremendous results in prediction.
Because human marks the sentiments, this
feature is one of the strongest features for III. IMPORTANCE OF BIG DATA IN SHARE PRICE
stock prediction. PREDICTION
3) Sentiment classification: To utilize the
remaining 84.4% of the messages without the Data required for the stock market prediction placed
explicit sentiments, the build model extracted and generated at different places of the world.
the sentiments for those messages. A Distinguished agencies are maintaining this data in its
classification model will be trained from the own format. So, to cater need of current scenario, big
messages with annotated sentiments on the data architecture for the stock market has framed. Big
training dataset. Then it can used to classify data architecture for stock market divided into three
the remaining messages into five classes parts. i) First part is to identify different data and its
(Strong Buy, Buy, Hold, Sell, and Strong Sell). sources required for the future prediction of the
4) LDA-based method: In this model, we market.ii) Second Part is to store, acquire and process
consider each message as a mixture of hidden of the heterogeneous, unstructured and temporal

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


66
Nitha K. P., Dr. S. Sivakumari Int J S Res CSE & IT. January-February-2019 ; 5(1) : 63-69

data.iii) Last part of the architecture is to attain its Stock market is affected by all information. As per
goal, the prediction of the stock price. the efficient market hypothesis, stock market must
react on all of the information. Small and Middle
News and Information flows from any corner have class investors have not such type of resources to
impact on investment. Every second data generated know all happenings instantaneously and take their
for the market and its format is different from positions respectively. Loss fear factor had been
previous dataset. To store and integrate these playing a crucial role for the small and middle class
temporal data for the stock market is the biggest population and that is why they are keeping
challenge. In current scenario, cost required for these distance from the market. Easy availability of this
data is affordable by few big investors. Investors from integrated data from all corners of the globe is the
third and fourth world countries have acute scarcity key issue.
of resources. Big data architecture with latest
technologies will provide easy access of information Due to temporal nature of the data required in the
flow; it is relatively affordable to all investors. Little, stock market, big data architecture ,being
and slow access of information became big reason for implemented in such a way that all data must be
the big loss in the market for small investors. All updated and integrated on timely basis and reports
information need would be integrated very fast and should be generated based on the current
provide information to the small and medium information. To achieve efficient market hypothesis,
investors at very reasonable & affordable cost. these information must reach all participates of the
market. Due to shared resources and other above
Traditional database will not be able to handle and dis- cussed benefits of big data technologies, small
integrate this type of heterogeneous, temporal, investors too have equal opportunity to access
unstructured and dynamic data stored at different excess relevant data. This will increase participation,
places, so a new architecture is required for the participation will increase more investments,
integration of the heterogeneous database. Data will investments will increase more stable market with
not be available at one place or one database and it market hypothesis, more realistic market hypothesis
is in the different format too. It requires an will lead to more accurate prediction of the market
integrated mechanism which provides all temporal and finally it leads to stable and mature market.
data at one place and easy to process for the
decision-making process in desired format. Cloud based big data engine is used to integrate
different heterogeneous and unstructured data. Cloud
based technology will provide all analytical services
to the clients at reasonable costs. Distributed file
systems like HDFS (Hadoop Distributed File System)
and parallel processing in the cloud will increase
efficiency in the processing of data with quick
response. Big data integration will affect greatly on
the growing markets like India. There is no need to
store same data repeatedly. Intermediaries and other
interested parties gain benefits by two ways. First,
Fig. 1. Big Data Analysis Pipeline Architecture they can get data without wasting time and cost as
per their own desired format. Secondly, they exploit

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


67
Nitha K. P., Dr. S. Sivakumari Int J S Res CSE & IT. January-February-2019 ; 5(1) : 63-69

the benefit of other existing assistances of Big or from Tweets for Event Prediction, ,
cloud computing like SaaS, PaaS, and IaaS for International Conference on Intelligent
forecasting the share prices. A model developed based Computing, Communication & Convergence.
on the framework of big data analytics along with the Procedia Computer Science pp: 84 89,July 2015
blend of the fundamental, Technical and Sentimental [7]. Raghavendra Kune1,Pramod Kumar
factors will surely be a promising model for the Konugurthi1, Arun Agarwal,Raghavendra Rao
accurate prediction of stock prices in Indian stock Chillarige and Rajkumar Buyya "The anatomy
market. of Big Data Computing", SOFTWARE:
PRACTICE AND EXPERIENCE,Wiley Online
IV. REFERENCES Library (wileyonlinelibrary.com). DOI:
10.1002/spe.2374,9 October 2015
[1]. V. Sandhiya, T.Revathi, A.Jayashree, A.Ramya, [8]. Dr B Charumathi and Suraj E S "An Artificial
S.Sivasankari "Stock Mar- ket Prediction on Neural Network Approach on Improving
Bigdata Using Machine Learning Algorithm", Relative Valuation Accuracy for Bank Stocks in
Inter- national Journal of Engineering Science India Asian Research Journal of Business
and Computing,Volume 7 Issue No.4,April Management Issue 2 (Vol. 3) Issn: 2321-
2017 9246,2015
[2]. Rupawari Jadhav, M. S. Wakode "Survey : [9]. Kavitha S, Raja Vadhana P, Nivi A N "BIG
Sentiment Analysis of Twitter Data for Stock DATA ANALYTICS IN FINANCIAL
Market Prediction", International Journal of MARKET", IJRET: International Journal of
Advanced Research in Computer and Research in Engineering and
Communication Engineering,Vol. 6, Issue 3, Technology,Volume: 04 Issue: 02 - Feb-2015
March 2017. [10]. Aditya Bhardwaja, Yogendra Narayanb,
[3]. Venkata Sasank Pagolu,Kamal Nayan Reddy Vanrajc, Pawana, Maitreyee Duttaa Sentiment
Challa,Ganapati Panda,Babita Majhi "Sentiment Analysis for Indian Stock Market Prediction
Analysis of Twitter Data for Predicting Stock Using Sensex and Nifty, , 4th International
Market Movements", International conference Conference on Eco-friendly Computing and
on Signal Processing, Communication, Power Communication Systems. Procedia Computer
and Embedded System (SCOPES)- 2016 Science 70 (2015) pp: 85-91
[4]. Varunesh Nichante, Prof. Sulabha Patil "A [11]. Krishna Kumar Singh, Priti Dimri, Krishna
Review: Analysis of stock market by using big Nand Rastogi "The Impli- cations of Big Data in
data analytic technology", International Journal Indian Stock Market", International Journal of
on Recent and Innovation Trends in Computer Applications, Volume 99,Issue:5,
Computing and Communication,Volume: 4 August 2014
Issue: 1,ISSN: 2321-8169,January 2016 [12]. Dr B Charumathi and Suraj E S "Comparing
[5]. Aparna Nayak, M. M. Manohara Pai and Stock Valuation Models for Indian Bank Stocks
Radhika M. PaiPrediction Mod- els for Indian International Journal of Accounting and
Stock Market, , Twelfth International Multi- Taxation, Vol. 2, No. 2, pp. 111-127,June 2014
Conference on Information Processing-2016 [13]. Carson Kai-Sang Leung,Richard Kyle
(IMCIP-2016),pp : 441-449 MacKinnon,Yang Wang "A Ma- chine
[6]. P.G Preethia,V. Umab ,Ajit kumar ,Temporal Learning Approach for Stock Price
Sentiment Analysis and Causal Rules Extraction

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


68
Nitha K. P., Dr. S. Sivakumari Int J S Res CSE & IT. January-February-2019 ; 5(1) : 63-69

Prediction", IDEAS14, Porto, Portugal,ACM


978-1-4503-2627-8/14/07,July 2014
[14]. Yoosin Kim1, Seung Ryul Jeong1, Imran Ghani
"Text Opinion Mining to Analyze News for
Stock Market Prediction", International Journal
in Advanced Soft Computing and Applications,
Volume 6,Issue:1, March 2014
[15]. Qasem A. Al-Radaideh,Adel Abu Assaf,Eman
Alnagi "Predicting Stock Prices Using Data
Mining Techniques", The International Arab
Con- ference on Information Technology
(ACIT2013)
[16]. Dr B Charumathi and Suraj E S "Value
relevance of earnings and book value for bank
stocks An empirical analysis using ohlson
valuation model Global Journal of Management
and Research, July 2013.
[17]. Phaisarn sutheebunjard "Stock exchange of
Thailand Index prediction using back
propogation networks Second international
conference on computer and network
technology, IEEE 376-381, Thailand,2010
(2010).

Cite this article as :

Nitha K. P., Dr. S. Sivakumari, "Survey on Factors


Determining the Share Price Movement in Indian
Stock Market - A Big Data Approach ", International
Journal of Scientific Research in Computer Science,
Engineering and Information Technology
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
63-69, January-February 2019. Available at doi :
https://doi.org/10.32628/CSEIT195113
Journal URL : http://ijsrcseit.com/CSEIT195113

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


69
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1838110

Enhancing Privacy Preserving Query Retrieval Using Locality


Sensitive Hashing
Archana M.S., K. Deepa
1Assistant Professor, Department of Computer Science, Nilgiri College of Arts and Science, Thaloor, The
Nilgiris, Tamil Nadu, India
2Assistant Professor, Department of Computer Science, Providence College for Women, Coonoor, Ooty, Tamil
Nadu, India

ABSTRACT

The usage of smart phones is tremendously increasing day by day. Due to this, Location Based Services (LBS)
attracted considerably and becomes more popular and vital in the area of mobile applications. On the other
hand, the usage of LBS leads to potential threat to user’s location privacy. In this paper, the famous LBS provide
information about points of interest (POI) in spatial range query within a given distance. For that, a more
efficient and an enhanced privacy-preserving query solution for location based, Efficient Privacy-Location
Query (EPLQ) is proposed along with Locality Sensitive Hashing (LSH) reduces the dimensionality of high
dimensional data. Experiments are conducted extensively and the results show the efficiency of the proposed
algorithm EPLQ in privacy preserving over outsourced encrypted data in spatial range queries. The proposed
method performs in spatial range queries and similarity queries of privacy preserving.
Keywords: Location Based Services, points of interest, Efficient Privacy-Location Query, NICT

I. INTRODUCTION combination of various technologies such as


Geographical Information System (GIS) and other
Location Based Services (LBSs) are progressively in spatial and situating advances, the Internet and the
style in today’s society. It's reportable that up to a Web, and new data and correspondence advances
4.77 billion users are enjoying LBSs as of 2018 (NICTs).
compared to 150 million during 2014 [1]. Users can
get the required location information from LBS A framework is context-aware in the event that it
provider to get personalized services like restaurant utilizes context to give pertinent information as well
recommendations, taxi reservations and map as services to the client, where pertinence relies upon
directions. the client's task. Information or services can be
activated with context information with several
The idea of Location Based Services depends on long parameters mirroring the context of the client task.
periods of advancement and intermingling of various These parameters might be subdivided into spatial,
advances and in addition a development of personal, social, technical and physical context. The
information society where context and customization issueconnected with context- aware services is
of information is one of the principal needs for displaying those parameters in quantifiable and
clients. Considering the background of LBS, it is a computable way. The spatiotemporal information

CSEIT1838110 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 70-73 ]
70
Archana M.S., K. Deepa Int J S Res CSE & IT. January-February-2019 ; 5(1) : 70-73

about the client is one of few parameters that are III. DESIGN GOAL
clear and genuinely simple to quantify and utilize.
This is connected to the term Location-Aware It is very challenging to propose a privacy preserving
Services that may be characterized as context-aware query for implementing LBS over the encrypted
services that use the location of the client to embrace location data. Enhancing Processing of Query (EPQ)
the conduct of the services. It is a unique way of enables the cloud server to perform query processing
Location Based Services in a most broad sense as its without exposing client’s location details. EPQ can
services are connected with location data. insist on a client location coordinate query as per
client’s present location. Practically, the LBS
II. RELATED WORKS
application is mainly developed for the user who
used to search his point of interest such as searching
M. Gruteser et al investigates a complimentary
hotels nearest to him. In this work, the user query is
approach that concentrates on the principle of
secured with the keyword and encrypted. This will
minimal collection. In this approach location-based
accurately identify the requested location and the
services collect and use only de-personalized data
distance between the target locations are classified
that is practically anonymous data. But query
and ranking will be given. Our system provides a
encryption is not done and database is not outsourced.
multiuser environment as any number of users can
log at a time. The following are the three main
M. F. Mokbel introduces Casper; a novel framework
processes of our proposed work.
in which mobile users can entertain location based
services without the need to disclose their private
1. To propose a LBS query scheme which is more
location information. Per query privacy and multi-
efficient and privacy preserving to protect the
location query are absent and inefficient in handling
location data and the user’s request. For this, a
frequent attack are some of the drawbacks of this
hybrid encryption technique to encrypt the
work.
outsourced location data is proposed. This scheme
R Shokri et al proposed an approach that protects the
also supports multiuser cloud environment.
location privacy. In this approach, the mainstream
✓ System Construction Module
approach to protecting the location-privacy of mobile
✓ LBS User Module
users in location-based services (LBSs) is to alter the
✓ LBS Provider Module
users' actual locations in order to reduce the location
✓ Dimensionality Reduction Module – Locality
information exposed to the service provider. This
Sensitive Hashing
method is inefficient in terms of handling frequent
✓ Spatial Range Query For Privacy Preserving
attacks. M. E. Andres et al.

1.1 POI Query Processing


G. Ghinita et al employed the Private Information
Retrieval theory to guarantee privacy in location-
POI query processing in privacy preserving is
dependent queries. The main drawback is query
classified into two ways (i) Public LBS and (ii)
encryption is present but lacking in database
Outsourced LBS. In public LBS, POI of spatial
outsourcing.
database is recorded by an LBS provider in plaintext.
These queries are initiated at the provider’s site.
Outsourced LBS consists of three kinds of entities,
LBS provider, Cloud and LBS users.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


71
Archana M.S., K. Deepa Int J S Res CSE & IT. January-February-2019 ; 5(1) : 70-73

1) LBS data is abundant in LBS provider and these


data are nothing but POI records. Authorized users
are allowed by the LBS provider to access these data
by passing location based queries. the LBS provider
provides the query services through the cloud for
getting operational and financial benefits of data
outsourcing. But, the more valuable LBS data is not
disclosed by LBS provider to the cloud. Hence, the
LBS provider will outsource the encrypted LBS data
to the cloud.
Figure 4.1 (a) POI in Coimbatore
2) The cloud has abundant computing resources and
storage. The encrypted LBS data from the LBS
provider are stored and also provides query services
for the users. So, in cloud, the encrypted POI records
are searched in local storage for finding the data that
matches the queries of LBS users.

IV. RESULT DISCUSSION

In the previous section, we have discussed about the


modules for achieving the privacy preserving of Figure 4.2 (b) POI in Calicut
location data. These modules are implemented using
Java and MySQL. For accessing the location, we have V. CONCLUSION AND FUTURE WORK
implemented using J2ME as a mobile application. The
experiments are carried out on Intel Core(TM) I3 In this paper, the famous LBS provide information
Processor (2.27 GHz) about points of interest (POIs) in spatial range query
within a given distance is discussed. For that, a more
efficient and an enhanced privacy-preserving
The system takes input from the users, processes it
location-based query solution, Enhancing Privacy-
and produces an output. Input design is link that ties
Location Query (EPLQ) is proposed.
the information system into the world of its users.
The system should be user-friendly to gain
In specific, to obtain privacy preserving spatial range
appropriate information to the user.
query, the first predicate-only encryption scheme for
inner product range (IPRE) is proposed. This method
The project gives the low time consumption to make
is used to detect a privacy preserving way to find a
the sensitive application made simple. The fund that
position whether it is within a given circular area.
the company is willing to spend into the research and
The IPRE scheme has two vectors, attribute and
development of this system is very limited. The
predicate vectors. The problem with the proposed
design of input covers all the phases of input from the
method is, when the numbers of attributes are
creation of initial data to actual entering of the data
increased, the latency time for retrieving data is also
to the system for processing.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


72
Archana M.S., K. Deepa Int J S Res CSE & IT. January-February-2019 ; 5(1) : 70-73

increased. Here, we apply locality sensitive hashing location based queries,” IEEE Transactions on
to reduce dimensionality thereby supporting k- Knowledge and Data Engineering, vol. 26,no.5,
nearest neighbor search. pp. 1200-1210, 2014.
[8]. L. Sweeney, "Achieving k-anonymity privacy
VI. REFERENCES protection using generalization and
suppression,” International Journal of
[1]. Statista, "Number of location-based service Uncertainty, Fuzziness and Knowledge-Based
users in the United States from2013 to 2018 (in Systems, vol. 10,no. 5, pp. 571-588, 2002.
millions),” Statista; 2017.
https://www.statista.com/statistics/436071/locat Cite this article as :
ion-based- service-users-usa
[2]. K. Xie, X. Ning, X.Wang et al., "Recover Archana M.S, K.Deepa , "Enhancing Privacy
corrupted data in sensornetworks: a matrix Preserving Query Retrieval Using Locality Sensitive
completion solution,” IEEE Transactions on Hashing", International Journal of Scientific Research
Mobile Computing, vol. 16, no. 5, pp. 1434- in Computer Science, Engineering and Information
1448, 2017. Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
[3]. M. Li, H. Zhu, Z. Gao et al., "All your Location 5 Issue 1, pp. 70-73 , January-February 2019.
are belong to us: breaking mobile social Available at doi :
networks for automated user location tracking,” https://doi.org/10.32628/CSEIT1838110
in Proceedings of the 15th ACM International Journal URL : http://ijsrcseit.com/CSEIT1838110
Symposium on Mobile Ad Hoc Networking and
Computing, MobiHoc 2014, pp. 43-52, USA,
August 2014.
[4]. J. Shao, R. Lu, and X. Lin, "FINE: a fine-grained
privacy preserving location-based
serviceframework for mobile devices,” in
Proceedings of the IEEE INFOCOM, pp. 244-
252, IEEE, Ontario, Canada, May 2014.
[5]. G. Ghinita, P. Kalnis, A. Khoshgozaran, C.
Shahabi, and K.-Tan, "Private queries in
location based services: anonymizers are not
necessary,” in Proceedings of the
ACMSIGMOD International Conference on
Management of Data (SIGMOD ’08), pp. 121-
132, ACM, 2008.
[6]. B. Gedik and L. Liu, "Protecting location
privacy with personalized k-anonymity:
architecture and algorithms,” IEEE
Transactions on Mobile Computing, vol. 7, no.
1, pp. 1- 18, 2008.
[7]. R. Paulet, M. G. Kaosar, X. Yi, and E. Bertino,
"Privacy preserving and content-protecting

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


73
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1838111

Enhanced Security for Dynamic Multi Keyword Ranked Search


Using Greedy Best First Search
Bibin Baby1, Sharmila Banu2
1
Assistant Professor, Department of Computer Science, Nilgiri College of Arts and Science, Thaloor,
The Nilgris, Tamil Nadu, India
2
Assistant Professor, Department of Computer Science, Providence College for Women, Coonor,
The Nilgiris, Tamil Nadu, India

ABSTRACT

Today, due to the enormous growth of data technology in cloud computing, the data owners are stimulated to
outsource their data in data management to reduce cost and for the convenient. Data confidentiality, in general,
can be obtained by encrypting the data before it is outsourced. The client stores the encrypted data to the cloud
using Searchable encryption schemes and applies keyword search techniques over cipher text domain. But the
main problem in outsourcing is the lack of security and privacy for the sensitive data. So, to overcome this, for
privacy requirement, the sensitive data can be encrypted before it is outsourced. Various methods were
proposed to preserve the privacy and to provide security to the cloud data which are encrypted. Here in this
paper, we proposed a tree-based search method over the encrypted datain the cloud that supports dynamic
operation and multi-keyword ranked search. Clearly, the commonly used “inverse document frequency (IDF)
term frequency (TF)” model and the vector space method are joined in the query generation and index creation
to give multi-keyword ranked search. To get high search efficiency, a tree-type index structure, “Greedy Best-
first Search” algorithm is proposed based on the tree- index.
Keywords : Cloud Computing, Security, Multi Keyword Search, Greedy Best First Search, Inverse Document
Frequency, Term Frequency

I. INTRODUCTION themselves. No matter the various blessings of cloud


services, outsourcing touchy information (which
Nowadays, cloud computing has been taken into includes e-mails, non-public health statistics,
consideration as a new edition of IT infrastructure, enterprise business enterprise finance records,
that may put together large beneficial useful resource authority’s files, and plenty of others) to far flung
of computing, storage and programs, and allow users servers brings privacy worries. The Cloud Carrier
to revel in ubiquitous, convenient and on-demand Vendors (CSVS) that preserve the facts for users may
network get admission to a shared pool of also additionally get admission to users’ sensitive
configurable computing sources with splendid overall information without authorization. A fashionable
performance and minimal monetary overhead technique to shield the information confidentiality is
.Attracted by way of those appealing features, each to encrypt the statistics earlier than outsourcing but,
individuals and firms are stimulated to outsource this will motive a big rate in phrases of records
their facts to the cloud, rather of buying software usability. For an example, the present techniques on
program and hardware to manipulate the facts keyword-based totally information retrieval, which

CSEIT1838111 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 74-78 ]
74
Bibin Baby, Sharmila Banu Int J S Res CSE & IT. January-February-2019 ; 5(1) : 74-78

are widely used on the plaintext records, cannot be vector space model supporting solidified and
immediately implemented at the encrypted records. particular document look. The searchable file is
Downloading all of the statistics from the cloud and fabricated utilizing Multidimensional B tree. Its
decrypt regionally is impractical. On the opposite, limitations are the likeness rank score of the record
greater sensible unique motive solutions, including vector completely relies on upon the kind of the
searchable encryption (SE) schemes have made report Jiadi [7] proposed this pursuit utilizing two
unique contributions in phrases of performance, round searchable encryption (TRSE). The withdrawal
capability and safety. and restricting is utilized to diminish figure content
size, still the key size is too extensive. The
II. RELATED WORKS correspondence elevated will be high, if the
scrambled trapdoor's size is too vast. It doesn't make
This section discusses the literature survey of various powerful searchable file redesign. These are some of
papers related to Multi-keyword ranked search over the drawbacks of this method.
encrypted cloud data in cloud. It describes the
research work already done by the authors III. PROPOSED METHODOLOGY
corresponding to searchable encryption in cloud
computing. In this section, the Dynamic Multikeyword Ranked
Search for Plaintext (DMRSP)method is explained for
Qin Liu [3] proposed the pursuit that gives catch implementing our proposed method. Based on the
phrase protection, information protection and vector space model and Best First tree based indexing,
semantic secure by open key encryption. The main DMRSP is constructed. Apart from this, a secure
drawback of this system is the correspondence and search scheme, Simple DMRS (SDMRS) is also
computational expense for encryption and decoding explained.
is more. Cong Wang [4] proposed seek which
unravels preparing overhead, information and 3.1 Index Construction using Best First Search
catchphrase security, least correspondence and .
calculation aeronautical. The information proprietor
The search algorithm, Best-first search explores
assemble list alongside the catchphrase recurrence
a graph by increasing the most likely node chosen as
based importance scores for documents. Butit doesn't
per the specified rule.
play out numerous catch phrase seeks and Minimal
overhead in record building. C Wang et al [5] Judea Pearl explains best-first search as calculating
proposed that searchable encryption idea agree to the the promise of node ‘n’ by a "heuristic evaluation
client to solidly search for over scrambled function. It may depend on the depiction of n, the
information through watchwords without first goal and the information gathered. It also collects
applying decoding on it. But this system is too slow in additional knowledge about the problem domain.
processing the request and is not suitable for huge A priority queue is the efficient way of selecting the
volume of data. best candidate.

The first successor of the parent is expanded using


Wenhai Sun [6] proposed an inquiry that gives
greedy algorithm. Once a successor is generated,
comparability based item positioning, catchphrase
security, Index and Query classification and Query
Unlink capacity. The scrambled record is worked by

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


75
Bibin Baby, Sharmila Banu Int J S Res CSE & IT. January-February-2019 ; 5(1) : 74-78

1. If the heuristic of successor is best, then the For achieving this, we need to visit all nodes before
successor is put at the front of the queue and we reach the destination. Since priority queue uses
restarts the loops. Max or Min heap, insert and delete operations takes
O(log n).So the performance of the algorithm highly
2. or else, insert the successor into the queue
depends on evaluation function or the cost is
designed.
The algorithm for Best first search is explained below
for the following fig. 3.1 IV. RESULT DISCUSSION

These modules are implemented using Java and


MySQL. For accessing the location, we have
implemented using J2ME as a mobile application. The
experiments are carried out on Intel Core(TM) I3
Processor (2.27 GHz).

4.1 Precision
In the proposed method, dummy keywords affects
Fig. 3.1 Best First Search Tree the search precision. Search Precision can be defined
1. We start from source S and search for destination as
Ith node using Best First search with the given cost.
2. P and Q initially contains S Pkd=kd’= k wherekd’ = top k- documents
3. Remove S and process unvisited neighbors of S to To obtain higher precision, the very small standard
pq deviation σ is set as the random variable for ∑ 𝜀𝑣
4. Now pq contains {A, C, B} (C is put before B
because C has lesser cost) Table 4.1 Storage Consumption of Index Tree
5. Remove A from pq and process unvisited neighbors
of A to pq.
6. Now pq now contains {C, B, E, D} Size of the
1000 2000 3000 4000 5000
dictionary
7. Remove C from pq and process unvisited neighbors
of C to pq. GDFS (MB) 78 151 230 335 400
8. Now pq contains {B, H, E, D}
BFS (MB) 72 137 210 297 380
9. Remove B from pq and process unvisited neighbors
of B to pq.
10. Now pq contains {H, E, D, F, G}
From the above table it is clearly noted that, the
11. Remove H from pq.
memory occupied by Best first search is very less
12 Now return I
compared to the greedy depth first search.

Worst case time complexity : Best First Search : O(n


* Log n)
- n is number of nodes.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


76
Bibin Baby, Sharmila Banu Int J S Res CSE & IT. January-February-2019 ; 5(1) : 74-78

Table 4.2 Precision Percentages achieve greater search efficiency by implementing


our “Best first Search” algorithm.Even though our
proposed method works efficiently in terms of multi-
NO Precision NO Precision
1 89% 9 90% keyword search, still there are some serious issues
2 95% 10 96% which we will do in the future work. For instance,
3 96% 11 96% every time data owner generates updating
4 97% 12 87% information and this will be sent to the cloud server.
5 100% 13 100% So he has to store the plaintext tree index and IDF
6 86% 14 100%
values for recalculation.
7 88% 15 95%
8 90% 16 87%
VI. REFERENCES

We have randomly chosen 5 keywords as input for


each test and the precision is noted for the top 100 [1]. Q. Liu, G. Wang, and J. Wu, "Secure and

results. This test is repeated 16 times and the average privacy preserving keyword searching for cloud
precision obtained is 93%. The below table gives the storage services,' Journal of Networks and
precision percentage obtained for multi-keyword Computer Applications, vol. 35, no. 3, pp. 927–

search on searching the documents. 933, May 2012


[2]. C. Wang, N. Cao, K. Ren and W. Lou, -

V. CONCLUSION AND FUTURE WORK Enabling secure and efficient ranked keyword
search over outsourced cloud data, IEEE
A tree-based search method is proposed over the Transactions on Parallel and Distributed
encrypted data in the cloud that supports dynamic Systems, vol. 23, issue 8, (2012) August, pp.

operation and multi-keyword ranked search on the 1467–1479.


document collection. Precisely, the widely-used [3]. C. Wang, N. Cao, J. Li, K. Ren, and W. Lou,
“term frequency (TF) ×inverse document frequency "Secure ranked keyword search over encrypted

(IDF)” model and the vector space model are cloud data,' in Proc. of ICDCS'10, 2010.
combined in the index creation and query generation [4]. W. Sun, B. Wang, N. Cao, M. Li, W. Lou, YT,

to afford multi-keyword ranked search. In order to Hou and H. L, -Privacy-preserving multi-


get high search efficiency, a tree-based index keyword text search in the cloud supporting

structure is constructed and for that a “Best First similarity-based ranking, In Proceedings of the

Search” algorithm is proposed based on this tree 8th ACM SIGSAC symposium on Information,
index. The secure kNN algorithm is utilized to computer and communications security, (2013),

encrypt the index and query vectors, and meanwhile pp. 71-82.

ensure accurate relevance score calculation between [5]. N. Cao, C. Wang, M. Li, K. Ren, and W. Lou,
encrypted index and query vectors. From the results, "Privacy-Preserving Multi-Keyword Ranked

it is clearly states that the special structure, tree- Search over Encrypted Cloud Data,' in IEEE

based index can easily achieve sub-linear search time Transactions on Parallel and Distributed
and compact with the insertion and deletion of Systems, vol. 25, no. 1, pp. 222–233, 2014.

documents. Due to the tree-based index, the search [6]. Yu, Jiadi& Lu, Peng& Zhu, Yanmin&Xue,
complexity of the proposed work is basically kept to Guangtao& Li, Minglu. (2013). Towards Secure
logarithmic. The proposed scheme, in practice, can Multi-Keyword Top-k Retrieval over
Encrypted Cloud Data. IEEE Transactions on

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


77
Bibin Baby, Sharmila Banu Int J S Res CSE & IT. January-February-2019 ; 5(1) : 74-78

Dependable and Secure Computing. Cite this article as :


10.1109/TDSC.2013.9.
[7]. W. Sun, B. Wang, N. Cao, M. Li, W. Lou, YT Bibin Baby, Sharmila Banu, "Enhanced Security for
,Hou and H. L, -Privacy-preserving multi- Dynamic Multi Keyword Ranked Search Using
keyword text search in the cloud supporting Greedy Best First Search", International Journal of
similarity-based ranking, In Proceedings of the Scientific Research in Computer Science, Engineering
8th ACM SIGSAC symposium on Information, and Information Technology (IJSRCSEIT), ISSN :
computer and communications security, (2013), 2456-3307, Volume 5 Issue 1, pp. 74-78, January-
pp. 71-82. February 2019.
[8]. Cengizorencik , ErkaySavaş, An efficient Available at doi :
privacy-preserving multi-keyword search over https://doi.org/10.32628/CSEIT1838111
encrypted cloud data with ranking, Distributed Journal URL : http://ijsrcseit.com/CSEIT1838111
and Parallel Databases, v.32 n.1, p.119-160,
March 2014
[9]. Chen, Lanxiang, et al. "DMRS: an efficient
dynamic multi-keyword ranked search over
encrypted cloud data." Soft Computing 21.16
(2017): 4829-4841
[10]. C Orencik, M Kantarcioglu, E SavasA practical
and secure multi-keyword search method over
encrypted cloud data 2013 IEEE Sixth
International Conference on Cloud Computing,
390-397
[11]. Peng, Tao, Qin Liu, and Guojun Wang. "a
multilevel access Control scheme for Data
security in Transparent Computing."
Computing in Science & Engineering (2016).
[12]. Shen, Jian, et al. "An authorized identity
authentication-based data access control
scheme in cloud." Advanced Communication
Technology (ICACT), 2016 18th International
Conference on. IEEE, 2016.
[13]. Shen, Zhirong, JiwuShu, and Wei Xue.
"Keyword Search With Access Control Over
Encrypted Cloud Data." IEEE Sensors Journal
17.3 (2017): 858-868

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


78
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT19519

Cyber Security, Challenges, Some Practical Solutions


Y. V. Sai Bharadwaj, 2Sai Bhageerath Y.V, 3Prof. Y.V.S.S.S.V. Prasada Rao
1

1Central University of Hyderabad, Hyderabad, Telangana, India


2 National Institute of Technology, Warangal, Hanamkonda, Telangana, India
3Pattabhipuram, Guntur. Andhra Pradesh, India

ABSTRACT

Cybercrime continues to surge without a slowdown in sight. The cyber security threat continues to worsen. In
the first half of 2018, the number of cyber breaches soared over 140% from a year earlier, leading to 33 billion
compromised data records worldwide. Cyber Security news such as Marriott hack in Nov 2018 is dominating
headlines and becoming a serious headache for business leaders. Malicious outsiders sparked more than half of
the 944 breaches and accounted for roughly 80% of stolen, compromised or lost records. Identity theft
continues to lead data breach types, but financial access incidents are escalating in severity as well. The United
States continues to be the favorite target, and data breaches at major US enterprises continue to grab the
headlines. In 2018, the most notable breaches have occurred at Adidas, FedEx, Jason’s Deli, Macy’s, Under
Armour, Nordstrom’s and the most popular Facebook. [1].

Keywords : Cyber Security, Quantum Computing, Crypto-Agility, Phishing, Cyber Insurance, Phishing attack,
JDLR, Emotional Blackmail, CCO, TalkTalk Data Breach

I. INTRODUCTION breaches have occurred at Adidas, FedEx, Jason’s Deli,


Macy’s, Under Armour, Nordstrom’s and the most
Cybercrime continues to surge without a slowdown popular Facebook. [1].
in sight. The cyber security threat continues to
worsen. In the first half of 2018, the number of cyber II. Cyber security in small and medium sized
breaches soared over 140% from a year earlier, companies
leading to 33 billion compromised data records 2.1. Issue :
worldwide. Cyber Security news such as Marriott Cyber security in small and medium-sized companies
hack in Nov 2018 is dominating headlines and is a big issue, but businesses can take steps to protect
becoming a serious headache for business leaders. themselves. The issue of cyber security in small and
Malicious outsiders sparked more than half of the 944 medium-sized companies is becoming vital. [2]
breaches and accounted for roughly 80% of stolen,
compromised or lost records. Identity theft continues As for cyber security in small and medium-sized
to lead data breach types, but financial access companies, many are realising that they are viewed as
incidents are escalating in severity as well. The attractive a target as the larger companies. Cisco’s
United States continues to be the favorite target, and 2018 SMB Cyber Security Report found that 53% of
data breaches at major US enterprises continue to mid-market companies in 26 countries experienced a
grab the headlines. In 2018, the most notable breach. For these companies, the top security

CSEIT19519 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 79-87 ]
79
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

concerns are targeted phishing attacks against 2. Ensure you have a proper backup system.
employees, advanced persistent threats, ransom ware, And make sure it is easy to access in case you
denial-of-service attacks and the proliferation of need to restore one piece of the system rather
employees allowed to use their own mobile devices. than the entire system. Enterprise-level cloud
systems can help.
Malware of all types is a huge problem. It is 3. Examine all the entry points into your system
becoming more difficult to combat as cyber-attackers and consider where they are vulnerable.
get more adept at developing software that can evade These include all your workstations,
traditional detection and employ more sophisticated communications and mobile devices as well
malware. For small and medium businesses, one as employee access cards, the internet and
breach often puts a victim out of business. That’s cameras.
because 54% of all cyber-attacks cause financial 4. Assess your system threats. These include
damages exceeding $500,000, the 2018 Cisco SMB client lists, passwords, data logs, backups and
cyber security report shows. That price tag along emails, and anyone who specifically has
with a damaged reputation are hard to survive. If access to the system, including customers and
they do survive, they still face significant system vendors.
downtime that averaged eight hours or more in the 5. Put a prevention system in place to defend
last year. Further, such companies often lack the IT against intruders. Put yourself in the place of
talent, budget and technologies to prevent, uncover the cyber attacker and consider the possible
and respond to an attack. ways the attacker could access your system
and steal your data. If your internal IT staff
2.2. Prevention of Cyber Theft : isn’t experienced enough to handle, entrust a
third-party firm, because the prevention
Unsurprisingly, there is no easy solution – and none
system must cover physical and digital
is likely within the near future – to prevent data
security.
breaches. But all businesses, especially small and
medium sized companies can become better prepared III. Quantum Cryptography: The next-generation of
and more adept at protecting against cyber-crime. secure data transmission
3.1. Issue:
Here are five actions concerning cyber security in
Quantum Computing will render much of today's
small and medium-sized companies that can be taken
encryption unsafe. But Quantum Cryptography could
to become more security-conscious:
be the solution. China has recently successfully tested
– for the first time – its quantum communications
1. Conduct a security audit. Learn how secure
system, which cannot be hacked. [3].
your network and other security systems are,
where vulnerabilities exist and how to
3.2. How does quantum cryptography work ?.
resolve them. If you consider cybersecurity
insurance — currently the fastest-growing
Quantum cryptography provides a secure method for
insurance — or have coverage from a
generating and distributing secret keys between two
business insurer, the insurer can usually refer
parties via an optical network. This is achieved by
you to resources to assist in the audit.
capitalising on the inherent unpredictability in the
state of particles – such as photons or electron – to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


80
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

generate the random numbers needed for be resolved at the click of a button. Crypto-agility has
cryptographic applications. The technology then never been more important to ensure businesses can
harnesses this to create and share a secret digital key confidently protect themselves and their customers
that can be used to encrypt or authenticate from hackers. This is why organisations must invest
information via streams of encoded single photons, in a credible technology to automate the tracking of
which are sent through an optical communication certificates; it is no longer feasible to do this
network. manually, there are simply too many certificates to
track. Companies focusing too much on protecting
3.3. The arrival of quantum cryptography : usernames and passwords and not enough on
machine identities.
The arrival of quantum computing will ultimately
render much of today’s encryption unsafe. The 4.3. A future-proof solution :
current consensus that public key encryption is an
Google’s decision to distrust Symantec certificates
essential part of data security is starting to be
doesn’t need to be the end of the world. If companies
questioned in the face of new attack strategies –
are able to manage all their machine identities
which is subsequently driving uptake of robust
centrally and automate the process, it will enable
quantum cryptography solutions and services to
crypto-agility and ensure they can migrate quickly
deliver better data security. As a result, the global
when a flaw or vulnerability is discovered. By doing
quantum cryptography market is forecast to grow
so, companies can insulate themselves against the
from USD 285.7 Million in 2017 to USD 943.7
volatility of the CA market, protect their reputation
Million by 2022, a Compound Annual Growth Rate and ensure business continuity for online services.
(CAGR) of 27 per cent according to Research and
Markets. In reality, however, as things stand today V. What is phishing ?
quantum cryptography is not quite so close to
mainstream fruition.
The word ‘phishing’ was invented as a homophone of
IV. Crypto-agility: the key to ensuring long term ‘fishing’ as it involves creating a bait to lure
website security victims. Typically it involves an email; although
4.1. Issue : sometimes a telephone call – called Vishing – or a
In the age of the internet, you simply cannot afford text – called Smishing. These emails are often
to do business without being online, and keeping credible enough to deceive the recipient into clicking
your website secure should be a top priority [4]. on a link which could then release malware – viruses,
worms, Trojans or bots – onto the recipient’s
4.2. Crypto-agility to the rescue : computer or take the victim to a fake website.
This is why organisations need crypto-agility – i.e.
the ability to manage machine identities in real-time. 5.1. How to protect against phishing :
Crypto-agility enables businesses to quickly identify
The following are the tips against phishing.
and replace certificates in bulk when security events
or business needs call for it. Currently, many
organisations take days or even weeks to find and
replace certificates, which isn’t conducive to ensuring
5.1.1. Never, ever follow suspect links
security. By automating the process, this process can

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


81
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

There is no 100% guaranteed way to detect phishing If attachments with unknown file extensions (or PDF
but, if there is the slightest suspicion that the email files) suddenly appear as an e-mail attachment, it is a
may be fraudulent, do not click on any links it clear indication that something is wrong – especially
contains. Always enter the sender’s website address if you haven’t had any previous dealings with the
(not the link in the email) directly into your browser. sender.

5.1.2. Check out the sender 5.1.6. Personal salutation

Be warned if the part after the ‘at’ sign @ in an email Most companies address their customers by name.
address doesn’t match the purported sender; for But if the name is missing, misspelt or if there is no
example, if ‘PayPal’ sends you an email name at all and it just says something like ‘Hey’ or
from paypal@emails.com or the URL is misspelled ‘Dear Customer’, it could be an indication that this is
as www.paypa1.com or something similar. This is a a fake email.
(fake) website owned by a cyber squatter. Some of
the most well-known companies in the world have 5.1.7. Trust is good but control is better
website impersonators including Facebook, Google,
DropBox and PayPal. By regularly checking your bank statements, you can
mitigate any potentially serious consequences of a
5.1.3. Don’t give in to emotional blackmail phishing attack. Any suspicious or unknown
transactions should be reported directly to the bank
Phishing mails almost always contain the same kind or credit card company immediately.
of content and requests. Sometimes, they ask you to
update your user account or password. But 5.1.8. Keep yourself up-to-date on current scams
sometimes they use psychology to get you to react:
the notification of a big lottery win, an offer to take Take the time to read up regularly on ways to protect
part in a once-in-a-lifetime business opportunity or, your digital safety. If you hear that a service provider
an appeal for a donation to a charity. has been hacked, be sure to follow their instructions
and change your password.
5.1.4. Banks never want to know this
5.1.9. Only use secure websites
There are some things that your bank will never ever
ask you. They don’t want your passwords or PINs to When conducting online transactions, go directly to
be sent by e-mail or text; they don’t want you to the website. If the special offer is genuine, it will be
authorise the transfer of funds to a new account; and available on the website. Look for a sign that the site
they don’t want you to meet a bank representative at is secure, such as a white padlock icon on the
your home to collect cash, bank cards or anything browser’s status bar or a “https” URL (where the “s”
else. stands for “secure”).

5.1.10. Protect your computer with a firewall, spam


filters, anti-virus and anti-spyware software.

Do some research to ensure you are getting the most


5.1.5. Beware of opening attachments up-to-date software, and update them all regularly to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


82
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

ensure that you are blocking new viruses and The following are the cyber security trends and issues
spyware. (more stringent regulation, creations of new roles
etc.,) the world can expect in 2019: [ 6 ].
5.1.11. Click in haste, repent at leisure
6.1. Cyber security regulations improvement
Many phishing e-mails put pressure on you to act
quickly or else, they threaten, something bad will
happen or you will miss out on something very
We need to see a continuing improvement in the
important. A ‘bank‘ might warn you that your
relevant regulations as apply to cyber security. The
account will be closed unless you act quickly; or a
dynamic and fast-moving nature of cyber security
company might tell you that you have won a major
outpaces regulation which is far too slow and clumsy
cash prize, but only if you can claim it in the next 24
to be of any benefit and might actually hinder
hours. Don’t act in haste. Take your time to satisfy
security by building a culture of compliance with
yourself that the message is genuine.
regulations and a false sense of security against
enemies who are agile, motivated, and clever.
5.1.12. Genuine messages don’t make threats

6.2. Data theft turning into data manipulation


Although most phishing scams involve trying to trick
or persuade people into handing over sensitive We can expect to see attackers changing their
information, some fraudsters use fear and methodology from pure data theft and website
intimidation to scare their victims. For example, hacking to attacking data integrity itself. This type of

threatening to send embarrassing videos or photos to attack, in comparison to a straightforward theft of


contacts unless a ransom is paid. Try not to react data, will serve to cause long-term, reputational
immediately to an email take a few minutes to calm damage to individuals or groups by getting people to
down and think rationally. Why would this person question the integrity of the data in question.
be emailing you, specifically about this, all of a
6.3. Demand will continue to rise for security skills
sudden?.
A global shortage of cyber security skills in the
5.2. The JDLR rule : workplace arguably makes organisations more
desirable targets for hacking. Demand for expertise
Criminals are now able to put together professional- will rise as companies realise that their current IS
looking messages and web pages which can trick strategy is not sufficient. Also, with companies
even the most discerning person into giving away increasingly insourcing their security needs, internal
personal information when they are tired, busy or training and skills growth has to continue to
stressed. Check your privacy settings on popular accelerate. Tailored training programmes are crucial.
social networks to restrict how much personal
information you are making public and above all, 6.4. Cyber security and Internet of Things (IoT)
follow the JDLR rule. If it ‘Just Doesn’t Look Right‘, ‘Secure by design’ will garner much copy, but
then it probably isn’t. [ 5 ]. probably will not deliver until 2019 or beyond. We’ll
have to wait and see with this, as connected devices
are increasing in circulation by the day, and perhaps
it is only a matter of time before the security
VI. Cyber security trends in 2019

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


83
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

vulnerabilities are exposed — could there be a repeat the Deep Web, in order to successfully hide and to
of the Mirai Botnet in 2019?. Indeed, the next communicate with other criminals.
generation of AI-powered attacks will be crafty
enough to emulate the behaviours of specific users to 6.8. Breaches will get more complicated and harder to
fool even skilled security personnel.This might beat
include the ability to craft complex and bespoke Cybercriminals will look to grow their malicious
phishing campaigns that will successfully fool even activities using malicious code in ever more devious
the most threat-conscious among us. ways.Such a ransomware variant has already been
discovered using an innovative system to increase
6.5. Attackers will continue to target consumer infections: the software turns victims into attackers
devices by offering a pyramid scheme-style discount.If the
Ransomware is a recognised problem for companies victim passes on a link to the malware and two or
of all shapes and sizes, epitomised by the large scale more people install this file and pay, the original
WannaCry attack that decimated the UK’s National victim has their files decrypted for free.
Health Service (NHS) and organisations around the
world. In 2019 and beyond, will we start to see 6.9. Cyber risk insurance will become more common
consumers being targeted across a range of connected This type of insurance will increasingly become part
objects? This is a likely scenario, with examples of operational risk strategy however, the insurance
coming out of child predators targeting IoT devices in industry needs to tailor products specific to client
toys (designed for children). Attackers might even needs and not just provide blanket cover as
target the smart TV in your house via a ransomware extensions to existing risks. As the industry evolves
attack that would require you to pay a fee to unlock we might see cyber insurance covering for loss of
it. reputation and trust with their customers, loss of
future revenue from negative media or other
6.6. Attackers will become bolder, more commercial exposure, and improvement costs for security
less traceable infrastructure or system upgrades.

6.10. New job titles appearing – CCO (Chief


Hackers will look to become more organised and Cybercrime Officer)
more commercialised, perhaps even having their own In the aftermath of the TalkTalk data breach, MPs
call centres – something already seen with fraudulent recommended appointing an officer with day-to-day
dating sites. They will look to base themselves in responsibility for protecting computer systems from
countries where cybercrime is barely regarded as a attack.
crime and thereby placing themselves outside their
victims’ police jurisdictions. 6.11. Will 2019 see organisations looking to appoint a
chief cybercrime officer?
6.7. Attackers will get smarter
The CCO would be responsible for ensuring that an
Attackers capability to write bespoke targeted code organisation is cyber-ready, would bear the
will continue to improve faster than the defenders responsibility for preventing breaches, would take
ability to counter or get ahead of it.They will the lead if a breach did occur and provide a robust
continue to exploit the Dark Web, a small portion of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


84
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

connection between the board and the rest of the 7.3. Raise awareness, stay vigilant :
company. Awareness should always underpin your prevention
plan. SMEs are often targeted as the way in for
financial attacks through phishing or impersonation
with the aim to extract financial data or currency.
VII. Cyber security prevention plan :
This is often due to it being easier to take advantage
of human behaviour to target specific individuals
While there is no simple, singular solution to combat
who may not have had the right training. Whether
hack attacks and deter criminals, a layered
it’s lack of awareness or just lazy decisions, they need
prevention plan is crucial. There are five main steps
to be properly trained on the risks and repercussions,
that a company should take to prevent serious
along with potential hacking tricks as the majority of
damage from a cyber attack. They are as follows. [ 7 ]
malware still requires a human action to initiate it.
7.1. Assess your current cyber hygiene :

7.4. Layered prevention approach :


The UK Government backs an innovative scheme
called Cyber Essentials which is designed to help all There are business-grade security solutions that can

companies improve their cyber hygiene. It’s an help protect the network and users from would be

essential requirement for any business bidding on attacks such as anti-virus, anti-spam and business
government contracts, but it’s beneficial for any grade firewalls. Businesses should always invest in a
company, as it’s believed to reduce the risk of an layered defence strategy, as the more comprehensive
attack by up to 80%. It results in a certification which the set up is, the less chance an attack will succeed as
demonstrates a commitment to protecting business it has to pass through the various layers. Avoid free
and stakeholder data from threats – crucial for products and solutions which claim that they can
building customer trust. The scheme covers all keep malware off your PC. Windows Defender, for
essential controls such as firewalls, malware instance, doesn’t stop adware or Potentially
prevention and up-to-date software. Unwanted Programs (PUP) and doesn’t possess the
accuracy and effectiveness that more sophisticated
7. 2. Good housekeeping : prevention tools do. Paid-for prevention tools are a
small price to pay for the reassurance of digital safety.
There’s no replacement for the basics. Updates and
The cost of a hack to a business would be much more
patching should be performed regularly —
than the cost of the prevention technology.
WannaCry, the ransomware that caused chaos across
the world, exploited unpatched Windows systems to
7.5. Better safe than sorry :
spread malware. Businesses should steer away from
legacy systems like Windows XP, as these no longer While it’s important to be proactive, businesses need
receive updates and are especially vulnerable to a recovery plan in place in the event of a disaster or
attack. Get in the habit of conducting regular routine any downtime. A staggering 60% of businesses that
maintenance and audits, and seek a service provider encounter an attack go out of business in their first

if necessary, which can use specialist software to year, because of attacks to their network and users.
block ransomware strains. While the big dogs like the NHS, Sony and Equifax
dominate the headlines when they suffer attacks, it’s
the Small & Medium Enterprises (SMEs) that are
secretly suffering the worst due to factors such as the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


85
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

human element and not prioritising investments, as claim while operations are down and sales are halted.
mentioned above. There could be consequential damages if this causes
supply chain issues with its partners who are not
Under General Data Protection Regulation (GDPR) receiving necessary component parts, thereby leading
legislation, you must have a plan in place to be able to third party claims. That's why cyber insurance can
to restore data, whether it’s a cyber attack, file be a smart precaution for any size business.
corruption or simple data loss, otherwise you risk
non-compliance. IT continuity is the bread and 8.1. What is cyber insurance?
butter of so many businesses, and you must make
Cyber insurance generally covers your business'
sure that you have a backup and business continuity
liability for a data breach involving sensitive
plan to prepare, should the worst happen. Seek
customer information, such as Social Security
Disaster Recovery as a Service (DRaaS) to protect
numbers, credit card numbers, account numbers,
critical business data and get operational again after a
health records etc., General liability insurance covers
disaster.
bodily injuries and property damage resulting from
your products, services or operations. Cyber
While there is not a simple, singular solution to
insurance is often excluded from a general liability
combat hack attacks and deter criminals, a layered
policy.
prevention plan is crucial. Follow these steps, and
data misuse and abuse can be minimised, while still
8.2. Cyber insurance coverage :
enabling your business to take advantage of the
The cyber insurance covers the following.
growing opportunities the internet can harness.

VIII. Cyber Insurance • Legal fee and expenses


• Notifying customers about a data breach
Unfortunately, data breaches and other cyber crimes • Restoring personal identities of affected
are becoming way too common. In the past couple of customers
years, data breaches have resulted in major fines and • Recovering compromised data
legal fees – not to mention headaches – for a discount • Repairing damaged computer systems
retail chain, one of the nation’s largest banks, a well-
known health insurer, an entertainment network and
IX. CONCLUSION
the government. But it’s not just large organizations
that are susceptible to being hacked or getting a virus.
Throughout the year 2018, many cyber breaches
Nearly 55% of small businesses have experienced a
were reported. With a different issue almost every
data breach and that 53% have had multiple
week throughout the year, cyber security is at the top
breaches?. A data breach can damage more than just
of bysiness owners’ minds. Also, this is not something
your small-business computer system – it also can
that experts foresee ending any time soon. As such
damage your reputation and put your customers
there are some of the things you shall have to pay
and/or employees at risk. A bad actor deploys ransom
attention in the year 2019.Latrobe University claims
ware into a manufacturing company’s network
that it takes a combination of education and
environment which freezes up operations and delete
awareness to prevent cyber attacks from occurring.
files until it receives ransom payment. This could
Standard defense practices are only as strong as their
lead to direct revenue loss in a business interruption
weakest link, which is why it’s so important to train

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


86
Y. V. Sai Bharadwaj Int J S Res CSE & IT. January-February-2019 ; 5(1) : 79-87

your employees regarding phishing and spam emails. Cite this article as :
These act as an entryway for malware, so one should:
Y. V. Sai Bharadwaj, Sai Bhageerath Y.V, Prof.
• Use reliable, well-known security and antivirus Y.V.S.S.S.V. Prasada Rao, "Cyber Security,
software Challenges, Some Practical Solutions", International
• Make sure devices update automatically Journal of Scientific Research in Computer Science,
• Don’t click on email attachments from an Engineering and Information Technology
unknown source (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
• Don’t add people to your social media network 79-87, January-February 2019.
unless you know them Available at doi :
• Don’t use apps or software from free sites – Buy https://doi.org/10.32628/CSEIT19519
them from the official website instead Journal URL : http://ijsrcseit.com/CSEIT19519
• Routinely change passwords, making sure they
contain a unique combination of numbers and
letters (in both upper and lower case)
• Don’t use old, unsupported browsers

With these practices in place, we can expect a higher


level of cyber security in the year 2019.

X. REFERENCES

[1]. Gemalto, International data security company


[2]. Michael Fitzgibbon, Slice Insurance
Technologies
[3]. Dr. Andrew Shields, Cambridge Research
Laboratory of Toshiba Research Europe,
[4]. Scott Carter, Senior Manager – US, Venafi
[5]. Jan Oetjen, CEO of GMX
[6]. Andy Taylor, lead assessor, APMG
International
[7]. James Healey, Managing Director, Air IT

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


87
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195110

Prediction of Fish Yields in Lakes and Reservoirs from simple


Empirical Models using Artificial Neural Network (ANN) : An Review
*1D. Karunakaran, M. Balakrishnan2
1Research Scholar, Research and Development Centre, Bharathiar University, Coimbatore, Tamilnadu, India
2Principal Scientist, National Academy of Agricultural Research and Management, Hyderabad, Telangana, India

ABSTRACT

Prediction of reservoir yield is an important for fisheries managers to use appropriate scientific management
practices to increase the fishery production. Many mathematical or applied mathematics and Artificial Neural
Networks models were developed to predict fish production forecast of reservoirs. Ecology of reservoirs is
dynamic, extraordinarily advanced and nonlinear in nature. There are several drivers have an effect on the
fisheries, both internal and external environmental parameters. Many researchers have assessed fish yield
potential based on leaner models using multiple linear regressions. Accurate modelling to predict fish yield of
the reservoirs and lakes helps to understand behaviour of the system and managers can formulate appropriate
management practices to improve fish yield. This paper provides an in-depth review on existing model
developed from simple empirical estimation to high-level non-linear model for assessing fishery potential of
lakes and reservoirs.
Keywords : Fish Yields, Artificial Neural Network, MEI, CPUE, EBP, Automata Networks

I. INTRODUCTION foreign exchange earnings to the country. Most


important, the fisheries sector plays a great role in
Fish is nutritious food and rich in protein, minerals, the nation building by providing nutritional health
vitamins and essential nutrients. As fishes are very security.
cheap comparatively other animal proteins, many
low-income groups prefer fish in their diet. Most In India, reservoirs are prime resource in terms of
important, the fisheries sectors play a great role in both surface area and production. It offers immense
the nation building by providing nutritional health scope for increasing fish production. More than 3
security and livelihood support to many people in the million ha of manmade reservoirs available in the
country particularly in rural area and give many country can increase the production. As populations
ecosystem services to the mankind. India is richly are increasing rapidly, another blue revolution is
endowed with vast inland open waters in form of necessary to double the fish production from these
rivers and canals (1,71,334 km), reservoirs (3.15 resources to meet the demand. Accurately predicting
million ha), floodplain wetlands (0.24 million ha), reservoir yield will help fisheries managers to use
estuaries (0.27 million ha) and ponds and tanks (2.25 available various scientific management practices to
million ha) (Sinha 1999). These resources provide enhance the fishery production in the reservoir.
employment and livelihood support to many people
as well as contribute more to Indian economy by

CSEIT195110 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 88-100 ]
88
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

Ecology of reservoirs is dynamic, very complex and calculated (Wijayaratne and Amarasinghe, 1984;
nonlinear in nature. There are many drivers Nissanke et al., 2000) and (Hasan et al., 2001)
influencing the fisheries both internal and external predicted fish yield based on chlorophyll-a, Secchi
environmental parameters. Many researchers have depth and morpho-edaphic index in Bangaladesh.
tried to assess fish yield potential based on leaner
models by using multiple linear regression analysis
techniques (Jowett, 1993). These empirical models Others have found that phytoplankton production
are capable of solving many linear problems, but can predict fish yield more accurately (Oglesby,
some time it cannot give accurate results when 1977) or total phosphorus and macro benthos
relationships among the variables are non-linear in biomass/mean depth found to be the best predictors
nature. Advance tools like neural network, fuzzy of fish yield better than MEI (Hanson and
logic, wavelet transform and genetic algorithm will Legget,1982). Surface area alone or Lake Shoreline
give better and accurate results for non-linear development is useful predictor variables of fishyield
problems. in North American lakes (Youngs and Heimbuch,
1982). Hrbacek (1969) found that significant
II. PREDICTION OF FISH YIELD USING correlation between primary production and carp
EMPIRICAL MODEL yield in European ecosystems. Melack (1976) and
Toews and Griffith (1979) reported that primary
Many researchers predicated fish yields in lakes and production is an estimator of fish yield. Stocking
reservoirs for the last 70 years using simple empirical densities of fingerlings in inland reservoirs can be
estimation. Rawson (1952) demonstrated the first calculated based on estimation of potential fish
model to estimate fish yields of lakes based on mean production using morpho-edaphic index (MEI)
depth. Some others also tried to assess fish yield (Welcomme, 1976). Surface area can alone be the
potential from biotic and morphometric parameters best predictor of fish yield reported by researchers
(Hayes, 1957; Northcote and Larkin, 1956). Ryder from North American lakes (Youngs and Heimbuch,
(1965) improves existing estimation further by 1982).
developing morpho-edapic index (MEI) by using total
dissolved solids or conductivity/depth using 23 Models have been developed using reservoir depth,
temperate lakes. Later many researchers used other volume and area for estimation of fishery yield
dependent variables such as lake surface area, (Rawson, 1952; Jenkins and Morais, 1971; Moreau
temperature and other parameters along with MEI and De Silva, 1991). Downing et al., (1990) reported
for prediction of fish yields in lakes and reservoirs in annual phytoplankton production correlated (r2=
many countries. (e.g. Toews & Griffith, 1979; Jenkins, 0.79) than MEI, total phosphorus concentration and
1982; Schlesinger & Reglier, 1982; Machena & Fair, macro benthos biomass/mean depth, were the best
1986). Ryder (1965) applied this method in Canada. univariate predictors of fish yield (r2= 0.84 and r2=
Matuszek (1978) used mean depth and total dissolved 0.48, respectively). Hanson and Leggett (1982)
solids concentration of large North America lakes. reported catchment land use patterns are key
Morpho-edaphic index along with soci-economic parameters for fish yield prediction of reservoir (De
variables such as numbers of fisher men, boat and Silva et al., 2001). Rawson (1938) grouped lake
effort are used to predict yield of African lakes productivity parameters in three groups i.e climatic,
(Henderson and Welcomme, 1974). Maximum morphometric, and edaphic
Sustainable fish yield of Sri Lanka reservoirs were

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


89
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

and fishing effort. Crul)1992( correlated with catch )t


A. Potential Fish Yield from Morphological y-1( and area )km2( for 46 lakes and 25 reservoirs in
parameters Africa.

Reservoir area, volume, depth, and shoreline


Brämick, U and Roland (2003) estimated fish yield
development or gradient is some of the key
potential based on data of 786 lakes in north-east
morphological parameters to assess the productivity
Germany using primary production and total
(Ryder, 1978(. The mean depth alone a single most
phosphorus. Relationship between fish yield
important parameter to assess the fish yield )Rawson,
potential and total phosphorus or chlorophyll a or
1952; Henderson and Welcomme, 1974; Ryder et al.,
primary production of phytoplankton have been
1974; Mehner et al. 2007). Henderson & Welcomme
published and reviewed (Nurnberg 1996; Knösche, R.,
(1974) first applied Ryder's morphoedaphic index to
& Barthelmes, D. 1998). John A. Downing and Céline
tropical fisheries in African lakes and derived
Plante(2011) estimated annual fish production using
relationship from 17 fully exploited lakes where more
independent variables temperature, phosphorus
than 1 fisherman fishing km-2 lake area. Oglesby
concentration, chlorophyll a concentration, primary
(1977) reported that mean depth was not relevant
production, and pH .Knösche, R., & Barthelmes, D.
variable for assessing fish yield, where lake having
(1998) estimated lake fisheries yield from primary
more than 25m deep. Schlesinger & Regier (1982)
plankton production or total phosphorus. John Mark
found that fish yield - MEI relationships accurate
HansonandWilliam Leggett (2011) reported that total
only to lakes within regional level and generalizing
phosphorus concentration and macro benthos
the relationship need to include other dependent
biomass/mean depth best predictor than
variable such as mean annual air temperature into the
morphoedaphic index. Kolding, J., & Van Zwieten, P.
model. Hanson & Legget (1982) developed many
A. (2012) reported water-level fluctuation and mean
equations based on total dissolved solids (TDS)and
depthis a simple estimatorfor fish productivity in
total phosphorus concentration (TP( which was
tropical lakes and reservoirs.
highly correlated and much better predictor of fish
yield. Downing et al. (1990) has reported based on
Wijeyaratne and Amarasinghe )1987( showed that
his study that fish production is closely correlated
maximum sustainable yield )MSY( in several
with annual phytoplankton production (R2 = 0.79),
reservoirs of Sri lanka were correlated with the
mean total phosphorus concentration (R2 = 0.67), and
morphoedaphic index. Amarasinghe et al. )2002(
annual average fish standing stock (R2 = 0.67) but
have shown catchment features like ratio of
least correlated with the morphoedaphic index
catchment, land use patterns and reservoir capacity
(p>0.05). Schneider & Hadrich, (1989) reported that
are important predictor variables in Sri Lankan.
fish landings varied proportion to lake area.
Ramakrishna )1990( reported that catchment has a
Moreau & De Silva )1991( developed fish yield model positive impact on the Indian reservoirs productivity,
using multiple regression for lakes and reservoirs of provided the catchment is moderately fertile
Sri Lanka, Thailand and Philippines using area and
(Natarajan, 1976, 1977, 1979; Jhingran, 1986(.
effort. Moreau & De Silva )1991( also tested a number
of models using predictor variables catchment/lake Vollenweider (1969( reported that Flushing rate
area ratio, mean depth, transparency, total alkalinity, )inflow/storage capacity( is important variable for fish
chlorophyll a concentration, primary productivity production as it regulates nutrient loading.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


90
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

B. Potential Fish Production from Primary Rawson,1952; Depth, Models have


Production Jenkins and Morais, Volume and been
1971; Moreau and Area developed
Oglesby (1977) has derived relationships between De Silva, 1991; based on
fish and summer standing crop of phytoplankton. Bernascek, 1997 morphometric
Jones & Hoyer (1982), correlated sport fish harvests features
with summer chlorophyll a in US reservoirs and Ryder (1965) Total Morpho-
dissolved edaphic Index
lakes. Liang et al., (1981) derived relationship
solids or )MEI(
between net fish yield and gross primary production conductivity predictive
in sub-tropical Chinese lakes and ponds. Relative fish divided by yield model
biomasses expressed as gillnet CPUE were found mean depth for lakes and
correlated to chlorophyll a, total phosphorus, total reservoir
nitrogen and total organic matter in a large number D. R. Toews and J. S. Morphoeda Predicted fish
Griffith (1979) phic- yield using 31
of Argentinean lakes and reservoirs (Quiros, 1990).
index)MEI( African lakes.
Schlesinger & Regier Air Prediction of
Xiong )1996(, shown that 15 factors such as mean (1982) temperature fish yield from
depth, catchment area, water temperature, and MEI reservoirs
precipitation, dissolved oxygen content, total having less
phosphorus, phytoplankton biomass and the number than 25 m
of stocked fingerlings responsible for fish yield. depth
C. Fish Yields from Socio-economic Variables Oglesby (1977) and Chlorophyll Predictive
Biro & Vörös (1988) -a yield model
Fishing effort, number of fishermen or number of
boats or units of fishing gear are important predictor Ryder (1978) Reservoir Predict the
area, reservoir or
socio-economic variables for predict fish yield
Volume, lake
(Henderson & Welcomme, 1974; Bayley, 1988). Depth, and productivity
Shoreline
developmen
TABLE – I. LIST OF MODELS DEVELOPED TO
t or gradient
PREDICT FISH YIELDS IN LAKES AND Biro&Vörös (1988) Ratio of Predicting the
RESERVOIR and Moreau & De surface area fish yields of
Silva (1991) to lakes and
Author(s) Predictor Remarks catchment reservoirs in
variable area Sri Lanka
Rawson (1952) Mean depth Fish yield )CA(
prediction of
Lakes Nissanka, Ratio of the Better
Crul)1992( Area Establishedrela Amarasinghe & De CA to predictor
tion-ship Silva (2000) reservoir variable for
between catch capacity fish yield
and area from
46 lakes and Hrbacek (1969) Primary Reported
25 reservoirs production highly
in Africa significant

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


91
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

relationship cover )FC( Sri Lanka


between and shrub-
primary land )SL( to
production reservoir
and carp yield surface area
from European )RA( and/or
ecosystems. reservoir
capacity
Melack (1976), Gross Useful )RC(
Toewsand Griffith primary estimator of Amarasingheet al. Ratios of Prediction of
(1979) and production fish yield )2002( forest cover fish yield
Plante&Lalonde based on data and/or
(1990) of African shrub cover
lakes. to reservoir
Hanson and Total Better capacity or
Legget(1982) phosphorus predictors of reservoir
fish yields area
than the MEI Downing et al. Phyto- Prediction of
Jenkins and Morais Surface area Opined that )1990( plankton fish yield
(1971); Youngs and alone surface area production,
Heimbuch (1982) can alone be Total
the predictor Phosphorus
of fish yield of concentratio
North n , Annual
American average fish
lakes standing
Henderson and Electrical Studies based stock
Welcomme(1974) conductivity on Africa and Xiong)1996( Phyto- Methodology
Srilanka plankton developed to
primary predict silver
Morpau and De Silva Watershed Asian reservoir production carp and
(1991) area and bighead carp
Mean depth production(Ch
Ramakrishniah(1990 C/A radio Modified MEI ina)
), Ramakrishniah el along with model Xiong (1996) Mean depth, Reported
al. 1998) MEI incorporating Catchment major factors
the drainage area, Water influencing
parameter, the temperature fish yield
ratio of ,
catchment to Precipitatio
reservoir area n, Dissolved
(C/A), (based oxygen
on 19 Indian content,
reservoir) Total
phosphorus,
S. S. De Silvaet. Catchment GIS Model Phyto-
al.(2001) parameters using 11 plankton
i.e forest reservoirs of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


92
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

biomass and (1981) Primary )FYP(or Fish


Number of Hanson&Leggett production Biomass)FB(.
stocked (1982) of
fingerlings Peters (1986) phytoplankt
Henderson & Fishing Socio- Leach et al. (1987) on )PP(
Welcomme, 1974; effort economic Downinget al. (1990)
Bayley, 1988 ( number of variable to Qutros (1990 &
fishermen) predict fish 1991)
or number yield Barthelmes (1992)
of boats or Downing&plante
units of (1993)
fishing gear Nurnberg (1996)
John A. Downing Temperatur Estimated Knosche&Barthelme
and Céline Plante e, Lake annual s(1998)
(2011) phosphorus production of Reiner Knösche and Primary The most
concen- fish D. Barthelmes (1998)
plankton promising
tration, populations production limnological
Chlorophyll )PP( or parameters for
-a concen- Total fish yield
tration, phosphorus estimation
Primary JeppeKoldingandPau Water-level Reported
production, l A.M.van fluctuation simple and
and pH. Zwieten)2012( and mean robust
depth indicator of
Water-level Simple robust fish
Jeppe Kolding and fluctuation indicator of productivity in
PaulA.M.van and mean fish tropical lakes
Zwieten(2012) depth productivity in and reservoirs
tropical lakes Hoyer (1982) chlorophyll Correlated
Chlorophyll All these a sport fish
C Nissanka; U S a, Dissolved parameters harvests in US
Amarasinghe; and S phosphorus were found reservoirs
S De Silva and Total positively
phosphorus, influence with III. ARTIFICIAL NEURAL NETWORKS (ANNs)
Alkalinity to fish yield in
mean depth reservoirs
Artificial Neural Networks (ANNs) are mathematical
)MEIa( and
models designed to mimic the information processing
Conductivit
y to mean functions of a network of neurons in the brain
depth and (Hinton1992; Jensen 1994). Humans and animals are
CA/RC processing information by neurons. Computer
ratios algorithms mimic the way biological systems are
Dillon &Rigler Total Empirically functioning called artificial neural networks. The
(1974), phosphorus derived
brain is a highly complex, nonlinear and parallel
Oglesby (1977) )TP(, relations
Lianget al. (1981) Chlorophyll between Fish processing system computing many times faster than
Bulon&Vinberg -a or Yield Potential digital computer. It is widely used in many
disciplines for modelling. An artificial neural

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


93
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

network is highly popular because of similarity in Mediterranean Sea using ANN. The input variables
biological systems and has anability to learn from are biomass/abundance ratio, depth of the water
experiences and improving its performance. It is a column, latitude and longitude, and
very powerful tool for modelling non-linear and monthlysampling of675 sample stations of North
complex, imprecise and noisy data. It has an ability to Aegean Sea.Error back-propagation (EBP) algorithm
manipulate large amounts of data and generalize is used to learn the network.
results.
Baran, Pet.al.(1996) developed stochastic models to
A. Prediction of Fish Yields in Lakes and Reservoirs predict trout population density or biomass on a
using Artificial Neural Network mesohabitat scale using neural networks.
Backpropagation algorithm was used to train habitat
ANN have been used in many ecological modeling, variables width, gradient, mean depth, coefficient of
phytoplankton production (Scardi, 1996), fish species variation of depth, mean bottom velocity, coefficient
richness prediction (Brosse, Set al., 1998), and of variation of bottom velocity, froude number, area
prediction of density and biomass of various fish of cover, area of shelter, pool deep water area and
populations (Baran et al., 1996; Lek et al., 1996a,b; elevation. The result shows that the back propagation
Mastrorilloet. al., 1997). Many authors have reported neural network estimatedmore accuratelythan the
that ANN predicts more accurate than multi leaner multiple regression.
regression (Ehrman et al .1996; Leket. al. 1996b;
Scardi1996). Brosse, S. et. al. (1999) predicted fish spatial
occupancy and abundance in a mesotrophic reservoir
Laë, R., Lek, S., & Moreau, J. (1999) predicted fish using Artificial Neural Network.The input databased
yield of African lakes using Artificial neural networks on 306 observations collected using electrofishing.
using six input variables such as catchment area, The relationships between physical parameters and
maximum area, fishing effort, conductivity, depth, the abundance fish species are studied. Eight
altitude and latitude. The structure of the feed independent environmental variables depth, distance
forward neural network is six input neuron with one from the bank, slope of the bottom, flooded
hidden layer of five neurons, sigmoid functions, and vegetation cover, percentage of boulders, percentage
backpropagation algorithm for the training of the of pebbles, percentage of gravel and percentage of
ANNs. The result shows that correlation coefficients mud are used to quantify fish density. Structure of
between the estimated and observed values were the network is feed-forward with eight input neurons,
significantly very high. one hidden layer with 10 nodes with bias and
Zhang, H., & Zimba, P. V. (2017) developed multi- aoutput layer (8-10-1). The network was trained
layer neural networks to assess effects of estuarine using back-propagation algorithm to predict
freshwater fluxes on fish abundanceusing artificial accurately with minimum error.
neural network. The predictor variable consists of
inflow, evaporation, precipitation and annual catch Scardi, M., et.al (2008) evaluated ecological integrity
rate of fish species of Nueces Estuary. The network of streams and rivers using an expert system based
was trained using Levenberg-Marquardt back fish assemblage model. Twenty-seven environmental
propagation algorithm. Maravelias, C. D., Haralabous, parameters along with fish assemblage composition of
J., & Papaconstantinou, C. (2003) predicted 63 locations in Latium streams and rivers were used
distributions of demersal fish species in the in this study. Neural network was trained using most

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


94
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

popular error back-propagation algorithm. Aoki, I., maximum depth, shoreline perimeter, elevation, total
and Komatsu, T. (1997) predicted the winter catch of dissolved solids, pH, summer stratification (0, 1) and
young Japanese sardine (Sardinopsmelanostictus) littoral-zone predator (0, 1). Artificial neural
from climatic, hydrological and biological parameters networks predicted accurately abundance as well as
in the Joban-Boso Seas of the Pacific coast of central occurrence of fish species.
Japan using neural network. A feed-forward three
layers an input, hidden and an output layer is used. Kılıç, H., et. al (2007) developed ANN models to
Predictor variables are hydrological and biotic predict primary production of reservoir by using
parameters. Hydrological parameters consist of preprocessing technique of an Automata Networks
southern limit of the Oyashio, path type of Kuroshio, (AN) to find suitable variables for subsequent ANN
northern limit of the Kuroshio Extension, sea surface modeling. The AN based preprocessing followed by a
temperature in the northeastern sea area of Japanand ANN application predicted primary productivity
sea surface temperature in the fishing ground. accurately using Chl-a. The correlation coefficient as
Zooplankton, climatic southern oscillation index, far high as 0.83 and RMSE was as low as 2.69g/l was
east zonal index and east sea index are biological achieved withdouble hidden layer structure with 10
parameters. The network trained by back- neurons.
propagation algorithm to predictchanges in the
sardine abundance. Kuo, J. T., Hsieh, M. H., Lung, W. S., & She, N. (2007)
predict reservoir water quality using Artificial Neural
Joy, M. K., & Death, R. G. (2004) developed Network with back-propagation algorithm. The input
predictive model and spatial mapping of decapod variables are dissolved oxygen (DO), total phosphorus
assemblages using combination of GIS and neural (TP), chlorophyll-a (Chl-a), and secchi disk depth
networks. Single hidden-layer feed forward multi- (SD). Results show that correlation coefficients
layer perceptron trained by back propagation error between predicted values and measured data are 0.7
algorithm are used in this network. Catchment area, with reasonable accuracy.
average catchment elevation, elevation upstream end
of reach, average catchment slope, average annual IV. CONCLUSION
catchment rainfall, average catchment air
temperature, estimated river flow, reach length, Prediction of reservoir and lakes fish yield is the
latitude, stream order, catchment rainfall, distance important factor for the fishery managers to improve
from the coast, lake catchment area, catchment land the decision-making processes of reservoir fishery
use proportion and catchment geology proportions management and use appropriate management
(surface rock) are input variables. The decapod was practices to enhance the fish production. There is
predicted with high degree of accuracy from extensive literature available over the past decade on
geospatial landscapepredictor variables. prediction of fish yield in reservoir and lakes both in
linear and non-linear models. Many authors reported
Olden, J. D., & Jackson, D. A. (2001). Developed fish– that Artificial Neural Network model is giving high
habitat models for nine fish species using Artificial level of accuracy than Multilinear Regression Model.
Neural Networks. The study was conducted using
data of 128 lakes from the Madawaska river drainage
and 32 lakes from the Oxtongue river drainage of
Canada. The habitat input variables are area,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


95
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

II. REFERENCES fish production in different waters. The basis of


freshwater ecosystem investigation, 5-10.
[1]. Amarasinghe, U. S., De Silva, S. S., & Nissanka, [10]. Crul, R. C. (1992). Models for estimating
C. (2002). Evaluation of the robustness of potential fish yields of African inland waters.
predictive yield models based on catchment CIFA Occasional Paper (FAO). no. 16.
characteristics using GIS for reservoir fisheries [11]. D'Angelo, D. J., Meyer, J. L., Howard, L. M.,
in Sri Lanka. Fisheries Management and Gregory, S. V., & Ashkenas, L. R. (1995).
Ecology, 9(5), 293-302. Ecological uses for genetic algorithms:
[2]. Aoki, I., & Komatsu, T. (1997). Analysis and predicting fish distributions in complex physical
prediction of the fluctuation of sardine habitats. Canadian Journal of Fisheries and
abundance using a neural network. Oceanolica Aquatic Sciences, 52(9), 1893-1908.
Acta, 20(1), 81-88. [12]. De Silva, S. S., Amarasinghe, U. S., Nissanka, C.,
[3]. Aoki, I., & Komatsu, T. (1997). Use of a neural Wijesooriya, W. A. D. D., & Fernando, M. J. J.
network to analyse and predict the winter catch (2001). Use of geographical information systems
in the Joban-Boso Seas off the Pacific coast of as a tool for predicting fish yield in tropical
central Japan of young Japanese sardine reservoirs: case study on Sri Lankan reservoirs.
(Sardinopsmelanostictus) Fisheries Management and Ecology, 8(1), 47-60.
[4]. Baran, P., Lek, S., Delacoste, M., & Belaud, A. [13]. De Silva, S.S. and Amarasinghe, U.S. (eds.).
(1996). Stochastic models that predict trout (2009) Status of reservoir fisheries in five Asian
population density or biomass on a mesohabitat countries. NACA Monograph No.2. Network of
scale. Hydrobiologia, 337(1-3), 1-9. Aquaculture Centres in Asia-Pacific, Bangkok,
[5]. Bayley, P. B. (1988). Accounting for effort when Thailand. 116p.
comparing tropical fisheries in lakes, river‐ [14]. Dillon, P. J., &Rigler, F. H. (1974). The
floodplains, and lagoons. Limnology and phosphorus‐chlorophyll relationship in lakes 1,
Oceanography, 33(4part2), 963-972. 2. Limnology and oceanography, 19(5), 767-773.
[6]. Bramick, U., & Lemcke, R. (2003). Regional [15]. Downing, J. A., & Plante, C. (1993). Production
application of a fish yield estimation procedure of fish populations in lakes. Canadian Journal of
to lakes in north-east Germany. Limnologica- Fisheries and Aquatic Sciences, 50(1), 110-120.
Ecology and Management of Inland Waters, [16]. Downing, J. A., Plante, C., & Lalonde, S. (1990).
33(3), 205-213. Fish production correlated with primary
[7]. Brosse, S., Guegan, J. F., Tourenq, J. N., &Lek, S. productivity, not the morphoedaphic index.
(1999). The use of artificial neural networks to Canadian Journal of Fisheries and Aquatic
assess fish abundance and spatial occupancy in Sciences, 47(10), 1929-1936.
the littoral zone of a mesotrophic lake. [17]. Fausch, K. D., Hawkes, C. L., & Parsons, M. G.
Ecological modelling, 120(2-3), 299-311. (1988). Models that predict standing crop of
[8]. Brosse, S., Lek, S., & Dauba, F. (1999). Predicting stream fish from habitat variables: 1950-85. Gen.
fish distribution in a mesotrophic lake by Tech. Rep. PNW-GTR-213. Portland, OR: US
hydroacoustic survey and artificial neural Department of Agriculture, Forest Service,
networks. Limnology and Oceanography, 44(5), Pacific Northwest Research Station. 52 p, 213.
1293-1303. [18]. Hanson, J. M., & Leggett, W. C. (1982).
[9]. Bulon, V. V., &Vinberg, G. G. (1981). Empirical prediction of fish biomass and yield.
Proportions between primary production and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


96
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

Canadian Journal of Fisheries and Aquatic reservoirs. Transactions of the American


Sciences, 39(2), 257-263. Fisheries Society, 111(2), 176-179.
[19]. Hasan, M. R., Mondal, M. A. W., Miah, M. I., & [29]. Jowett, I. G. (1993). A method for objectively
Kibria, M. G. (2001). Water quality study of identifying pool, run, and riffle habitats from
some selected Oxbow Lakes with special physical measurements. New Zealand journal of
emphasis on chlorophyll-a. In. Reservoir and marine and freshwater research, 27(2), 241-248.
culture-based fisheries: Biology and [30]. Joy, M. K., & Death, R. G. (2004). Predictive
management. De Siva, S. S. (Ed.)2001. 384 pp. modelling and spatial mapping of freshwater fish
[20]. Hayes, F. R. (1957). On the variation in bottom and decapod assemblages using GIS and neural
fauna and fish yield in relation to trophic level networks. Freshwater Biology, 49(8), 1036-1052.
and lake dimensions. Journal of the Fisheries [31]. Kılıç, H., Soyupak, S., Tüzün, İ., İnce, o.,&
Board of Canada, 14(1), 1-32. Başaran, G. (2007). An automata networks based
[21]. Henderson, H. F., & Welcomme, R. L. (1974). pre-processing technique for artificial neural
The relationship of Morphoedaphic Index and network modelling of primary production levels
numbers of fishermen in African inland in reservoirs. Ecological modelling, 201(3-4),
fisheries. Committee for Inland Fisheries of 359-368.
Africa. Occasional Paper I, FAO, Rome, Italy. [32]. Knosche, R., & Barthelmes, D. (1998). A new
[22]. Henderson, H. F., &Welcomme, R. L. (1974). approach to estimate lake fisheries yield from
The relationship of yield to morpho-edaphic limnological basic parameters and first results.
index and numbers of fishermen in African Limnologica, 28(2), 133-144.
inland fisheries. Food and Agriculture [33]. Kolding, J., & van Zwieten, P. A. (2012).
Organization of the United Nations. Relative lake level fluctuations and their
[23]. Hinton, G. E. (1992). How neural networks influence on productivity and resilience in
learn from experience. Scientific American, tropical lakes and reservoirs. Fisheries Research,
267(3), 144-151. 115, 99-109.
[24]. http://www.fao.org/docrep/005/y3610e/y3610E3 [34]. Kuo, J. T., Hsieh, M. H., Lung, W. S., & She, N.
0.htm (2007). Using artificial neural network for
[25]. Jenkins, R. M., & Oglesby, R. T. (1982). The reservoir eutrophication prediction. Ecological
morphoedaphic index - concepts and practices: modelling, 200(1-2), 171-177.
The morphoedaphic index and reservoir fish [35]. Lae, R., Lek, S., & Moreau, J. (1999). Predicting
production. Transactions of the American fish yield of African lakes using neural
Fisheries Society, 111(2), 133-140. networks. Ecological modelling, 120(2-3), 325-
[26]. Jensen, M. C., & Meckling, W. H. (1994). The 335.
nature of man. Journal of applied corporate [36]. Leach, J. H., Dickie, L. M., Shuter, B. J.,
finance, 7(2), 4-19. Borgmann, U., Hyman, J., &Lysack, W. (1987).
[27]. Jhingran, A.G., 1986. Reservoir Fisheries A review of methods for prediction of potential
Management in India. Bulletin No. 45, Central fish production with application to the Great
Inland Fisheries Research Institute, Barrackpore. Lakes and Lake Winnipeg. Canadian Journal of
[28]. Jones, J. R., & Hoyer, M. V. (1982). Sportfish Fisheries and Aquatic Sciences, 44(S2), s471-
harvest predicted by summer chlorophyll-α s485.
concentration in Midwestern lakes and [37]. Lek, S., Delacoste, M., Baran, P., Dimopoulos, I.,
Lauga, J., &Aulagnier, S. (1996). Application of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


97
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

neural networks to modelling nonlinear [47]. Natarajan, A.V., 1977.Fifth Workshop of AICP
relationships in ecology. Ecological modelling, on Ecology and Fisheries of Freshwater
90(1), 39-52. Reservoirs. Ukai, 7-8 September, 1977.
[38]. Liang, Y., Melack, J. M., & Wang, J. (1981). [48]. Natarajan, A.V., 1979. Ecosystem oriented
Primary production and fish yields in Chinese approach for reservoir fisheries development in
ponds and lakes. Transactions of the American India. Souvenir on the ICAR Golden Jubilee
Fisheries Society, 110(3), 346-350. Year, CIFRI, Barrackpore, West Bengal, Part I,
[39]. Machena, C., & Fair, P. (1986). Comparison of 107-111.
fish yields from prediction models between [49]. Nissanka, C., Amarasinghe, U. S., & De Silva, S.
Lakes Tanganyika and Kariba. Hydrobiologia, S. (2000). Yield predictive models for Sri Lankan
137(1), 29-32. reservoir fisheries. Fisheries Management and
[40]. Maravelias, C. D., Haralabous, J., & Ecology, 7(5), 425-436.
Papaconstantinou, C. (2003). Predicting [50]. Northcote, T. G., & Larkin, P. A. (1956). Indices
demersal fish species distributions in the of productivity in British Columbia lakes.
Mediterranean Sea using artificial neural Journal of the Fisheries Board of Canada, 13(4),
networks. Marine ecology progress series, 255, 515-540.
249-258. [51]. Nürnberg, G. K. (1996). Trophic state of clear
[41]. Matuszek, J. E. (1978). Empirical predictions of and coloured, soft-and hard water lakes with
fish yields of large North American lakes. special consideration of nutrients, anoxia,
Transactions of the American Fisheries Society, phytoplankton and fish. Lake and Reservoir
107(3), 385-394. Management, 12(4), 432-447
[42]. McConnell, W. J., Lewis, S., & Olson, J. E. [52]. Oglesby, R. T. (1977). Relationships of fish yield
(1977). Gross photosynthesis as an estimator of to lake phytoplankton standing crop,
potential fish production. Transactions of the production, and morphoedaphic factors. Journal
American Fisheries Society, 106(5), 417-423. of the Fisheries Board of Canada, 34(12), 2271-
[43]. Mehrotra, S. N., &Jhingran, A. G. (1986). 2279.
Evaluation of limnological characteristics and [53]. Olden, J. D., & Jackson, D. A. (2001). Fish–
trophic status of Gulariya Reservoir. Proc. Nat. habitat relationships in lakes: gaining predictive
Acad. Sci. India, 53, 204-210 and explanatory insight by using artificial neural
[44]. Melack, J. M. (1976). Primary productivity and networks. Transactions of the American
fish yields in tropical lakes. Transactions of the Fisheries Society, 130(5), 878-897.
American Fisheries Society, 105(5), 575-580. [54]. Peters, R.H. (1986). The role of prediction in
[45]. Moreau, J., & De Silva, S. S. (1991). Predictive limnology.Limnol.Oceanogr. 31: 1143-1159.
fish yield models for lakes and reservoirs of the [55]. Quiros, R. (1990). Predictors of relative fish
Philippines, Sri Lanka and Thailand (No. 319). biomass in lakes and reservoirs of Argentina.
Food & Agriculture Org. Canadian Journal of Fisheries and Aquatic
[46]. Natarajan, A.V., 1976. Ecology and state of Sciences, 47(5), 928-939.
fishery development in some of the man made [56]. Raju, M. M., Srivastava, R. K., Bisht, D., Sharma,
reservoirs of India. In: Symposium on the H. C., & Kumar, A. (2011). Development of
development and utilisationof Inland Fishery artificial neural-network-based models for the
resources, Colombo, 27-29 October 1976. simulation of spring discharge. Advances in
Artificial Intelligence, 2011.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


98
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

[57]. Ramakrishnaiah, M., & Sarkar, S. K. (1982). S.H. Abidi, N.K. Thakur, R.S. Birader and L.
Plankton productivity in relation to certain Shenoy.(eds.) Vision on Indian Fisheries of 21st
hydrobiological factors in Konar reservoir Century. Central Institute of Fisheries
(Bihar). Journal of the Inland Fisheries Society Education, Bombay.
of India, 14(1), 58-68. [68]. Toews, D. R., & Griffith, J. S. (1979). Empirical
[58]. Ramakrishniah, M. (1990, October). estimates of potential fish yield for the Lake
Morphodrainage index, a fish yield predictor for Bangweulu system, Zambia, Central Africa.
Indian reservoirs. In 2nd Asian Reservoir Transactions of the American Fisheries Society,
Workshop, Hangzhou, PR China (pp. 15-19). 108(3), 241-252.
[59]. Rawson, D. S. (1952). Mean depth and the fish [69]. Unni, K.S., 1993.Limnology of Central Indian
production of large lakes. Ecology, 33(4), 513- lakes, reservoirs and ponds – a review. In:
521. Mishra, P.C. and Trivedy, R.K. (eds.). Ecology
[60]. Rempel, R. S., & Colby, P. J. (1991). A and pollution of Indianlakes and reservoirs.
statistically valid model of the morphoedaphic Ashish Publishing House, New Delhi.
index. Canadian Journal of Fisheries and Aquatic [70]. Vollenweider, R.A., 1969.A manual on methods
Sciences, 48(10), 1937-1943. for measuring primary productivity in aquatic
[61]. Rigler, F. H. (1982). Recognition of the possible: environment. IBP Hand Book No. 12. Blackwell
an advantage of empiricism in ecology. Canadian ScientificPub., U.K. 213 pp.
Journal of Fisheries and Aquatic Sciences, 39(9), [71]. Weimin, M., Liu, J., Li, Z., Xie, S., Vass, K.K.,
1323-1331. Sugunan, V.V., Pradhan, G.B., Amarasinghe,
[62]. Ryder, R. A. (1965). A method for estimating the U.S., Weerakoon, D.E.M. and Jutagate, T., 2006.
potential fish production of north‐temperate Status of Reservoir Fisheries in Five Asian
lakes. Transactions of the American Fisheries Countries.
Society, 94(3), 214-218 [72]. Welcomme, R. L. (1976). Approaches to
[63]. Ryder, R. A. (1982). The morphoedaphic resource evaluation and management in tropical
index—use, abuse, and fundamental concepts. inland waters. In FAO, Rome (Italy). Fisheries
Transactions of the American Fisheries Society, Dept. Indo-Pacific Fisheries Council.
111(2), 154-164. Symposium on the Development and Utilization
[64]. Ryder, R. A., Kerr, S. R., Loftus, K. H., &Regier, of Inland Fishery Resources. Colombo (Sri
H. A. (1974). The morphoedaphic index, a fish Lanka). 27 Oct 1976.
yield estimator—review and evaluation. Journal [73]. Welcomme, R.L.,1972. The inland waters of
of the Fisheries Board of Canada, 31(5), 663-688. Africa. FAO, CIFA Technical Paper, 1, 117pp.
[65]. Scardi, M. (1996). Artificial neural networks as Publ. FAO, Rome.
empirical models for estimating phytoplankton [74]. Wijeyaratne, M. J. S., &Amarasinghe, U. S.
production. Marine Ecology Progress Series, 139, (1987). Estimation of maximum sustainable fish
289-299. yield and stocking densities of fish fingerlings in
[66]. Schlesinger, D. A., & Regier, H. A. (1982). fresh water lakes and reservoirs.
Climatic and morphoedaphic indices of fish [75]. Wijeyaratne, M. R. S., &Amarasinghe, U. S.
yields from natural lakes. Transactions of the (1984). Estimations of maximum sustainable fish
American Fisheries Society, 111(2), 141-150. yields and stocking densities of inland reservoirs
[67]. Sinha, M. 1999. Vision of inland fisheries of of Sri Lanka. Journal of the National Aquatic
India of twenty first century. p. 154-168. In: Resources Agency, 31, 65-72.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


99
D. Karunakaran, M. Balakrishnan Int J S Res CSE & IT. January-February-2019 ; 5(1) : 88-100

[76]. Youngs, W. D., & Heimbuch, D. G. (1982).


Another consideration of the morphoedaphic
index. Transactions of the American Fisheries
Society, 111(2), 151-153.
[77]. Zhang, H., & Zimba, P. V. (2017). Analysing the
effects of estuarine freshwater fluxes on fish
abundance using artificial neural network
ensembles. Ecological modelling, 359, 103-116.

Cite this article as :

D. Karunakaran, M. Balakrishnan, "Prediction of Fish


Yields in Lakes and Reservoirs from simple Empirical
Models using Artificial Neural Network (ANN) : An
Review", International Journal of Scientific Research
in Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 88-100, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195110
Journal URL : http://ijsrcseit.com/CSEIT195110

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


100
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195111

Evolving Cryptographic Approach for Enhancing Security of Resource


Constrained Mobile Device Outsourced Data in Cloud Computing
K. Kaviya 1, K. K. Shanthini 1, Dr. M. Sujithra2
1MSc. Software Systems, Coimbatore Institute of Technology, Coimbatore, Tamil Nadu, India
2Assistant Professor, Coimbatore Institute of Technology, Coimbatore, Tamil Nadu, India

ABSTRACT

Mobile cloud computing in the increasing popularity among users of mobile device enables to store their data in
cloud. Security is the major concern when the sensitive information is stored and transferred across the internet.
It is essential to make sure that the data is secured and protected. In this paper, we discussed the problem of
privacy of data with reducing the resources usage. Moreover, mobile cloud computing has limitations in
resources such as power energy, processor, Memory and storage. Cryptography ensures the confidentiality,
authentication, availability, and integrity of the data. This is done through cryptographic algorithms such as
Data Encryption Standard (DES), Blowfish and Advanced Encryption Standard (AES). The experimental results
evaluated and compared the performance of the encryption algorithms. The performance metrics used are
encryption and decryption time, CPU and Memory utilization. Evaluation results showed a significant
improvement in reducing the resources, amongst all the techniques, choosing a suitable encryption algorithm
based on different parameters that are best fit to the future user requirements is considered.

Keywords : Mobile Cloud Computing, Blowfish, AES, Security, Privacy, Mobile Device.

I. INTRODUCTION

Mobile cloud computing uses cloud computing to


carry out the resource intensive tasks over the
internet to provide higher scope of functionality with
minimal pressure on mobile resources. Cloud
computing is a modern era computing technique that
has a greater future and bringing a lot of benefit to
Figure 1.1 Cloud Environments
the information technology. The major advantage of
using cloud is that it offers the cloud services to
There are various layered architectures available for
people by Pay-as-You-Go manner. Cloud computing
cloud computing to provide the services as a utility.
applies a virtualized platform with elastic resources
Cloud’s backbone layer consists of physical servers
on demand by provisioning hardware, software, and
and switches. The cloud service provider is
data sets dynamically.
responsible to run, manage, and upgrade cloud
hardware resources according to the requirements of
users. The backbone layer is also responsible to

CSEIT195111 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 101-106 ]
101
K. Kaviya Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 101-106

allocate hardware resources to users in an efficient, can/cannot access the data. The principle of
quick, and smooth way. The supervisor software confidentiality specifies that only the sender and
layer contains the system software to manage the the intended recipient should be able to access
cloud hardware resources. The system software the contents of a message.
permits application software to run and utilize  Integrity: The integrity mechanism ensures that
underlying resources in an efficient way.In the the contents of the message remain the same
context of mobile cloud security, cloud providers when it reaches the intended recipient as sent by
should ensure reliability and availability by the sender. It is defined as the rightness of data
integrating security technologies MCC satisfies stored in the cloud. The alterations between two
limitations of resource constrained devices by updates of a record violate the data integrity.
allocating cloud services for the purpose of  Security: In the traditional file systems data was
application executions and data storage. So, whenever stored within boundaries, but cloud data is stored
we talk about the security in MCC, it is necessary to outside the boundaries of an organization, say,
consider the security and privacy issues. and third-party storage using strong encryption
techniques.
To secure the mobile cloud environment, several  Authentication: Authentication mechanisms help
issues need to be considered regarding the data to establish proof of identities. This process
security, data integrity, data confidentiality, ensures that the origin of the message is correctly
authentication, authorization, network security, data identified.
violation issues etc. A secure framework is essential  Non- repudiation: Non-repudiation does not
to protect sensitive data of mobile users with minimal allow the sender of a message to refute the claim
performance degradation. The proposed scheme uses of not sending the message.
data encryption to protect sensitive data leakage.  Access Control: Access Control specifies and
When user’s data is transmitted to and stored on controls who can access the process.
cloud, it creates many possibilities of unauthorized  Availability: The principle of availability states
access of data either during transmission or from that resources should be available to authorized
storage devices. There are many cryptographic parties all the times.
algorithms to deal with security, but the selection
must be taken by considering security as well as III. CRYPTOGRAPHY
performance improvement, especially in the case of
resource constrained devices. Cryptography is a method of protecting information
and communications with codes so that only those
II. SECURITY ISSUES for whom the information is intended can read and
process it. Cryptography refers to secure information
 Data Storage: Cloud storage providers manage the and communication techniques derived from
data in multiple copies across many independent mathematical concepts and a set of rule-based
locations calculations called algorithms to transform messages
 Confidentiality: Confidentiality can be defined as in ways that are hard to decipher. Security has always
the sensitive data not being disclosed to been an important term in all the fields. As the paper
unauthorized process, devices and person. A is discussing on cloud storage systems, the proposed
cloud service provider knows where the user’s methodology considers the security required to the
public or private data is located and who data that is stored on the cloud. Cryptography has

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


102
K. Kaviya Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 101-106

come up as a solution to this security issue. block size. It includes 64 bits key that contains 56 bits
Cryptography is classified into symmetric are directly utilized by the algorithm as key bits and
cryptography, asymmetric cryptography and hashing. are randomly generated. The DES algorithm
processes the 64 bits input with an initial
Types of Cryptography permutation, 16 rounds of the key and the final
permutation. The DES was initially considered as a
 Secret Key Cryptography: When the same key is strong algorithm, but today the large amount of data
used for both encryption and decryption, DES, and short key length of DES limits its use.
Triple DES, AES, Blow Fish RC5 etc., may be the
examples of such encryption, then that Advanced Encryption Standard (AES)
mechanism is known as secret key cryptography.
 Public Key Cryptography: When two different The algorithm explains about by AES is a secret-key
keys are used, that is one key for encryption and algorithm which means of the same key is used for
another key for decryption, RSA, Elliptic Curve both encrypting and decrypting the data. This is one
etc., may be the examples of such encryption, reason why it has a comparably small number of
then that mechanism is known as public key rounds. AES encryption is fast and flexible. It can be
cryptography. implemented on various platforms especially in small
 Hash Algorithms where the input data (message) devices. All the operations in this algorithm involve
is recreated from the hash value (message complete bytes for effective implementation. Three
digest/digest) examples include: MD5, SHA, different key lengths such as 128, 192 and 256 block
MD2, MD4, MD6.SHA-256, SHA-512, SHA-1, size are supported by AES.
Whirlpool etc.
Table 4.1 Comparison of Symmetric Algorithms

Algorith DES AES BLOWFIS


ms/Metr H
ics
Structure Feistel Substitution- Feistel
Permutation

Key 56 bits 32-448 128,192 or


Length 256
Rounds 16 10, 12, 14 16

Block 64 bits 128 bits/192 0r 256 64 bits


Size
Figure 3.1 Types of cryptographic algorithms Through < AES < Blowfish High
put
Security Adequat Excellent Excellent
IV. PROPOSED METHODOLOGY e
Speed Slow Fast Fast

Data Encryption Standard (DES)


Blowfish

DES utilized the one secret key for encryption and


Blowfish is fast, license free, unpatented, freely
decryption process and key length is 56 bits and
available and alternative for existing encryption
performs the encryption of message using the 64 bits

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


103
K. Kaviya Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 101-106

algorithms. It uses the key length range up to 32-448- CPU Time = I * CPI * T, where I = number of
and 64-bits block. Blowfish algorithm employed 16 instructions in program, CPI = average cycles per
rounds for the encryption process. Each round instruction and T = clock cycle time.
contains a key dependent permutation and key- and CPU Time = I * CPI / R, where R = 1/T the clock rate,
data dependent substitution. Data Encryption T or R are usually published as performance measures
involves the iteration of a simple function of 16 times. for a processor, I requires special profiling software
Each round contains a key dependent permutation and CPI depends on many factors (including
and data dependent substitution. Sub key Generation memory).
involves converts the key up to 448 bits long to 4168
bits. Blowfish is very suitable algorithm for the Performance Calculation
platform of smart phones because of its high security
level and high speed. Seconds/Program = (Instructions/Program) x
(Clocks/Instruction) x (Seconds/Clock)
V. PERFORMANCE EVALUATION METRICS
Memory Consumption Calculation
CPU Time Calculation Total Memory - (Free + Buffers + Cached) = current
total memory usage

Table 5.1. Experimental Results: Time Comparison

Time Encryption Time Decryption


Algorithm File Size
(milliseconds) (milliseconds)
10 KB 300 299
15 KB 303 304
Blowfish 20 KB 310 310
25 KB 315 313
30 KB 320 317
10 KB 303 303
15 KB 306 307
AES 20 KB 316 313
25 KB 320 317
30 KB 325 320
10 KB 308 303
15 KB 309 309
DES 20 KB 314 317
25 KB 318 318
30 KB 321 324

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


104
K. Kaviya Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 101-106

Table 2. Experimental Results: CPU and Memory Consumption

E- CPU D- CPU E-Memory D-Memory


Algorithm File Size
(%) (%) (KB) (KB)
10 KB 21.1 16.2 17.6 8
15 KB 22.7 18.1 17.9 8.3
BlowFish 20 KB 17.4 18.1 17.9 17.4
25 KB 20.2 18.8 17.6 17.4
30 KB 20.8 24.2 17.9 17.6
10 KB 21.4 22.2 20.4 20.3
15 KB 25.3 23 20.5 20.3
AES 20 KB 20.4 34 20.5 20.6
25 KB 25 23.7 20.6 20.1
30 KB 25.5 31.7 20.7 20.7
10 KB 22.4 22.2 21.7 20.4
15 KB 23.3 23.6 21.9 21.3
DES 20 KB 24.4 25.2 22.9 22.8
25 KB 25.3 26.7 23.9 22.9
30 KB 25.7 27.7 24.6 22.5

Figure 5.1 Time Comparison for Encryption and Decryption

Figure 5.2 CPU and Memory comparisons for Encryption and Decryption

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


105
K. Kaviya Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 101-106

VI.CONCLUSION RSA algorithms over cloud storage", Journal of


Theoretical and Applied Information
The demonstration of results and discussion about Technology, Vol.80. No.2 ISSN: 1992-8645,
these algorithms are mainly focused on evaluation October 2015.
parameter like encryption and decryption time, [6]. HealeyM(2010) Why IT needs to push data
memory and CPU utilization which has more impact sharing efforts. Information Week. Source:
on the security, confidentiality, integrity, and http://www.informationweek.com/services/inte
reliability for secure communication. Security as gration/why-it-needs-to-push-data-sharing-
earlier discussed is the main challenge faced while effort/225700544. Accessed on Oct 2012
storing data in the cloud, the proposed system [7]. D. Elminaam, H. M. A. Kader, and M. M.
provides security for the data stored in the cloud Hadhoud, "Performance evaluation of
computing model. Based on the performance symmetric encryption algorithms,"
evaluation, the results of Blowfish, AES and DES International Journal of Computer Science and
provide more security based on the resources Network Security, vol. 8, no. 12, pp. 280-286,
availability. In future we can use encryption 2008.
techniques in such a way that it can consume less [8]. Shanthni KK., Kaviya K and Sujithra M.2018, A
time and minimum energy consumption. Survey on Cloud Computing: Data Security
Challenges and Their DefensiveMechanisms.
VII. REFERENCES Int J Recent Sci Res. 9(5), pp. 26497-26500.
DOI:
[1]. Gurpreet Kaur and Manish Mahajan (2013), http://dx.doi.org/10.24327/ijrsr.2018.0905.2070
Analyzing Data Security for Cloud Computing [9]. D. Salama, A. Minaam, H. M. Abdual-kader,
Using Cryptography Algorithms‖, International and M. M. Hadhoud, "Evaluating the effects of
Journal of Engineering Research and symmetric cryptography algorithms on power
Application, Vol.-3,782-786. consumption for different data types,"
[2]. Sujithra, M., G. Padmavathi, and Sathya International Journal of Network Security, vol.
Narayanan. Mobile Device Data Security: A 11, no. 2, pp. 78-87, 2010.
Cryptographic Approach by Outsourcing [10]. L Krithikashree ; S. Manisha ; M Sujithra,"
Mobile Data to Cloud. In: Procedia Computer Audit Cloud: Ensuring Data Integrity for
Science 47; 2015.p. 480-485. Mobile Devices in Cloud Storage",Published in:
[3]. Paresh D.Sharma, Prof. Hitesh Gupta(February 2018 9th International conference on
2014) An Implementation for Conserving Computing, Communication and Networking
Privacy based on Encryption Process to Secured Technologies (ICCCNT), IEEE , DOI:
Cloud Computing Environment‖ IJESRT 10.1109/ICCCNT.2018.8493963
Sharma, 3(2).
[4]. D. S. A. Elminaam, H. M. A. Kader, and M. M. Cite this article as :
Hadhoud, "Evaluating the performance of K. Kaviya, K. K. Shanthini, Dr. M. Sujithra, "Evolving
Cryptographic Approach for Enhancing Security of Resource
symmetric encryption algorithms," Constrained Mobile Device Outsourced Data in Cloud
International Journal of Computer Theory and Computing", International Journal of Scientific Research in
Computer Science, Engineering and Information Technology
Engineering, vol. 10, no. 3, pp. 343-351, 2009.
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp. 101-106,
[5]. M. Sujithra, and G. Padmavathi, "Ensuring January-February 2019. Available at doi :
Security on mobile device data with two phase https://doi.org/10.32628/CSEIT195111

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


106
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT183857

Semantic Web Search Engines : A Comparative Survey


Anita Kumari, Dr. Jawahar Thakur
Department of Computer Science, Himachal Pradesh University, Shimla, Himachal Pradesh, India

ABSTRACT

Search engines play important role in the success of the Web. Search engine helps the users to find the relevant
information on the internet. Due to many problems in traditional search engines has led to the development of
semantic web. Semantic web technologies are playing a crucial role in enhancing traditional search, as it work
to create machines readable data and focus on metadata. However, it will not replace traditional search engines.
In the environment of semantic web, search engine should be more useful and efficient for searching the
relevant web information. It is a way to increase the accuracy of information retrieval system. This is possible
because semantic web uses software agents; these agents collect the information, perform relevant transactions
and interact with physical devices. This paper includes the survey on the prevalent Semantic Search Engines
based on their advantages, working and disadvantages and presents a comparative study based on techniques,
type of results, crawling, and indexing.
Keywords : Semantic Web, Semantic Search Engines, Intelligent Search

I. INTRODUCTION which synchronies and arranges all the data over the
web in a disciplined manner.
The web is a medium for accessing a great variety of So moving from current web to semantic web
information stored in different parts of the world. particularly depends on enhancement of knowledge
Information is mostly in the form of unstructured and overcome the current web issues. Semantic search
data. As the data on the web grows at explosive rates, plays an important role because it intelligently
it has lead to several problems such as increased understands the content and give smart and relevant
difficulty of finding relevant information, extracting results according to the quires. Traditional search
useful knowledge. engines get the answers syntactically accurate but
larger in amount. However, semantic engines provide
Conventional web search engines like Google and accurate information in terms of well-defined
yahoo return thousands of results for users quires expressions. The main aim of semantic web is to lift
because they do not understand the exact meaning of the capacity of users and software agents to find
quires. They uses keyword based search approach, documents, information and knowledge from the web
thus new technology “Semantic Web” comes into and make the content to machine and human
existence, which works on the semantics (meaning) of readable. This is possible by adding the semantic layer
queries. Semantic web offers a smarter web services to the current web.

CSEIT195115 | Received : 10 Jan 2019 | Accepted : 22 Jan 2019 | January-February -2019 [ 5 (1) : 107-115 ]
107
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

The organization of this document is as follows. In pyramid. The research provides deep
Section 2 (Literature Survey), literature survey is Understanding of five main semantic search
explained with comparative study. In Section 3 engines based on comparative analysis that may
(Keyword Based Search), keyword based search help for future work for semantic web in general
engines and the limitations are discussed. Semantic and for semantic search engines in particular.
web, approaches to semantic web and few search 4) Nikhil CHITRE [8], analyses the semantic search
engines (Hakia, Falcons, Swoogle, DuckDuckGO and engines generation into the intelligent web. Their
Knigne) and comparison on these search engines are focus is on finding the difference between the
discussed in Section 4(Semantic Web Search) based traditional and semantic search engines. Explains
on few parameters. Section 5 (Conclusion) concludes the various techniques and advantages of few
the paper by providing a summary of all reported important semantic web search engines that has
work in semantic web search. been developed so far.
5) Yi Jin Zhuying Lin, Hongwei Lin [9], states that
II. LITERATURE SURVEY search engines plays important role in success of
web; search engine helps any internet user to
There is significant literature work that has been rapidly find relevant information. Architecture
done on semantic web area. Various authors and of semantic search engine is explained and how
researchers have contributed very well in this area. elements of semantic search engine are used in
performing the fundamental task of information
1) Vidhi Shah1, Akshat Shah [5], concluded that retrieval is explained. An improved algorithm
searching the internet today is a challenge and based on TFIDF algorithm is also proposed.
semantic search engine enhances the traditional 6) Kyumars Sheykh Esmaili, Hassan Abolhassani
web search engines. They discuss various [10], presents a categorization scheme for web
semantic search engines and present a search engines. For each category, components
comparison based on the approaches. are explained according to a proposed general
2) G.Anuradha, G.Sudeepthi et al [6], presents a brief architecture and various approaches are used in
survey of the existing literature regarding these components also represented. It should note
intelligent semantic web search engine that some of these engines are constructed
technologies. Also, review their characteristics through a portal and therefore they have not
respectively. They concluded that existing individual names.
techniques have limitations particularly in terms 7) Ranjana Jain, Neelam Duhan el at [11], states that
of time response, accuracy of results, importance web 2.0 search engines are unable to present
of the results and relevancy of results. An direct answers against the users fired query. Web
efficient semantic web search engines should 3.0 is semantic web using RDF format
meet out these challenges efficiently and organization information in more structure form,
compatibility with global standards of web which helps the semantic search engines to
technology. present results in a more direct way. Also include
3) Maliha Majid Qureshi, Bibi Assma [7], presents these semantic search engines, review their
the semantic search engines and broad survey to working, and compare them based on few factors.
analyze the semantic search engines has done. 8) Jyoti Chaurasia, Jagdish Raikwal[12], states that
Comparative analysis of the semantic search semantic web technology plays an crucial role in
engines has done based on Factors cited in the enhancing traditional web search. They proposed

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


108
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

a semantic search engine based on ontology. An difference between the keyword based and
algorithm is also introduced to overcome the semantic based search engines.
various challenges in the fields of unstructured 10) G.Madhu1 and Dr. A. Govardhan [14], presents a
data using semantic web technology and provide brief survey of the existing literature regarding
the complete description. intelligent semantic search technologies. Review
9) Ankita Malve , P.M. Chawan [13], concluded that their characteristics respectively. In addition, the
semantic based search engine have more issues within the reviewed intelligent semantic
advantages over keyword based search engine in search methods and engines are concluded based
terms of accuracy of getting results. Search on four perspectives differentiations between
process in semantic search engine is based on designers and users’ perceptions, static knowledge
semantics of queries. It provides assurance to user structure, low precision and high recall and lack
to get accurate and relevant results based on of experimental tests.
meaning of word. They also include the basic
TABLE I
COMPARISONS BASED ON LITERATURE SURVEY

Reference Semantic Search Language Objects Conclusion


no. Engines
[6] Falcons, Kosmix, RDF, OWL, OIL, Differentiate Combining semantic
Sunbot, Hakia, DAML+OIL,ODEX. between semantic search with current
Swoogle, XCDSearch, search engines and search engines to make
XSearch, Kngine, presents their content understandable
Powerset, advantages and by human as well as
DuckDuckGO. disadvantages. machine.
[7] Hakia, Sensebot, QDEX, NLP. Research is carried This paper presents
Powerset, Cognition, out to explore the pyramid based
Lexxe. different dimensions comparison. Values of
of semantic search pyramid includes: search
engines. environment, intrinsic
problem, query type, and
iterative and exploratory.
[8] DuckDuckGO, Hakia, RDF, DAML+OIL, Analyse semantic The semantic search
Kngine. OIL, OWL, search engines engine improves the
QDEXing. generation and the limitation of traditional
part of search search engine and
technologies and provide 100% relevant
intelligent web. result.
[9] Architecture of OWL, XML, TFIDF How fundament Shows that how the
semantic search engine algorithm. task of retrieval is fundamental elements of
and improved performed. semantic search engines
algorithm based on perform the information
TFIDF are presented. retrieval.
[10] Content meta semantic RDF,OWL. Differentiate Provide a categorization
engine, crawler based between ontology scheme for different type
ontology search engine, semantic search of search engines on
evolutionary search engines and semantic web.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


109
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

engines, semantic semantic search


association discovery eninge.
engine.
[11] Architecture of search RDF, OWL, CSV, Why we need web Comparative study of
engines, meta-search XML. search engines. Google, Falcons,
engines, semantic Swoogle, Hakia, SWSE.
search engines
(Swoogle, Falcons,
Hakia, SWSE).
[13] Google, DuckDuckGO OWL, RDF, XML, Comparison Search process in
HTML. between keyword semantic search engine is
based search based on semantic of
engines and query and provide
semantic search assurance to get 100%
engine. accurate data.
[12] Google, a new RDF, OWL, Propose a semantic Proposed an algorithm
algorithm proposed in DAML+OIL, OIL, search engine based with ontology’s to give
addition with Google. QDEXing. on domain better search experience.
ontology.

III. KEYWORD BASED SEARCH • Low value of Recall and current Precision
parameters.
Search Engines are used to find the information on • More network congestion because multiple
the internet and getting results in less time but it crawlers try to crawl the same web pages form
lacks the existence of the semantic structure so it web servers.
becomes difficult for the machine to understand the • Polysemy words:- engines do not differentiate
meaning of the terms and expressions provided by between the same spelled words having different
the users. As the results, search engines returns the meaning hence give irrelevant pages.
ambiguous data sets. The survey indicate that user do • Stemming: - also these search engines don not
not find the accurate results in the first set of URLs differentiate between the singular and plural
returned because of the increased size of the links on words. Hence may produce the confusion while
the web pages. Because sometime one word has searching a query.
several meanings and several words have same • Synonymy: - it is the problem of understanding
meaning, in this case user will not get the result what the different spelled words with the same
he wanted to search [13]. Semantic web is developed meaning and it produce the confusion while
to overcome the following limitations of the current searching.
web search:- • Lack of semantic understanding [6][13][12].

• These engines are based on the occurrence of IV. SEMANTIC WEB SEARCH
words in the documents, so it is difficult to get
A. Semantic Web
the relevant results.
• Lack of proper semantic structure according to
It is an extension to WWW and international
the representation of information.
certified standard by W3C, which aims in handling

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


110
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

mostly unstructured data. It describes the 3) The Third approach focus on natural language
relationship between things and properties of things. understanding.
The main idea of semantic web is to add the metadata 4) The Fourth approach makes use of ontology to
into the current web pages to improve the represent the knowledge about a Particular
description of online resources. domain and expand quires [6][8].

C. Semantic Based Search Engines


Tim Berners-Lee originally expressed the vision of
the semantic web as follows ―I have a dream for the The semantic search describes that search engine will
Web [in which computers] become capable of give most accurate search results possibly by
analysing all the data on the Web – the content, links, understanding searcher intent, query content and the
and transactions between people and computers. A relation between words. Alternatively, we can say
Semantic Web, which should make this possible, has that semantic search engines will understand the
yet to emerge, but when it does, machines talking to natural language as the way human can understand
machines will handle the day-to-day mechanisms of and it improves the search accuracy of query related
trade, bureaucracy and our daily lives. The intelligent data and delivers exact and smart results. Semantic
agent’s people have touted for ages will finally search engines uses ontology to get the relevant and
materialize [1]. It presents the easier way to share, accurate results in less time. Ontology provides the
reuse and discover the online resources [2]. Also related association between the contents. RDF
represent the information in the way that human and (Resources Description Framework) and OWL (Web
machines can understand using semantics [2]. Ontology Language) are used for semantic web
documents. Rather than using ranking algorithms
(PageRank to predict relevancy), Semantics are used.
It provides the guaranty that most relevant results
will retrieve depending upon the meaning and
relations of the words not a specific keyword. The
goal of semantic search is to use the meaning
(semantic) to improve the search experience of the
users.

In order to access structured data, a variety of search


engine has been introduced which works on
semantics(meaning) of data and provide the accurate
Figure 1. Architecture of Semantic Web results in less time as compare to the traditional
search engine. The broad categories of semantic
B. Approaches to Semantic Web
search engine according to their workings are [10]
There is not only one approach in the semantic
search. Every approach has a lot to contribute. These
1) Crawler Base Search Engines,
main four approaches are used; different search
2) Human powered directories,
engine may use one or more approaches’.
3) Hybrid Search Engines,
4) Meta Search Engines.
1) The First approach uses contextual analysis to
disambiguate the search queries.
Some of the important semantic search engines are:
2) The Second approach focus on reasoning.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


111
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

1) SWOOGLE: - It is a crawler-based indexing and 4) KNGINE: - It is intelligent search engine and the
retrieval system for semantic web documents result is divided into either web results or image
(SWD). It uses the crawler to discover the RDF results. They are preceded by the information
and HTML documents. It extracts metadata and about the “concepts”. For instance, attempting to
computes relations between documents. Swoogle find a town can be preceded by using statistics
is also a content-based search engine that used to about the metropolis, nearby points of interest,
analyse, discover, and index the knowledge from activities, climate and hotels. It provides the high
the web. It is based on the page rank algorithm degree of related data to the user’s queries. It
and OWL language [11]. analyze the main keywords from the users
2) HAKIA:- It is a meaning based search engine that queries and form the relations between them and
works on matching the meaning rather than return results. It is a question-answer engine that
keyword match. The searching capacity of this understands the questions and tries to answer
engine is not limited to single keyword; you can with the help of natural language processing and
search sentences, a phrases or questions. This machine learning [8][6].
engine gets the results using equivalent terms. 5) DUCKDUCKGO: -“The search engine that does
Hakia engine essentially builds around mainly not track you” it is a feature rich search engine
three evolving technologies i.e., Sense that gives many reasons to put Google away. It
Repository, Query Indexing Technique and focuses on maintaining user privacy and
Semantic Ranking algorithm [8][6][11]. therefore does not store any personalized search
3) FALCONS: - It is a keyword based ontology results. This engine guaranty that they does not
search engines. It is develop around the concept store users IP address, does not users login
sharing and ontology reusing. It display the information. This engine only uses cookies to
required information on snippets, therefore user maintain the privacy of the users. It is a semantic
does not need to explore the pages. This engine search technology based search engine that
provides the best user interface. This engine presents relevant match or concept match for
integrates these two terms i.e., Concept-Level- example if a user searches for a keyword that has
Search and Ontology- Level-Search by more than one connotation, it gives you the
recommending ontology and allows filtering the options to choose what you were originally
concepts using ontology’s. Through this simple searching for along with the disambiguate
process user can easily find the ontology’s that outcome. This engine has the most important
satisfy their need within a few seconds [11]. property that it provides users privacy which
Google does not give [8][6][10].
TABLE II
COMPARISONS BETWEEN DIFFERENT SEMANTIC SEARCH ENGINES

SWOOGLE FALCONS HAKIA KNGINE DUCKDUCKGO

APPROACH Content based Keyword Content match, knowledge cluster search,


approach. Based NLP based , NLP.
approach. Statistical

TECHNIQUE 1.Rational Surfer Object OntoSem(Sense knowledge- Cluster search,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


112
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

Model, Ranked based Repository), Based- NLP


2.It uses crawler on QDEX( Query Approach,
based indexing combination Indexing Statistical
semantic search of their Method), Approach.
technique that relevance to Semantic
search ontology the query and Algorithm.
and instance data. their
popularity.
Uses cosine
similarity
measures
between the
query and
virtual
documents of
objects.

TYPE OF Entity(OWL,RDF) Entity(gives Documents Web Summaries(provide


RESULTS the result in (Links and free Results or Privacy)
snippet itself) Text) Image
Results
INDEXING Swangling Inverted Query Provide Instant
TECHNIQUE technique. Indexing detection and Answers
extraction
IT CRAWEL RDF Data RDF Data HTML Web Web pages RDF Data
Pages
FEATURES 1.Find exact 1.It integrate 1.Easily 1.Display 1.Zero click
ontology’s, 2.Finds concepts level identify results in information,
exact instance data search and information the form of 2.Provide privacy,
structure of the ontology level from credible image, 3.Does not record
semantic web. search, 2.User sites, 2.It is users information,
can find 2.Save time, multi- 4.Produces the result
results 3.Excellent lingual, based on many
quickly. response time. 3.It allows source and also from
the user to its own crawler.
search in
parallel
manner.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


113
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

LIMITATIONS 1.Poor indexing of 1.This engine 1.URL 1.Does not 1.It provides itself an
documents, does not rank Canonization, provides option of privacy but
2.Poor response the objects 2.Privacy silent mode they are just as
time for find according to Session IDS, option. opaque as Google is,
queries. query. 3.Does not since they don’t
index open source their
everything, code, how do you
4.Need other know they are not
search engines tracking you?
too,
5.Contains
virtual content
and dynamic
content.

V. CONCLUSION [2]. Shadbolt, Nigel, Tim Berners-Lee, and Wendy


Hall. "The semantic web revisited." IEEE
This paper includes the survey on the prevalent intelligent systems 21.3 (2006): 96-101.
Semantic Search Engines based on their advantages, [3]. Berendt, Bettina, Andreas Hotho, and Gerd
working and disadvantages and presents a Stumme. "Towards semantic web mining."
comparative study based on techniques, type of International Semantic Web Conference.
results, crawling, and indexing. It concluded that Springer, Berlin, Heidelberg, 2002.
semantic based search has more advantages over the [4]. Al-Feel, Haytham T., Magdy Koutb, and Hoda
keyword-based search and provide the relevant and Suoror. "semantic web on Scope: A New
accurate results in first URL. Semantic Web organizes Architectural Model for the semantic web."
the information in more structure form by using RDF (2008).
or OWL format, which helps the semantic search [5]. Shah, Vidhi, Akshat Shah, and Asst Prof
engines to, presents the results in way that is more Khushali Deulkar. "Comparative Study Of
accurate. Semantic Search Engines." International
Journal Of Engineering And Computer Science
Future work can be extended to the development of 4.11 (2015).
new semantic search engine technologies that [6]. Sudeepthi, G., G. Anuradha, and M. Surendra
overcomes their limitations. Prasad Babu. "A survey on semantic web search
engine." International Journal of Computer
VI. REFERENCES Science Issues (IJCSI) 9.2 (2012): 241.
[7]. Qureshi, Maliha Majid, Bibi Asma, and Hikmat
[1]. Berners-Lee, Tim, James Hendler, and Ora Ullah Khan. "Comparative analysis of semantic
Lassila. "The semantic web." Scientific american search engines based on requirement space
284.5 (2001): 34-43. pyramid." International Journal of Future
Computer and Communication 2.6 (2013): 562.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


114
Anita Kumari, Dr. Jawahar Thakur Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 107-115

[8]. Chitre, Nikhil. "Semantic Web Search Engine." Cite this article as :
International Journal 4.7 (2016).
[9]. Jin, Yi, Zhuying Lin, and Hongwei Lin. "The Anita Kumari, Dr. Jawahar Thakur, "Semantic Web
research of search engine based on semantic Search Engines : A Comparative Survey",
web." International Symposium on Intelligent International Journal of Scientific Research in
Information Technology Application Computer Science, Engineering and Information
Workshops. IEEE, 2008. Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
[10]. Esmaili, Kyumars Sheykh, and Hassan 5 Issue 1, pp. 107-115, January-February 2019.
Abolhassani. "A categorization scheme for Available at doi :
semantic web search engines." Computer https://doi.org/10.32628/CSEIT195115
Systems and Applications, 2006. IEEE Journal URL : http://ijsrcseit.com/CSEIT195115
International Conference on.. IEEE, 2006.
[11]. Jain, Ranjna, Neelam Duhan, and A. K. Sharma.
"Comparative study on semantic search
engines." International Journal of Computer
Applications, ISSN (2015): 0975-8887.
[12]. Chaurasia, Jyoti, and Jagdish Raikwal.
"Enhanced Semantic Web Retrieval Technique:
Domain Ontology." International Journal of
Computer Applications 119.16 (2015).
[13]. Malve, Ankita, and P. M. Chawan. "A
Comparative Study of Keyword and Semantic
based Search Engine." (2015): 11156-11161.
[14]. Madhu, G., Dr A. Govardhan, and Dr TV
Rajinikanth. "Intelligent semantic web search
engines: a brief survey." arXiv preprint
arXiv:1102.0831 (2011).

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


115
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195121

Image Dehazing Technique Based On DWT Decomposition and


Intensity Retinex Algorithm
1Sunita Shukla, 2Prof. Silky Pareyani
M. Tech Scholar, Gyan Ganga College of Technology Jabalpur, Madhya Pradesh, India
1

2Assistant Professor, Gyan Ganga College of Technology Jabalpur, Madhya Pradesh, India

ABSTRACT

Conventional designs use multiple image or single image to deal with haze removal. The presented paper uses
median filer with modified co-efficient (16 adjacent pixel median) and estimate the transmission map and
remove haze from a single input image. The median filter prior(co-efficient) is developed based on the idea that
the outdoor visibility of images taken under hazy weather conditions seriously reduced when the distance
increases. The thickness of the haze can be estimated effectively and a haze-free image can be recovered by
adopting the median filter prior and the new haze imaging model. Our method is stable to image local regions
containing objects in different depths. Our experiments showed that the proposed method achieved better
results than several state-of-the-art methods, and it can be implemented very quickly. Our method due to its
fast speed and the good visual effect is suitable for real-time applications. This work confirms that estimating
the transmission map using the distance information instead the color information is a crucial point in image
enhancement and especially single image haze removal.
Keywords : AF: Adaptive filter, AHE: Adaptive Histogram Equalization, LOE: Lightness Order Error

I. INTRODUCTION captured under bad visibility often has a contrast and


many of its features are difficult to see.
Restoration of hazy image is an important issue in
outdoor vision system. Image enhancement and II. METHODOLOGY
dehazing remain a challenging problem as well as an
important task in image processing. Image In this paper, we proposed a new method for single
enhancement is really an important issue in image image dehazing using the NAM (Non-symmetry and
processing applications such as digital photography, Anti-packing Model)-based decomposition and
medical image analysis, remote sensing and scientific contextual regularization. We estimated the airlight
visualization [1]. Image dehazing and enhancement is by decomposing the image using non-symmetry and
the process by which the appearance and the anti-packing model [11] to eliminate the false
visibility of an image are improved such that the estimation at the boundary or the over-bright object.
obtained image is suitable for visual perception of Then, the scene transmission was calculated using the
human beings or for machine analysis. It is useful not combination of the boundary constraints, the
only from an aesthetic point of view but also helps in contextual regularization and the optimization
image analysis and object recognition, etc. Image proposed by Meng et al. [12]. The proposed method
had better color visual and haze-free image when it

CSEIT195121 | Received : 10 Jan 2019 | Accepted : 23 Jan 2019 | January-February -2019 [ 5 (1) : 116-122 ]
116
Sunita Shukla, Prof. Silky Pareyani Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 116-122

comes to the image dehazing problem, a dehazing Then the haze-free image could be estimated:
model called Atmospheric Scattering Model is widely ( )
( )
used in ( ( ) )
I (x) =J (x)t(x) + A(1-t(x))
where is usually a small constant (0.001) in order to
where I(x) represents the hazy image, and J(x) avoid division by zero.
denotes the origin haze-free image(also called scene
radiance). A is the airlight which is the global light in Proposed an algorithm which combines the merits of
the atmosphere while t(x) denotes scene transmission transform color space algorithm and wavelet
function (0<t(x)<1). t(x) usually correlates with the transform algorithm.
scene depth. When the observed object is further
away from the camera, it’s more affected by the haze. First, the RGB image is converted to the HSI color
As a result, t(x) is closer to zero when the object space, Then histogram equalization is applied to
depth in the scene is greater. In the dehazing process, intensity component I to enhance the contrast of
the goal is to estimate the A (airlight) and construct image and the segmentation
the most optimal transmission function t(x).

Input Median R
Original Filter G
Image
B

Histogram Equalization

RGB to RGB Stretching


HSI

Hue Saturation Intensity

DWT

HH HL LH LL

Retinex
Algorithm

IDWT

Dehazed
Image

Figure 1. Flow diagram of proposed work

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


117
Sunita Shukla, Prof. Silky Pareyani Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 116-122

Exponential enhancement algorithm is applied to Let input image is x which is a RGB image
saturation component S, and then the intensity First Histogram Equalization need to be done
component I is divided into high and frequency sub-
( )
bands with wavelet transform and then Retinex
algorithm is applied to the low-frequency sub-band
( ) ( )
to reduce the effect of haze and adjust image ( )
√( ) ( )( )
luminance, a fuzzy enhancement algorithm is applied
to high-frequency sub-band to achieve the
enhancement and de-noising for the image details. (
( )
Finally, utilize the inverse wavelet transform to Let ‘img’ is the HSI image and its intensity block is of
reconstruct the I component and then the 3x3 is as below, and the intensity need to enhance
reconstructed component I will be synthesized with with K coefficient
H and S components to get a clear RGB image, and
the proposed algorithm is represented by following
flowchart [ ]

Table 3.1 Histogram equalization algorithm

Pixel intensity A B C D e

Pixel value f1 f2 f3 f4 f5

Probability f1/9 f2/9 f3/9 f4/9 f5/9

Cumulative F1/9
probability

CP*k K*F1/9
{ } { } { } { }

Floor Na=
rounding floor(K*F1/9) [{ } ] [{ } ]

[{ } ] [{ } ]

[ ] ( ) ∑ ( ) ( )

( ) ∑ ( ) ( )
is the intensity frame of HSI image of MxN DWT
applied on ‘I’
Table 1 below shows the symlet type 4 HPF and LPF ( )
filter coefficients. Proposed work use ‘sym4’ type ( )
wavelet for decomposition of Cover image, figure 1
below shows HPF and LPF decomposition using
DWT.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


118
Sunita Shukla, Prof. Silky Pareyani Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 116-122

Retinex let (x,y) are the pixels coordinates of ‘p’ in accelerate the processing speed, a fast single image
space domain then W is the reflection component dehazing method has to be developed in near future
and Z illumination component then which will be based on the atmospheric scattering
model and gray projection. our future method will
( ) ( ) ( ) have transmission map is estimated roughly using
minimum filtering which is based on dark channel
( ) ∑ ∑ ( ) ( )
prior and refined with fast average filtering. The gray
( ) projection method is adopted to obtain the
( )
atmospheric light. The subsection mechanism is
designed to avoid the high light of sky region in
Where c is Gaussian scale and is a constant that
recovered image, and Weber-Fechner Law is
makes F(x, y) equal to 1.
employed for color compensation in restored image. a
( ) ( )* ( ) ( )+
optimized approach is planned to be develop in near
( ) ( ( ))
future, in which the transmission map is roughly
( ( )) ( ( )
obtained with DCP and processed with a fast average
( ))
( ) ( ) filtering to improve the running speed. By gray
projection, the atmospheric light is obtained. And
Let (u,v) are the pixels coordinates of ‘q’ in space after the restoration of haze-free image, the intensity
domain compensation is employed to increase the visual
effect. The designed algorithm must achieve satisfied
( ) result with a computation cost.
( )

III. RESULTS AND DISCUSSION

( ) ( ( ) ) LOE measure is based on the lightness order error


( ) ( )( ) between original image X and enhanced image Y
( ) ( ) .The LOE measure is defined as
Mq is the final enhanced high frequency component
q
∑∑

∑ { ( ) ( ) }
is the relative order difference

Outdoor images are degraded by haze. The degraded


images have image contrast. This seriously affects ∑∑( ( ) ( ))
many computer vision applications such as video
surveillance systems and vehicle visual systems.
Dehazing methods have attracted much interest in The lightness L of an image is the maximum of its
recent years. Generally, it is difficult to prevent the three color channel.
reflection of haze. To restore the image contrast, ( )( )

dehazing methods need to be performed. To


Simulation is been taken for Crop image

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


119
Sunita Shukla, Prof. Silky Pareyani Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 116-122

Figure 2. GUI for the test image Crop

Figure 5. Test image Crop with histogram


enhancement

Figure 3. Histogram analysis of test image Crop

Figure 6. Enhanced test image Crop

Tables below shows the observed results of the


proposed simulation for test images 'house’,
‘building’, ‘Man’, Crop’ and 'Ground’.

Table 1 Observed results of MSE, PSNR and LOE

SN Test MSE PSNR LOE


Image
1 House 0.237564 48.97 0.890866
Figure 4. Test image Crop RGB to HSI with haze 2 Building 0.223472 48.4388 0.838019
enhancement 3 Man 0.300008 50.997 1.12503
4 Crop 0.135195 44.0736 0.506982
5 Ground 0.202547 47.5849 0.759551
Average 0.19652 47.234 0.862502

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


120
Sunita Shukla, Prof. Silky Pareyani Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 116-122

Table 2 LOE comparison for Building image changing when the homogeneity of RGB channels
have a big difference. The results obtains for the
Author/ Journal/Year LOE for Image of thesis work are better in terms of LOE and SNR then
Building available works.
Yuhei Kudo / IEEE 2018 1.932
Proposed 0.838019
V. REFERENCES

Table 3 Mean Square Error comparison of Crop


[1]. Yuhei Kudo, Akira Kubota,Image Dehazing
Image
Method by Fusing Weighted Near-Infrared
Image, 978-1-5386-2615-3/18/ ©2018 IEEE
Work MSE for image of
[2]. Elisee A Kponou, Zhengning Wang, Ping wei,
Crop
Efficient Real-time Single Image Dehazing
Proposed 0.135195
Based on Color Cube Constraint, 2017 IEEE
Yunping Zheng/ IEEE 2017 0.1393 2nd International Conference on Signal and
Image Processing, 978-1-5386-0969-
9/17/©2017 IEEE
IV.CONCLUSION [3]. Yunping Zheng, Zhenfeng Xie, Changting Cai,
Single Image Dehazing Using Non-symmetry
The proposed work studies different types of methods and Anti-packing Model Based Decomposition
and technologies that have been used for image and Contextual Regularization, 2017 13th
dehazing and observed that the low contrast and International Conference on Natural
noise remains a barrier to visually pleasing images in Computation, Fuzzy Systems and Knowledge
dehazing conditions. In that condition, to find out a Discovery (ICNC-FSKD 2017), 978-1-5386-
more accuracy in image enhancement process there is 2165-3/17/©2017 IEEE
need to detect and measure the intensity level of [4]. Takuya Mikami, Daisuke Sugimura and
individual pixel channel as well as have to present an Takayuki Hamamo to, CAPTURING COLOR
appropriate enhancement factor for enhancement AND NEAR-INFRARED IMAGESWITH
purpose, so that effective and efficient image DIFFERENT EXPOSURE TIMES FOR IMAGE
enhancement process will be created. A new method ENHANCEMENT UNDER EXTREMELY
for image dehazing based on DWT Decomposition HAZE SCENE, 978-1-4799-5751-4/14/2014
and intensity Retinex. The air-light of image is IEEE
estimated by decomposing the image using non- [5]. Xiankun Sun, Huijie Liu, Shiqian Wu, Zhijun
symmetry and anti-packing model to refine the Fang, Chengfan Li, and Jingyuan Yin, Haze
illumination value. Next, the scene transmission Image Enhancement Based on Guided Image
function is calculated using the combination of the Filtering in Gradient Domain, Hindawi,
boundary constraints and the contextual International Journal of Digital Multimedia
regularization. The proposed method produces high Broadcasting, Volume 2017, Article ID
quality dehazed picture in most cases and decrease 9029315, 13 pages,
artifact. What’s more, the tone of the image is https://doi.org/10.1155/2017/9029315
natural. But there’s still some points needing to be [6]. Pixel Binning Yoonjong Yoo , Jaehyun Im and
improve such as the time consuming and the tone’s Joonki Paik, Haze Image Enhancement Using

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


121
Sunita Shukla, Prof. Silky Pareyani Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 116-122

Adaptive Digital, Sensors 2015, 15, 14917-


14931; doi:10.3390/s150714917, ISSN 1424-
8220, www.mdpi.com/journal/sensors
[7]. Zhenqiang Ying, Ge Li, Yurui Ren, Ronggang
Wang, and Wenmin Wang, A New Haze
Image Enhancement Algorithm using Camera
Response Model National Science Foundation
of China (No.U1611461), Shenzhen Peacock
Plan (20130408183003656), and Science and
Technology Planning Project of Guangdong
Province, China (No. 2014B090910001 and No.
2014B010117007).
[8]. Zhuang Feng, Haze Image Enhancement by
Refining Illumination Map with Self-guided
Filtering, 2017 IEEE International Conference
on Big Knowledge, 978-1-5386-3120-1/17 2017
IEEE, DOI 10.1109/ICBK.2017.37

Cite this article as :

Sunita Shukla, Prof. Silky Pareyani, "Image Dehazing


Technique Based On DWT Decomposition and
Intensity Retinex Algorithm", International Journal
of Scientific Research in Computer Science,
Engineering and Information Technology
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
116-122, January-February 2019. Available at doi
: https://doi.org/10.32628/CSEIT195121
Journal URL : http://ijsrcseit.com/CSEIT195121

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


122
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195116
Data Extraction in Malignant Environments
Amith Mandal1, Dr. R. P. Ramkumar2
1M. Tech Scholar, Department of CSE, Malla Reddy Engineering College (A), Telangana, India
2Professor, Department of CSE, Malla Reddy Engineering College (A) Telangana, India

ABSTRACT

Deliberate or inadvertent escape of classified data is undoubtedly one among the premier extreme security
dangers that associations look inside the advanced period. The risk right now stretches out to the private lives:
an abundance of private information is out there to interpersonal organizations and great telephone providers
and is in a roundabout way exchanged to undependable outsider and fourth gathering applications. amid this
work, a bland data genealogy structure LIME (Data Lineage in the Malicious Environment), is utilized for data
stream over different elements that take two trademark, guideline parts (i.e., proprietor and buyer). It
characterizes the exact security ensures required by such a data heredity component toward recognizable proof
of a blameworthy substance, and decide the disentangling non-denial and trustworthiness suspicions. At that
point create and break down a totally special dependable data exchange convention between two substances
among noxious surroundings by expanding upon unmindful exchange, solid watermarking, and mark natives.
At long last, an exploratory investigation to exhibit the helpfulness of convention what's more, apply structure
to the important data run projections of learning outsourcing and the informal organizations. In general,
LIME(Data Lineage in the Malicious Environment),lineage structure for information exchange, to be a key
advance towards accomplishing responsibility by plan.
Keywords: Classified Data, LIME, Trademark, Malicious Environment, Micro Benchmarking, Pairing Based
Cryptography, GNU

I. INTRODUCTION In the past few decades there has been an exponential


growth in technology in which the major part was
Data Extraction avoidance is the type of solutions the revolution of computers. This revolution has
which can help an institute to apply control for the taken upon every aspect both resourcefully and
prevention of unwanted, unintentional or leakage of individually, this growth has surely changed our lives
specific information to illegal entities inside or by making it much easier. In the process of this
outside the institute. Here specific information may growth there has been a severe risk which everyone
refer to the institute’s internal processing of may not be conscious of which is data extraction.
documents, tactical business plans, academic property, Data extraction is basically a scenario where any
economic statements, security policy, network intruder will access our personal data and make the
architecture, blueprints etc. extracted data public or the intruder might use it for
any other intentions. Today we have a huge access of
smartphones and gadgets which makes data

CSEIT195116 | Received : 15 Jan 2019 | Accepted : 25 Jan 2019 | January-February -2019 [ 5 (1) : 123-130 ]
123
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

extraction much easy as the impostor can now easily outcome .Amid this distribution the prerequisite for
transfer the data. general answerability instrument in learning
exchange is demonstrated which may specially go
We as a consumer find many apps on internet and we with confirmable measures and control the condemn
knowingly handle our data to the respective blameworthy. This philosophy of identification is
organizations but is our data really safe in the hands named as learning root. Endeavors are made by
of the organization, what if any employee of the numerous associations to utilize learning source
fellow organization decides to make a step that ruins inside the type of watermarking or including of false
the organizations name, what if any employee information or data in specially appointed way has
wantedly access your data and misuses it. The answer not been that flourishing.
to that problem is represented in the following work
which will determine the guilty entity with proof II. EXISTING SYSTEM
and held for explanation if any data extraction occurs.
There are few safety mechanisms such associate In the digital era, data outflow during intentional
secret writing of valuable data, however it's a exposures, or un-intentional damage by untruthful
proactive measure that cannot verify the info leak staff and harmful external entities, always provide
once it happens. Once the intruder decrypts the serious threats to organizations. it's not laborious to
information then there's no potential live which will believe that this is simply the edge of an iceberg, as in
prevent the info leak of the personal data and most cases information of knowledge outflow go
business data by any desired measures. Following are unreported thanks to concern of loss of client
few measures that we often see and there are a confidence or restraining penalties huge amount of
number of situations where the event is determined. digital data can be derived at nearly no price and
might be unfold through the net in terribly short
1. We should take facebook for instance wherever we time to boot, the risk of obtaining or getting caught
can see assortment of outsiders that approaches our for information outflow or the extraction is very low,
insight, if any outsider is experienced in learning spill as there square measure presently nearly no
it winds up impractical to call attention to or see that answerableness mechanisms. For these reasons, the
wherever extremely the data was misused as there matter of information loss has reached a replacement
square measure a few outsiders advertised. dimension these days.
2. Consultancies have all the valuable data of
representatives/applicants caning to seek out work 2.1 Disadvantages
which outsider contracts another outsider then all Even with access management mechanisms,
the three organizations will right now approach the wherever access to sensitive information is restricted,
individual information then it may turn out to be a malicious approved user will publish the valuable
horrendously difficult to undeniably relate one the information as soon as he receives it. Primitives like
three partnerships for learning spill, in each the coding provide protection solely as long because the
circumstances witnesses can either don't target the info of interest is encrypted, however once the
exact learning safety or the intruder may attentively recipient decrypts a message, nothing will stop him
reveal classified data which cannot be tied to any of from publication the decrypted content. Therefore it
the handlers. At times there comes experience the looks not possible to stop information outpouring
explanatory procedures however it's horribly proactively.
expensive and it also doesn't create a terribly sensible

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


124
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

2.2 Proposed System The key advantage of this representation is that it


In this work, the requirement for a general applies answerableness right from design; i.e., it
answerability mechanism in information transfers is drives the system designer to contemplate doable
got wind. This answerability are often directly information leakages and also the corresponding
related to demonstrably detective work a answerableness constraints at the look stage. This
transmission history of information across multiple helps to beat the prevailing state of affairs wherever
entities ranging from its origin. This can be called most lineage mechanisms ar applied solely once an
information place of origin, information lineage or escape went on.
supply tracing.
III. MODULES
Here the matter is dignified of practically associating
the problem to the extraction, and work on the 3.1 Data Owner
information lineage methodologies to unravel the In this module owner uploads their pictures with
matter of knowledge leak in numerous leak situations. their contents data to the net server. For the safety
This system defines LIME, a generic information purpose to the information, owner assigns the digital
lineage framework for information flow across sign and stores it inside the internet and in addition
multiple entities within the malicious atmosphere. performs the subsequent operations to transfer image
with its digital signature with supported title,
Here it's observe that entities in information flows description
assume one amongst 2 roles: owner or shopper and Below are the few operations an Owner can perform
introduce an extra role within the sort of auditor, 1 .Upload image or a file,
whose task is to see a problem for any data extraction, 2. Verify image or file details
and summarize the accurate properties for 3. View Consumer requests and
communication between these roles.In the process 4. Deleted image details
Associate in nursing non-mandatory non-repudiation
assumption created between two house owners, 3.2 Web Server
Associate in nursing and non-mandatory trust The Web service supplier manages an internet to
(honesty) assumption created by the auditor supply information storage service. And performs the
concerning the house owners. subsequent operations reminiscent of Storing all
image files with their signature, read all image Files
The second contribution, associate in nursing with its details, view all image comments, and view
responsible for information transfer protocol to all information house owners and Users, read all
provably and securely transfer valuable information attackers.
between the two parties is bestowed. It deals with
associate in nursing and untrusted sender, associate in A web server will be monitoring the system held
nursing and untrusted receiver situation related to responsible for a secure login including analyzing of
data transfer between any two customers, the all the data that has been transferred between two
protocol we use is an interesting combination of the entities
sturdy watermarking, oblivious transfer, and
signature primitives. The basic operations that are performed by the web
server are as follows
2.2.1 Advantages of Proposed System 1. View Owner and consumer Details

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


125
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

2. View all the files details usage of the Cox calculation is utilized for vigorous
3. View the attacked files picture and set the negative factor, which decides the
4. View Download history quality of the watermark, to a estimation of 0.1.Now
3.3 Auditor execute the explore different avenues regarding
Auditor has capability to access, manage or monitor diverse parameters to examine the execution. The
the transferred knowledge or data beneath the sender and beneficiary piece of the convention will
entrustment of information of owner, auditor has be executed in a similar program, i.e., don't break
experience and capabilities that a standard user down system sending, yet just computational
doesn't have, for occasionally auditing the outsourced execution.
knowledge. This audit service is considerably
necessary for digital forensics and credibleness in 4.2 PBC (Pairing Based Cryptography)
Web and performs the subsequent operations
1.View image/file details Blending based cryptography is a generally youthful
2. View Attacked files territory of cryptography that rotates around a
3. Block or unblock an attacker certain capacity with exceptional properties.
4. View messages from a consumer
The PBC (Pairing-Based-Cryptography) library is a
3.4 Data Consumer free C library based on the GMP library that plays
The net user of an Organization encompasses a great out the numeric activities and fundamental matching
quantity of data to be stored in servers and have the based on cryptosystems.
permissions to access and manipulate and keep image
and its data. The PBC library is planned to be the foundation of
executions of matching based cryptosystems, in this
The consumer can search information and accessing manner speed and portability are key objectives. It
the image data if he's approved and performs the provides schedules by comparing it to the elliptic
subsequent operations bend era, elliptic bend number juggling and blending
1.View all the files calculation. Due to the GMP library, in spite of being
2. Download a file composed in C, pairings times are shabby. On a 1GHz
3. Send download request Pentium III:
4. Send message to the auditor
• Fastest matching is 11ms
IV. TECHNIQUES USED • Short matching is 31ms

4.1 Micro Benchmarking The API is sufficiently dynamic that the PBC library
The execution of the convention is done in as a will be used despite of the fact that the associated
proof-of-idea and to determine its result. For the researcher has exclusively a rudimentary
unaware exchange sub-convention is executed to the comprehension of pairings. There’s no must be
convention utilizing the PBC library, which itself constrained to get some answers concerning elliptic
makes utilization of the GMP library. For marks BLS bends or a significant measure of collection
plot is executed, likewise utilizing the PBC library. hypothesis. (The base necessity is a few information
For symmetric encryption a usage of AES from the of cyclic gatherings and properties of the blending.)
Crypto++ library is utilized. For watermarking a

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


126
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

The PBC library can likewise be utilized to construct concerning once every year. Since variant about six,
customary cryptosystems. GMP is appropriated beneath the twin licenses,
wildebeest LGPL v3 and wildebeest GPL v2. These
licenses make the library freed to utilize, share, and
enhance, and allow you to kick the bucket the
outcome. The wildebeest licenses offer flexibilities,
Fig 1. Two party one round key agreement protocol however conjointly set firm limitations on the use
with without non programs. GMP's primary target
3.3 The GNU Multiple Precision Arithmatic Library stages are Unix-type frameworks, reminiscent of
GMP is attentively designed to be fast for small GNU/Linux, Solaris, HP-UX, waterproof OS
operands as well as for large operands. The speed is X/Darwin, BSD, AIX, and so on. It is additionally
accumulated by use of adequate words as basic celebrated internationally to figure on Windows in
arithmatic type, by use of quick algorithms, with each 32-bit furthermore, 64-bit mode.
immensely advanced assembly code for the most
familiar inner loops for a lot of CPUs, by a general 3.4 BLS Scheme
insistence on speed. In cryptography, the Boneh– Lynn– Shacham (BLS)
signature topic allows a client to confirm that an
GMP could be a free library for subjective accuracy endorser is bona fide. The subject uses a straight
number juggling, in task on marked whole numbers, blending for confirmation, Associate in nursing
objective numbers, and gliding point numbers. marks square measure segments of an elliptic bend
there's no sensible cutoff to the accuracy aside from group. working in Associate in Nursing elliptic bend
those verifiable by the available memory inside the bunch gives some safeguard against list math assaults
machine GMP keeps running on. GMP highlights a (with the proviso that such assaults square measure
made arrangement of capacities, and furthermore the still achievable inside the objective bunch of the
capacities have a regular interface. The fundamental matching), allowing shorter marks than FDH marks
target applications for GMP are cryptography for an indistinguishable level of security. Marks made
applications and investigation, web security by the BLS signature subject square measure
applications, unadulterated science frameworks, normally talked as short marks, BLS short marks, or
process unadulterated arithmetic examination, and so just BLS marks. The mark subject is undeniably
forth. secure.

GMP is thoroughly intended to be as brisk as feasible, 3.5 Cox Algorithm


each for little operands and for huge operands. The The Cox– Zucker machine is a calculation made by
speed is accomplished by exploitation full words in David A. Cox and Steven Zucker. This calculation
light of the fact that the fundamental number decides whether a given arrangement of areas gives a
juggling sort, by exploitation brisk calculations, with premise (up to torsion) for the Mordell– Weil
to a great degree enhanced gathering code for the gathering of an elliptic surface E → S where S is
principal regular inward circles for loads of CPUs, isomorphic to the projective line. The recipe was
and by a general weight on speed. The primary GMP introductory uncovered inside the 1979 paper
unharness was made in 1991. it's consistently created "Crossing point quantities of segments of elliptic
and kept up, with a substitution unharness surfaces" by Cox and Zucker and it totally was later

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


127
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

named the "Cox– Zucker machine" by Charles inside the archive, this can be not an inconsequential
Schwartz in 1984. assignment.
V. RESULTS

Fig 3. file for auditor

Fig 2.Working of cox algorithm

3.6 Possible Data Distortion


Here a simple cacophonic algorithmic program is
utilized. Here the picture is spilt into n similarly
estimated squares. In any case, once utilized a
powerful watermark for the modest components
(that will be that the - issue used by the Cox
algorithmic program is zero.5), varieties between Fig 4.web server
neighboring components ended up noticeable in spite
of the way that the one watermarks zone unit in
cognoscible.This outcome turns out to be
considerably more grounded when numerous
emphases as decided. Now and again, this bending
would potentially affect the ease of use of the record.
The Cox algorithmic program is utilized with relates
alpha issue of zero.1 and no mutilation is noticeable.
It would entrance to examine if this drawback are
regularly dodged by exploitation extra detailed
cacophonic calculations. As most watermarking plans
assemble utilization of the nearness of information Fig 5. consumer download page

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


128
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

create the predefined assurance for touchy data.


LIME is adaptable as we have a tendency to separate
between beyond any doubt senders (typically
proprietors) and untrusted senders (generally buyers).
Inside the instance of the beyond any doubt sender, a
dreadfully simple convention with next to no
overhead is achievable. The untrusted sender needs a
considerable measure of modern convention,
however the outcomes don't appear to be bolstered
put stock in presumptions furthermore, along these
Fig 6. Attacker being detected and blocked
lines they should be prepared to change over an
unbiased substance (e.g. a judge). Our work
conjointly rouses any investigation on data release
recognition procedures for differed report sorts
furthermore, projections. To Illustrate, it'll be a
propelling future examination heading to style a
certain ancestry convention for determined data.

VII.REFERENCES

Fig 2. A message request from a non guilty consumer [1]. "Chronology of data breaches,"
http://www.pri"acyrights.org/data-breach.
VI. CONCLUSION AND FUTURE ENHANCEMENT [2]. "Data breach cost,"
http://www.symantec.com/about/news/release/
We are blessing LIME, a model for capable data article.jsp?prid=2011030801
exchange over different elements. We tend to plot [3]. "Pri"acy rights clearinghouse,"
teaming up parties, their interrelationships and give a http://www.pri"acyrights.org.
solid inner portrayal for a data exchange convention [4]. "Electronic Pri"acy Information Center (EPIC),"
utilizing a novel blend of negligent exchange, strong http://epic.org, 1994.
watermarking and computerized marks. We have a [5]. "Facebook in Pri"acy Breach,"
tendency to demonstrate its rightness and show that http://online.wsj.com/article/SB1000142405270
it's feasible by giving little seat checking comes about. 2304772804575558484075236968.html.
By introducing a general relevant structure, we have [6]. "Offshore outsourcing,"
a tendency to present answerableness as right on time http://www.computerworld.com/s/article/1099
as inside the style area of a data exchange foundation. 38/Offshore outsourcing cited in Florida data
Despite the fact that LIME doesn't effectively prevent leak.
data release, it presents responsive answerableness. [7]. A. Mascher-Kampfer, H. Stogner, and A. Uhl,
Along these lines, it'll deflect vindictive gatherings "Multiple re-watermarking scenarios," in
from unseaworthy individual reports and can Proceedings of the 13th International
energize legitimate (however reckless) gatherings to Conference on Systems, Signals, and Image

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


129
Amith Manda, Dr. R. P. Ramkumar. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 123-130

Processing (IWSSIP 2006). Citeseer, 2006, pp. Cite this article as :


53-56.
[8]. P. Papadimitriou and H. Garcia-Molina, "Data Amith Manda, Dr. R. P. Ramkumar, "Data Extraction
leakage detection," Knowledge and Data in Malignant Environments", International Journal of
Engineering, IEEE Transactions on, "ol. 23, no. Scientific Research in Computer Science, Engineering
1, pp. 51-63, 2011. and Information Technology (IJSRCSEIT), ISSN :
[9]. "Pairing-Based Cryptography Library (PBC)," 2456-3307, Volume 5 Issue 1, pp. 123-130, January-
http://crypto.stanford.edu/pbc. February 2019. Available at doi :
[10]. I. J. Cox, J. Kilian, F. T. Leighton, and T. https://doi.org/10.32628/CSEIT195116
Shamoon, "Secure spread spectrum Journal URL : http://ijsrcseit.com/CSEIT195116
watermarking for multimedia," Image
Processing, IEEE Transactions on, "ol. 6, no. 12,
pp. 1673-1687, 1997.
[11]. B. Pfitzmann and M. Waidner, "Asymmetric
fingerprinting for larger collusions," in
Proceedings of the 4th ACM conference on
Computer and communications security, ser.
CCS ’97, 1997, pp. 151-160.
[12]. S. Goldwasser, S. Micali, and R. L. Ri"est, "A
digital signature scheme secure against adapti"e
chosen-message attacks," SIAM J. Comput., "ol.
17, no. 2, pp. 281-308, 1988.
[13]. A. Adelsbach, S. Katzenbeisser, and A.-R.
Sadeghi, "A computational model for
watermark robustness," in Information Hiding.
Springer, 2007, pp. 145-160.
[14]. J. Kilian, F. T. Leighton, L. R. Matheson, T. G.
Shamoon, R. E. Tarjan, and F. Zane, "Resistance
of digital watermarks to collusi"e attacks," in
IEEE International Symposium on Information
Theory, 1998, pp. 271-271.
[15]. M. Naor and B. Pinkas, "Efficient obli"ious
transfer protocols," in Proceedings of the
Twelfth Annual ACM-SIAM Symposium on
Discrete Algorithms, 2001, pp. 448-457

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


130
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1| ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195119

Diagnosis of Various Thyroid Ailments using Data Mining


Classification Techniques
Umar Sidiq1, Dr. Syed Mutahar Aaqib2, Dr. Rafi Ahmad Khan3
1 Research Scholar, Department of Computer Science, Mewar University, Rajasthan India
2Assistant Professor, Department of Computer Science, Amar Singh College, Srinagar, Jammu and Kashmir,
India
3Assistant Professor, University of Kashmir Srinagar, Jammu and Kashmir, India

ABSTRACT

Classification is one of the most considerable supervised learning data mining technique used to classify
predefined data sets the classification is mainly used in healthcare sectors for making decisions, diagnosis
system and giving better treatment to the patients. In this work, the data set used is taken from one of
recognized lab of Kashmir. The entire research work is to be carried out with ANACONDA3-5.2.0 an open
source platform under Windows 10 environment. An experimental study is to be carried out using classification
techniques such as k nearest neighbors, Support vector machine, Decision tree and Naïve bayes. The Decision
Tree obtained highest accuracy of 98.89% over other classification techniques.
Keywords : Thyroid disease, K-Nearest Neighbor, Support Vector Machine, Decision Tree, Naïve Bayes.

I. INTRODUCTION hormones released from the thyroid gland which is


one of the important organlocated in the front of the
Classification techniques play a vital as well as major neck and below the Adam's apple of our body. The
role in analyzing survivability of diseases and secretion of thyroid hormones from the thyroid gland
providing facilities to reduce the cost to the patients. are of two types i.e. levothyroxine or T4 and
Now-a-days, Disease diagnosis has become very triiodothyronine or T3. These hormones help in
crucial because of occurrence of so many diseases production of balanced amount of proteins,
every year. People from all over the world have been regulating the temperature of body, and maintaining
suffering from various health issues like diabetes, overall production of energy [18]. Thyroid disease
heart disease, typhoid, tuberculosis, kidney disease occurs when thyroid gland stop to functioning
etc [16] [17].Beside these health issues , thyroid properly and are mainly divided into hypothyroidism
disease have also been detected worldwide and thus and hyperthyroidism [4].
become a serious endocrine health problem and an
issue of concern. It is expected that in India about 42 The excess and deficient secretion amount of thyroid
million people suffer from thyroid disorders [2]. As hormone causes hyperthyroidism and
per resent studies, women are 5 to 8 times more hypothyroidism respectively. The common
prone to thyroid disorders than men worldwide. It is symptoms of hyperthyroidism are sudden weight loss,
caused by the improper secretion of thyroid rapid heartbeat, nervousness, etc. and

CSEIT195119| Received :15Jan 2019| Accepted : 25Jan2019 | January-February-2019[ 5 (1) : 131-136]


131
Umar Sidiqet al. Int J Sci Res CSE & IT. January-February-2019 ;5(1):131-136

hypothyroidism has weight gain, tiredness, weakness, neuro fuzzy method emplaced in ESTDD system for
feeling cold etc. One of the most common cause thyroid diseases diagnosis and achieved 95.33% of
occurred due to hyperthyroidism is graves’ disease [3]. accuracy. ESTDD is one of significant tool for
The underestimated thyroid disease causes thyroid endocrinologists or students studying endocrinology
storm and myxedema which may lead to death [12]. for testing their knowledge by comparing their
In this research work, a classification model is predictions with ESTDD[21].
trained using classification algorithms like K-nearest
neighbor (KNN),support vector machine (SVM), In year 2009, Temurtas performed diagnosis of
decision tree (DT) and Naïve bayes (NB) for the thyroid diseases with the help of artificial neural
diagnosis of thyroid diseases. The Decision Tree network with 3-fold and 10-fold cross-validation
outperformed over other techniques. approaches were used to estimate the performance of
the used neural networks. In this work highest
The rest of the paper is followed as: - Section 2 accuracy of 94.81% was achieved [12].
represents related work in diagnosis of thyroid
diseases. Section 3 contains dataset and methods. In year 2011, EsinDogantekin et al. worked on
Section 4 represents the results and discussion. Wavelet Support Vector Machine (WSVM) with
Section 5 contains conclusion and at last references Generalized Discriminant Analysis (GDA) methods
are mentioned. for diagnosis of thyroid diseases and achieved 91.86%
classification accuracy [22].
II. RELATED WORK
K.Saravana Kumar et al in 2014, proposed KNN and
By studying literature, it seems that several SVM classification methods for the diagnosis of
classification methods have been developed for thyroid disease. They showed that the prediction
thyroid diseases diagnosis. In year 2002, Ozyilmaz et accuracy of KNN and SVM are 96.34% and 94.43%
al in [14] showed that feed forward neural network respectively [1].
could be successfully used for diagnosis of thyroid
diseases. In this work, three architectures of MLP, In year 2017 MP Gopinath proposed SVM, FKNN
RBF and CSFNN were used and achieved accuracy and Decision tree for the diagnosis of thyroid diseases.
89.80%, 79.08% and 91.14% respectively. This work The SVM obtained highest classification accuracy of
shows CSFNN gives the best classification accuracy 96.30% over other techniques [23].
and takes less training times than all other algorithms.
III. DATASET AND METHODS
In 2007, Polat et al. studied on the artificial immune-
recognition system (AIRS) classification method for The dataset used in this investigate work is a clinical
diagnosis of thyroid diseases and achieved 81% dataset. The dataset was taken from one of the
classification accuracy. On the other hand AIRS with leading diagnostic lab in Kashmir. The dataset
Fuzzy weighted pre-processing was classified and contains the record of 807 patients of almost all age
showed the accuracy of 85.00% .This means AIRS groups. Out of 807 patients (224 Males and 583
with fuzzy outperformed than AIRS [15]. Females) 553 belongs to normal, 218 belongs to
hypothyroidism and 36 belongs to hyperthyroidism.
In year 2008, keles et al. proposed an expert system The dataset has 6 attributes as: age, gender, TSH, T3,
for thyroid disease diagnosis (ESTDD) and using T4 and added classification attribute for indication of

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


132
Umar Sidiqet al. Int J Sci Res CSE & IT. January-February-2019 ;5(1):131-136

normal or hyperthyroidism or hypothyroidism. Table hyper plane is generally done, that has largest
1 shows the description of dataset. distance to the closest training data point of any class
(so-called functional margin), since in general the
greater the margin the smaller the generalization
Table 1. Attributes used in our dataset. error of the classifier [19].

Serial Attribute description Value Decision tree


No. name Decision Tree (DT) is tree like graph known as one of
1 Age Age in years Numeric the most admired classification data mining
technique that splits the dataset into parts on the
2 Gender M-Male Nominal
decisions [5]. In decision tree, each internal node or
F- Female
non-leaf node represents a test on a particular
3 TSH Continuous Numeric
attribute, each branch denotes the outcome of that
4 T3 Continuous Numeric test, and each leaf node has a class label. The paths
through which a particular test data is to classify
5 T4 Continuous Numeric from root to leaf represent classification rules based
6 Results Normal Nominal on maximum information gain [6].
Hyperthyroidism
Hypothyroidism Naïve Bayes
Naïve Bayes (NB) is a simple classification algorithm
K Nearest Neighbour for predictive modeling with clear semantics,
K-nearest neighbour (KNN) is a supervised technique representing and the probabilistic learning method
as well as non-parametric in nature. The input of K- based on Bayesian theorem [13]. Naive Bayes
NN depends on the K closest instances present in the classifier assumes that the value of one attribute is
feature space. The generated output depends on not dependent on the value of any other attribute,
whether KNN is Classification or regression methods and it assumes that the presence or absence of
[7] [8].When prediction is required for undetected particular attribute doesn’t affect the prediction
data instances, the KNN algorithm will search process. Suppose there are m classes say C1, C2….Cn
through the training data instances for the k-most having a unidentified data sample X, Naive Bayesian
similar instances. The prediction attribute of the most classifier will predict an unknown sample X to the
similar instances is summarized and returned as the class Ci on the basis of class having highest
prediction for the undetected instance [19]. probability [20].
P( Ci |X > P( Cj |X ) for 1≤j≤m , j ≠ i
Support Vector Machine
Support vector machine (SVM) is a supervised Anaconda
learning classification technique that are used to For implementation of our methods we used
analyzethe data for regression and classification ANACONDA3-5.2.0 64 bit a free and open source
methods [9].It constructs an optimal hyper plane in platform distribution of python and R programming
a high- or infinite-dimensional space in which new language with number of modules, packages and
examples are assigned to one group or the other libraries that provides multiple ways of achieving
one[10].The separation of data is achieved by the classification problems. ANACONDA can be
downloaded from the website [11].

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


133
Umar Sidiqet al. Int J Sci Res CSE & IT. January-February-2019 ;5(1):131-136

In this work, Out of four classifiers, The Decision


Tree shows highest test accuracy of 98.89% over
other classifiers. The accuracy comparison among
K- Fold cross- validation classifiers is shown in figure 2. The highest accuracy
In k-fold validation, the whole dataset is divided into of our method is higher than some of the methods
K equal size subsets and one of the subset K is taken available in literature as shown in table 2.
as test data and the remaining K-1 folds acts as
training data. Thus different test results exist on each
iteration and at last average of these results gives the
test accuracy of the algorithm [12]. In this study 10-
fold cross-validation is used to find out the
classification accuracy by the classification methods.

System implementation
The thyroid dataset has three output categories. First
one is Normal, second one is hypothyroidism and
third one is hyperthyroidism. The thyroid dataset of
five input attributes Age, Gender, TSH, T3 and T4 is
Figure 2. Accuracy comparison of classifiers
supplied to the classifiers of KNN, SVM, Decision
Tree and Naïve bayes in PYTHON to classify the data.
Study Method Highest
The performance of each classifier is evaluated in
Accuracy
terms of accuracy as shown in figure 1.
Ozyılmaz and MLPNN with BP 91.14%
Yıldırım [14 ] (3×FC)
Data mining MLPNN with FBP
classification Performance
techniques measure (3×FC)
RBF
KNN (3×FC)
CSFNN
Thyroid Accuracy
dataset SVM (3×FC)
Polat et al. [15 ] AIRS 85.00%
Decision
Tree
(10×FC)
AIRS with Fuzzy
(10×FC)
Naïve
Bayes weighted
preprocessing
keles et al. [21] ESTDD 95.33%
F. Temurtas [12 ] MLNN with LM 94.81%
Figure 1. Thyroid disease prediction process (3×FC)
PNN (3×FC)
IV. RESULTS AND DISCUSSION LVQ (3×FC)

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


134
Umar Sidiqet al. Int J Sci Res CSE & IT. January-February-2019 ;5(1):131-136

MLNN with LM
(10×FC) [1].K.Saravana Kumar, Dr. R. ManickaChezian,
PNN "Support Vector Machine and K- Nearest Neighbor
(10×FC) Based Analysis for the Prediction of Hypothyroid.
LVQ International Journal of Pharma and Bio
(10×FC) Sciences",volume - 2,Issue - 5,page no-(447-453),2014
EsinDogantekinet GDA-WSVM 91.86% .
al.[22] [2].http://www.ncbi.nlm.nih.gov/pmc/articles/PMC3
MP SVM 96.30% 169866/(accessed dec 2015)
Gopinath.[23] FKNN [3].G. Zhang, L.V. Berardi, An investigation of neural
Decision Tree networks in thyroid function diagnosis, Health Care
Our study KNN 98.89% Manage. Sci. (1998)
(10×FC) [4].Xia C, Hsu W (2006) BORDER: efficient
computation of boundary points. In: IEEE, 2006
SVM
(10×FC) [5].Available from: http://en.wikipedia.org.Last
DT accessed on Dec24].
(10×FC) [6].Apte & S.M. Weiss, Data Mining with Decision
NB Trees and Decision Rules, T.J. Watson Research
(10×FC) Center,
http://www.research.ibm.com/dar/papers/pdf/fgcsapt
Table-2 shows the highest classification accuracy ewe issue_with_cover.pdf, (1997).
obtained by our method over other methods used for [7].Roychowdhury S (2014) DREAM: diabetic
the diagnosis of thyroid diseases. retinopathy analysis using machine learning. In:
IEEE, 2014
V. CONCLUSION [8].Chetty N, Vaisla KS, Patil N (2015) An improved
method for disease prediction using fuzzy approach.
The work has been done using classification data In: IEEE, 2015
mining techniques for the diagnosis of thyroid disease. [9].S. Sathiya Keerthi, Olivier Chapelle, Dennis
For this purpose, K nearest neighbor, Support vector DeCoste "Building Support Vector Machines with
machine, Decision tree and Naive Bayes classifiers Reduced Classifier Complexity" Journal of Machine
have been used. The Decision Tree classifier Learning Research, Vol: 7, PP 1493- 515, January -
outperformed over other classifiers. However, if we (2006).
merge it with any other classification technique such [10].Shen X, Lin Y (2004) Gene expression data
as neural network, then the result might be even classification using SVM-KNN classifier". In: IEEE,
better as compared to what we got with the current 2004
study. [11].www.anaconda.com.
[12].F. Temurtas, "A comparative study on thyroid
disease diagnosis using neural networks," Expert
Systems with Applications, vol. 36, 2009, pp. 944-
949.
VI. REFERENCES

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


135
Umar Sidiqet al. Int J Sci Res CSE & IT. January-February-2019 ;5(1):131-136

[13].G. H. John and P. Langley, "Estimating International Journal of Pure and Applied
Continuous Dis-tributions in Bayesian Classifiers," Mathematics Volume 117 No. 7 2017, 53-63.
Proceedings of the 11th Conference on Uncertainty Cite this article as :
in Artificial Intelligence, San Francisco, 1995, pp.
338-345. Umar Sidiq, Dr. Syed Mutahar Aaqib, Dr. Rafi
[14].Ozyılmaz, L., Yıldırım, T. (2002). Diagnosis of Ahmad Khan, "Diagnosis of Various Thyroid
thyroid disease using artificial neural network Ailments using Data Mining Classification
methods. In Proceedings of ICONIP’02 9th Techniques", International Journal of Scientific
international conference on neural information Research in Computer Science, Engineering and
processing (pp. 2033-2036). Singapore: Orchid Information Technology (IJSRCSEIT), ISSN : 2456-
Country Club. 3307, Volume 5 Issue 1, pp. 131-136, January-
[15].Polat, K., Sahan, S., & Gunes, S. (2007). A novel February 2019.
hybrid method based on artificial immune Available at doi :
recognition system (AIRS) with fuzzy weighted https://doi.org/10.32628/CSEIT195119
preprocessing for thyroid disease diagnosis. Expert Journal URL :http://ijsrcseit.com/CSEIT195119
Systems with Applications, 32, 1141-1147.
[16].Sehgal MSB, Gondal I (2014) K-ranked
covariance based missing values estimation for
microarray data classification. In: IEEE, 2004
[17].Bonner A (2004) Comparison of discrimination
methods for peptide classification in tandem mass
spectrometry. In: IEEE, 2004
[18].HalifeKodaz et al. Medical application of
information gain based artificial immune recognition
system (AIRS): Diagnosis of thyroid disease.
[19].Joel Jacob et al. "Diagnosis of Liver Disease Using
Machine Learning Techniques". (IRJET) Volume: 05
Issue: 04 | Apr-2018
[20].K. Pavya et al. "Diagnosis of Thyroid Disease
Using Data Mining Techniques: A Study". (IRJET)
Volume: 03 Issue: 11 | Nov -2016.
[21].Keles, A., and Keles, A., ESTDD: Expert system
for thyroid diseases diagnosis. Expert Syst. Appl.
34(1):242-246, 2008.
[22]. Dogantekin, E., Dogantekin, A., and Avci, D.,
An expert system based on generalized discriminant
analysis and wavelet support vector machine for
diagnosis of thyroid diseases. Expert Syst. Appl.
38(1):146-150, 2011.
[23].M.P.Gopinath "Comparative Study on
Classification Algorithm for Thyroid Data Set".

Volume 5, Issue 1, January-February-2019 | http://ijsrcseit.com


136
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195120
Analysis of An Indoor Air Quality Antenna for Asthma
Management in Children
N. Vijayalakshmi1, J. G. Rohini2
1 Assistant Professor, Department of Computer Science, Bon Secours College for Women, Bharathidasan
University, Thanjavur, Tamil Nadu, India
2M. Sc Computer Science, Bon Secours College for Women, Bharathidasan University, Thanjavur, Tamil Nadu,
India
ABSTRACT

The extensive of diseases related with Indoor Air Quality (IAQ) usually referred to as Sick Building Syndrome
(SBS) and Building Related Illness (BRI) has generated concern for technological approach in monitoring and
analysis of the quality of air in confine environments. In line with “prevention is better than cure” mantra, such
monitoring system is necessary for timely awareness of quality of surrounding air so as to take preventive
action not only at the individual level, but also aid prompt diagnosis by the medical personnel. Motivated by
the need for an affordable, simple and efficient indoor air quality device for general use at homes, offices and
other related confine environment, a sensor node with capability for wireless sensor networking is presented in
this study. The proposed device incorporates: sensors to measure major air quality indices: odor, carbon
monoxide, dust, humidity, and temperature, LED indicators for prompt visual status alert, memory device (SD
card) for off-line data access, and wireless capability for remote data transfer in applications such as web and
mobile-apps interfacing. The performance of the system is evaluated in a laboratory scale indoor environment.
The results show effectiveness of the system in capturing the air-quality status and providing timely
information for proactive action.
Keywords : Indoor Air Quality, Sick Building Syndrome, Building Related Illness, wireless capability, mobile-
apps.
I. INTRODUCTION has called for a pro-active solution proportionate to
the advances in the related technological. SBS and
The alarming rate of indoor-air quality related BRI are related and can be defined as sickness and
diseases commonly referred to as Sick Building discomfort experience by occupants due to poor
Syndrome (SBS) and Building Related Illness (BRI) indoor air environment with no specific causes

CSEIT195120 | Received : 01 Feb 2019 | Accepted : 10 Feb 2019 | January-February -2019 [ 5 (1) : 137-142]
137
N. Vijayalakshmi, J. G. Rohini Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 137-142

identify but related to amount of time spent inside. was explored and analyzed in comparison to
Symptoms such as headaches, sore throats, itchy eyes, acceptable standard set by ASHRAE [7]. In [8], the
breathing difficulties, lung cancer, asbestosis, indoor air parameters variation with respect to
mesothelioma, asthma, skin irritation and allergic microclimate condition and impact of outdoor
reactions are among the reported health effect of environment was investigated and reported with
poor air environment according to medical report reference to large museum exhibition halls for a
and study [1-6]. According to the World Health period of one year. The study showed the differences
Organization (WHO), 4.3 million people a year die in indoor microclimate in air-conditioned rooms and
from the exposure to household air pollution [4]. The non-air conditioned ones in the same building. In
problems are been compounded with the increasing addition, external conditions were reported to have a
amount of time people spend in a confine significant impact on the indoor air parameters. On
environment - either in home or in offices or in car, the other hand, at the system level, the pace of
or malls or in school-, and the increasing potential of research work on monitoring systems for IAQ is
air pollution due to cooking using solid fuels, driven by the dynamic rate of sensors and computing
smoking in and around buildings, poor ventilation, technologies advancement. Several studies have been
dust, humidity and temperature. For instance, it has reported in this direction [9-15]. A wireless
been reported that people spend a substantial monitoring system incorporating temperature,
proportion of their time in buildings [5]. In the arid humidity, CO and CO2 sensors for HVAC control
region (MENA) most people spend 90% of their time and IAQ monitoring is presented in [13]. An
indoors [1,2]. In response to these problems, research ATMega88 with a transceiver was adopted in the
studies have been directed to (i) qualitative study for the implementation, sample performance
monitoring and analysis of indoor air quality, results was presented to justify the success of the
including guidelines on IAQ [4] and (ii) development study. Sneha and Sandeep [10] presented an
of sensor node (monitoring system) for quantitative embedded system model for air quality monitoring
monitoring of the IAQ for several applications. For incorporating a LCD display unit to help an allergic
instance, assessment of indoor air quality pollutants patient aware of its environment. No real-life data
in the offices and their potential negative effect in results presented in the study. Hybrid sensor system
Dubai were carried out in [1] using both quantitative with both stationary and mobile sensors is reported
and qualitative techniques. In that study, the in [15] for accurate tracking of air pollutant.
concentration levels of specific indoor air quality Meanwhile, apart from the fact that no single system
indicators (e.g. TVOC, Ozone, Carbon Monoxide, exists that addresses all the pollutants nor
Carbon Dioxide, Relative Humidity and Temperature) incorporates all the performance requirements for a

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


138
N. Vijayalakshmi, J. G. Rohini Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 137-142

given situation and environment, most of the IAQ designated as follows: Green: indicates clean, Yellow:
devices are either expensive, thereby hindering indicates moderate, and Red: indicates danger. This is
affordability for general use, or lack simplified data to provide easy and prompt status information for the
presentation for a common end users. Hence, the end users for timely proactive action
current study proposes a wireless sensor node for
IAQ monitoring with simplicity of indication Apart from the cost and power consumption
(situation alert), affordability, easy of modification which are aimed to be low cost and low power,
and maintenance, and yet provides opportunity of respectively, the choice of the sensors stated above is
being fixed located at a place or used as a mobile based on characteristic of the local environment, and
hand-held device (i.e. mobility option). In addition, it common potential pollutant-generator indoor
is aimed to serve as a base for further research on the activities such as
ongoing IAQ data monitoring and analysis, and IAQ • Tobacco smoking,
wireless sensor networks applications suitable for the • Traditional charcoal-based air freshener
emerging Internet of Things (IoT). • Dust (penetration from outside dusty air, poorly
maintained HVAC systems, etc.)
The subsequent sections provide the essential of IAQ • Cooking gas leakage
followed by the design and development of the • Sewage pipe leakage
proposed system, result, discussion and further study. • Ambient thermal influence
The conclusion of the study is presented at the last
section.

II. METHODOLOGY

The simplified block and components interfacing


diagrams of the proposed IAQ Wireless Sensor Node
(WSN) are shown in Fig. 1. The system comprises of
five air parameter sensors - Carbon monoxide (CO),
Methane (CH4), dust, humidity and temperature Fig 1 : Simplified Block and Components Interfacing

sensors-, microcontroller board for processing, LEDs Diagrams.

display for status alert, micro SD card for data


capturing, Wi-Fi for wireless interface capability and Several compounds constitute air pollutants once

power supply. The three LEDs status alert are their concentration exceeded the Occupational

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


139
N. Vijayalakshmi, J. G. Rohini Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 137-142

Exposure Level (OEL) specified by appropriate Thermal anemometer, Direct-reading dust sampler,
regulations/standards. The types and composition of and microbial sampling [16]. It is noted that, (i) there
the indoor air pollutants depends on ambient outside is no single instrument to sample all the pollutants
air condition and nature of the confined environment which necessitates the need for localized device for a
(be it residential, office, markets/malls, sport center given situation, (ii) some of the testing and sampling
etc.). methods are only interpretable by IAQ professionals,
thereby limits their direct adoption and benefits to
Essentially for a given indoor environment, the IAQ the occupants (general users) and (iii) affordability in
issues involved identification of potential pollutants, terms of cost still remains an obstacle towards
their sources, health concerns, standards/regulations widespread adoption of IAQ monitoring system for
in respect of their OEL, monitoring and preventive preventive maintenance and safety. These issues have
measures to minimize human exposure, and initiating led to dynamic research and development of different
corrective action when necessary. Among the major types of air testing/sampling instruments to meet the
and common parameters in IAQ are Carbon demands for timely and informative monitoring of
monoxide, Carbon dioxide, Dust/particulates, Volatile the ambient indoor air condition in line with the
Organic Compounds (VOCs), Formaldehyde, state-of-the-art technology.
Methane, and temperature/Humidity for thermal
comfort. Others are Benzene, Nitrogen dioxide, III. RESULTS, DISCUSSION AND FUTURE STUDY
Sulfur dioxide, Radon etc. [16]. Table 1 gives the
summary of the sources, effects and standards of the The results of both web data transmission and real-
common IAQ parameters while the detailed can be time data monitoring show the successful
obtained in the literatures [4, 16]. implementation of the proposed IAQ-WSN. The web
data interface of Fig. 1 provides the detail
With respect to the management of IAQ issues (such information on the status of the air parameters and
as monitoring, prevention and control) the core working condition of the device. This is aimed at
requirement is air quality sampling and testing with providing remote data access for occupants and
the aid of appropriate instruments. Specific health officers for timely preventive action. The
procedures have been proposed and reported in the results of Fig. 5 show the effectiveness of the node in
sampling and measurements of different types of air monitoring the pollutant in real-time. All the
contaminants. The common types of instrument are: pollutants data were captured almost at the same time,
direct-reading Colorimetric Tubes, Direct-reading (around 6000seconds) based on the simultaneous
Portable monitors, Air samplers, Smoke tube, injection of the materials in the enclosure. This

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


140
N. Vijayalakshmi, J. G. Rohini Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 137-142

provides a qualitative performance of the design and [2]. Jennifer Bell, “Sick buildings are leading to
development of the proposed IAQ-WSN. Future sick UAE office workers, doctors say”, The
study is expected to conduct quantitative analysis and Nation, UAE, Spetember 12, 2016.
benchmark the system performance with standard / http://www.thenational.ae/uae/health/sick-
industrial IAQ sensors. In addition, network of the buildings-are-leading-to-sickuae- office-
nodes is to be developed with appropriate routing to workers-doctors-say. accessed date: 1/10/2017.
cover large indoor environment or complete building. [3]. Khawaja Mohammad Hassan, Sick Building
Syndrome: The killer within, The Khaleej
IV. CONCLUSION times,
http://www.khaleejtimes.com/article/2011011
The design and development of a wireless sensor 5/ARTICLE/3011 59995/1002, Filed on
node for Indoor Air Quality monitoring system was January 16, 2011, accessed date: 10/12/2017.
presented in this project. The evaluation results of [4]. World Health Organization (WHO),
both web data transmission and real-time data “Guidelines for Indoor Air Quality: Selected
monitoring show the successful implementation of Pollutants”.WHO,2010.http://www.euro.who.
the proposed IAQ-WSN. Future study is expected to int/__data/assets/pdf_file/0009/128169/e94535.
conduct quantitative analysis and benchmark the pdf.
system performance with standard / industrial IAQ [5]. HS. Brightman and N. Moss. Sick building
sensors. In addition, network of the nodes is to be syndrome studies and the compilation of
developed with appropriate routing to cover large normative and comparative values. Indoor Air
indoor environment or complete building. Quality Handbook. Editors: J. D. Spengler, J. F.
McCarthy, and J. M. Samet, New York:
V. REFERENCES McGraw-Hill, 2001.
[6]. US Environmental Protection Agency
[1]. Hanan Ahmad Ibrahim, Indoor Air Quality in (USEPA). Indoor air quality tools for schools
UAE Office Buildings and Their Effects on Communications guide. http://www.epa.gov,
occupants’ Health, Comfort, Productivity and accessed date: 10/12/2017
Performance”, Master thesis, Faculty of [7]. ASHRAE: The American Society of Heating,
Engineering and IT MSc Sustainable Design of Refrigerating and Air- Conditioning
the Built Environment, The British University Engineers. https://www.ashrae.org. accessed
in Dubai, May 2015. date: 10/12/2017.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


141
N. Vijayalakshmi, J. G. Rohini Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 137-142

[8]. Joanna Ferdyn-Grygierek Monitoring of Electrical, Electronics and System Engineering


indoor air parameters in large museum (ICEESE), Pages: 60 - 64.
exhibition halls with and without air- [13]. R. du Plessis, A. Kumar; GP Hancke and BJ
conditioning systems, Building and Silva, (2016). “A wireless system for indoor air
Environment Volume 107, October 2016, quality monitoring”, IECON 2016 - 42nd
Pages 113-126, Elsevier 2016. Annual Conference of the IEEE Industrial
[9]. Shaharil Mad Saad, Abdul Rahman Mohd Electronics Society, Pages: 5409 - 5414.
Saad, Azman Muhamad Yusof Kamarudin, [14]. Jung-Yoon Kim, Chao-Hsien Chu, and Sang-
Ammar Zakaria, and Ali Yeon Md Shakaff, Moon Shin, (2014). “ISSAQ: An Integrated
“Indoor Air Quality Monitoring System using Sensing Systems for Real- Time Indoor Air
Wireless Sensor Network (WSN) with Web QualityMonitoring”, IEEE Sensors Journal ,
Interface” proceeding International Volume: 14, Issue: 12,Pages: 4230 - 4244.
conference of electrical, electronics and [15]. Yun Xiang, Ricardo Piedrahita, Robert P.
system engineering, (ICEESE) Malaysia, 2014. Dick, Michael Hannigan, Qin Lv, and Li
[10]. Sneha Jangid and Sandeep Sharma, (2016). “An Shang, (2013). “A Hybrid Sensor System for
embedded system model for air quality Indoor Air Quality Monitoring”,2013 IEEE
monitoring”, 2016 3r International Conference International Conference on Distributed
on Computing for Sustainable Global Computing in Sensor Systems, Pages: 96 - 104.
Development, Pages: 3003 - 3008. [16]. Indoor Air Quality Tool Kit, (2009).
[11]. Tsang-Chu Yu, Chung-Chih Lin,Chun-Chang Government of Alberta, Employment and
Chen, Wei-Lun Lee, Ren- Guey Lee, Chao- Immigration.
Heng Tseng, and Shi-Ping Liu, “Wireless
sensor networks for indoor air quality
monitoring” Medical Engineering & Physics, Cite this article as :
Volume 35, Issue 2, February 2013, Pages 231-
235. N. Vijayalakshmi, J. G. Rohini, "Analysis of An
[12]. Shaharil Mad Saad, Abdul Rahman Mohd Indoor Air Quality Antenna for Asthma
Saad, Azman Muhamad Yusof Kamarudin, Management in Children", International Journal
Ammar Zakaria, Ali Yeon Md Shakaff, (2013).
of Scientific Research in Computer Science,
“Indoor air quality monitoring system using
Engineering and Information Technology
wireless sensor network (WSN) with web
interface”,2013 International Conference on

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


142
N. Vijayalakshmi, J. G. Rohini Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 137-142

(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1,


pp. 137-142, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195120
Journal URL : http://ijsrcseit.com/CSEIT195120

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


143
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195132
Home Automation using Google Voice Assistance with
Raspberry Pi
Undra Narasimha Babu1, S. N. Sreekanth2
1PG Scholar, Department of ECE, Siddartha Educational Academy Group of Institutions, Gollapalli, Andhra

Pradesh, India
2Associate Professor, Department of ECE, Siddartha Educational Academy Group of Institutions, Gollapalli,
Andhra Pradesh, India

ABSTRACT

Home automation popular in day to day to number of advantages. In previous and current we monitor the
home appliances remotely by emails or text messages. This paper gives an innovative idea of controlling the
home appliance using with Google voice assistance by using the Raspberry Pi. As a server system Internet of
Things (IoT) links smart objects to the internet. It can enable exchange of data in a more secure way. Rapid
adaptations of Smart City and Internet of Things technologies are assisting in urban planning to ensure
sustainable cities and lifestyles. Life is made simpler and more productive through IOT. As the Wi-Fi hotspots
are increasingly becoming common in the recent times, the existing infrastructure can be used to develop a
cost-effective solution to enable the existing appliances with IOT Raspberry Pi is single board computer (SBC)
is proposed in this paper to Raspberry pi known as a Single Board Computer (SBC) is proposed in this paper to
reach these goals. The voice commands will give to Raspberry PI using Google voice assistance. These
commands will control the any appliances or devices at home. This will provide a better communication in
automated home as compared to normal homes.
Keywords: IoT, WI-FI, Raspberry Pi.

I. INTRODUCTION there is a need to robotize everything, so the idea of


brilliant home mechanization framework is a thought
In this cutting-edge time, computerization of which is utilized to make the city shrewd. A Smart
everything is the need of great importance Home is one that gives solace, security and gives the
Automation is the utilization of control frameworks sentiment of home to house individuals. Shrewd
and data innovation to manage gear, modern homes likewise give Energy proficiency (low
hardware, and procedures, limiting the requirement working expense) and comfort consistently, for each
for human inclusion. Computerization assumes an person at home. Home computerization implies the
inexorably imperative job in the worldwide economy checking and control of family unit protests
and in step by step experience. Architects work to insightfully for powerful use. The family questions
connect mechanized gadgets with scientific and ought to be astutely interconnected and also give data
hierarchical devices to make complex frameworks for to better tasks. Home computerization increased with
quickly growing parameters of uses and human the Internet of Things (IoT) gives better adaptability
exercises. For the improvement of keen urban areas, in overseeing and controlling family unit questions in

CSEIT195132 | Received : 15 Jan 2019 | Accepted : 30 Jan 2019 | January-February -2019 [ 5 (1) : 143-146 ]
143
Undra Narasimha Babu, S. N. Sreekanth et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 143-146

a more extensive perspective. This will bolster the used as a piece of equipment wanders, and for an
interconnectivity of countless homes for better asset extensive part of the things that your work zone PC
usage in more extensive zone. does, like spreadsheets, word-getting ready and
diversions. It likewise plays superior quality video.
II. EXISTING SYSTEM We need to see it being utilized by kids everywhere
throughout the world to figure out how PCs
In the existing system, home appliances are function, how to control the electronic world around
controlled through Bluetooth, Zigbee and other them, and how to program.
devices. But they are limited to certain area. To
overcome this problem, we are moving to the
proposed system.

III. PROPOSED SYSTEM

In the proposed system we are control home


appliances using Raspberry Pi with Google assistance
in mobile. We can give command to the Google
assistance according to that command’s loads will
turn on or turn off. The user can have access to Figure 2 : Raspberry Pi3 module
operate the loads from anywhere in the world.

IV. METHODS AND MATERIAL

BLOCK DIAGRAM

Figure 3 : Pin layout of raspberry Pi GPIO used in


system

Figure 1 : block diagram of the system d) Motor


Electrical DC Motors are continuous actuators that
HARDWARE REQUIREMENTS convert electrical energy into mechanical energy.
The DC motor achieves this by producing a
a) Raspberry Pi 3 continuous angular rotation that can be used to rotate
pumps, fans, compressors, wheels, etc.
The Raspberry Pi is an ATM card sized PC that
fittings into your TV. It is a fit little PC which can be

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


144
Undra Narasimha Babu, S. N. Sreekanth et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 143-146

SOFTWARE COMPONENTS

a) Raspbian OS

Raspbian OS is such of the no buts about it operating


systems at hand without electronic commerce a dime
to transform and use. The apparatus is chiefly based
on Debian Linux and is optimized to employment
effectively by the whole of the Raspberry Pi laptop.
Figure 4 : DC motor
As we erstwhile know an OS is a reside of integral
packages and utilities that murmur a indisputable
b) Power Supply
hardware, in this position the Pi. Debian could be
literally lightweight and makes a amazing in a class
all by itself for the Pi. The Raspbian includes tools for
surfing, python programming and a GUI computing
device.

WORKING PRINCIPLE

We are using Raspberry PI to control the appliances


in the home through wireless Communication. The
Appliances in the Home are controlled through
Google assistance through the Wi-Fi Connectivity.
Figure 5 : Flow chart of power supply ON/OFF instructions are given through Wi-Fi
present in the user Mobile, the Data Transfers from
c) Relay mobile phone to Module and then instruction is
further given to the Controller and thus the loads are
Relay is electromagnetic switch that open or close the operated.
switches electrically or electromechanically. Relay is
mostly used to switch smaller circuits. V. RESULTS AND DISCUSSION

APPLICATIONS

➢ Used in homes,
➢ offices,
➢ industrial areas etc.,

ADVANTAGES
.
➢ Smart and secure.
Figure 6 : Relay ➢ Controlling home appliances from anywhere
in the world.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


145
Undra Narasimha Babu, S. N. Sreekanth et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 143-146

RESULTS VI. CONCLUSION

Home automation integrates any electrical devices


and in houses each other. In this paper we
implemented home appliances (fan, electric tubes,
refrigerator and washing machine) control using
Google voice assistance with help of raspberry
Pi.Development of such Smart Home achieve by
using Internet of Things technologies. This system is
low cost and more reliable.

VII.REFERENCES

Figure 7 : Hardware design of the project


[1]. HOME AUTOMATION USING ATmega328
MICROCONTROLLER AND ANDROID
APPLICATION, S.Anusha1, M.Madhavi2,
R.Hemalatha3. International Research Journal
of Engineering and Technology (IRJET),
Volume: 02 Issue: 06 | Sep-2015 www.irjet.net
[2]. Internet of Things: Ubiquitous Home Control
and Monitoring System using Android based
Smart Phone, Rajeev Piyare. International
Journal of Internet of Things 2013, 2(1): 5-11
Figure 8 : Google voice assistance for light on DOI: 10.5923/j.ijit.20130201.02
[3]. Design and Implementation of a WiFi Based
Home Automation System, Ahmed ElShafee,
Karim AlaaHamed. World Academy of Science,
Engineering and Technology International
Journal of Computer, Electrical, Automation,
Control and Information Engineering Vol:6,
No:8,2012.

Cite this article as : Undra Narasimha Babu, S. N.


Figure 9 : Google voice assistance for light off Sreekanth, "Home Automation using Google Voice
Assistance with Raspberry Pi", International Journal
to keep the city clean by informing about the garbage of Scientific Research in Computer Science,
levels of the bins by providing graphical image of the Engineering and Information Technology
bins via IOT Php web development platform. (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
143-146, January-February 2019. Available at doi :
https://doi.org/10.32628/CSEIT195132
Journal URL : http://ijsrcseit.com/CSEIT195132

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


146
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195130

Sentiment Analysis of movie reviews using Microsoft Text


Analytics and Google Cloud Natural Language API
S Jaswanth Sai Chowdary, Sk Waseem Farooq, Sk Abubakar Siddique, S Dinesh
Department of CSE, Vasireddy Venkatadri Institute of Technology, Guntur, Andhra Pradesh, India

ABSTRACT

Sentiment analysis is focused on the extraction of emotions and opinions of the people towards a particular
topic from a structured, semi-structured or unstructured textual data. In this paper we focus our task of
sentiment analysis on The Movie DB (TMDB) database which is an online database and contains information of
movies, television shows and video games including cast, crew and reviews. We examine the sentiment
expression to classify the polarity of the movie review on a scale of 1(highly disliked) to 5(highly liked) and
perform feature extraction and ranking and use these features to train our multilabel classifier to classify the
movie review into its correct label.We will collect the reviews for various movies and perform classification on
them. The classification for the reviews are done based on score obtained for the text. We are using two
approaches to generate scores for the reviews. These approaches are Microsoft cognitive services and Google
cloud natural language API.Finally we are able to generate the scores and can classify them based on movie
reviews.
Keywords : Sentiment Analysis, TMDB database, Microsoft cognitive services, Google cloud natural language
API, classification

I. INTRODUCTION websites are not formal reviews but are rather very
informal and are unstructured form of grammar.
Now a days Internet has become a huge Cyber Opinions expressed in movie reviews give a very true
Database which hosts gigantic amount of data which reflection of the emotion that is being conveyed. The
is created and consumed by the users. The database presence of such a great use of sentiment words to
has been growing at an exponential rate giving rise to express the review inspired us to derive an approach
a new industry filled with it, in which users express to classify the polarity of the movie using these
their opinions across channels such as Rotten sentiment words. Sentiment Analysis is the process of
Tomatoes, Facebook, Twitter, and Foursquare. computationally identifying and categorizing
Opinions which are being expressed in the form of opinions expressed in a piece of text, especially in
reviews provide an opportunity for new explorations order to determine whether the writer's attitude
to find collective likes and dislikes of cyber towards a particular topic, product, etc. is positive,
community. One such domain of reviews is the negative, or neutral. Sentiment Analysis is a
domain of movie reviews which affects everyone technology that will be very important in the next
from writers, users, film critics to the production few years. With opinion mining, we can distinguish
company. The movie reviews being posted on the poor content from high quality content. With

CSEIT195130 | Received : 15 Jan 2019 | Accepted : 30 Jan 2019 | January-February -2019 [ 5 (1) : 147-150 ]
147
S Jaswanth Sai Chowdary et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 147-150

thetechnologies available we can know if a movie has A. Feature Selection


more good opinions than bad opinions and find the Most researchers apply standard feature selection in
reasons why those opinions are positive or negative. their approach to improve computational
Most sentiment analysis studies are now focused on performance with a handful using more sophisticated
social media sources such as TMDB, IMDB [1], approaches. Papers focusing entirely on feature
Twitter [2] and Facebook, requiring the approaches selection to improve sentiment analysis are few. We
be tailored to serve the rising demand of opinions in fetched the review data based on content keyword by
the form of text. Furthermore, performing the eliminating unwanted fields from TMDB [3]
phrase-level analysis of movie reviews proves to be a database. Based on the movie id data related movie is
challenging task. generating which is shown in below figure.

II. METHODS AND MATERIAL

Sentiment analysis aims to determine the attitude of a


speaker or a writer with respect to some topic or the
overall contextual polarity of a document. The
attitude may be their judgment or evaluation,
affective state which is the emotional state while
writing, or the intended emotional communication
which is the emotional effect the writer wishes to
have on the reader. Sentiment is defined as any kind
of emotion and in the context of sentiment analysis is Figure 2 : Movie data send as response by TMDB
the opinion that is being expressed in the form of text database
or speech.
Here details about the author of the movie, id of user
and other unnecessary fields are eliminated. The
review given by user is with key content. The final
reviews are gathered from users by eliminating the
unwanted fields which is shown in the below figure.

Figure 3 : After eliminating unwanted fields reviews


Figure 1 : Process Model Diagram
of each movie

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


148
S Jaswanth Sai Chowdary et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 147-150

B. Microsoft Text Analytics SentiLabel = 4


The Text Analytics is a cloud-based service that Else if SentiScore >=0.4 &&<0.6,then:
provides advanced natural language processing over SentiLabel = 3
raw text, and includes four main functions: sentiment Else if SentiScore >= 0.2 &&<0.4,then:
analysis, key phrase extraction, language detection, SentiLabel = 2
and entity linking. The API is backed by resources Else
in Microsoft Cognitive Services[4], a collection of SentiLabel = 1
machine learning and AI algorithms in the cloud,
readily consumable in your development projects. III.RESULTS AND DISCUSSION

C. Google Cloud Natural Language API


Both the features analyse the reviews data set and
Google Cloud Natural Language [5] reveals the generate the SentiScore. Based on the algorithm
structure and meaning of text both through powerful SentiLabel is determined. Finally the class of the
pre trained machine learning models in an easy to use movie is obtained from the SentiLabel. We use this
REST API and through custom models that are easy scale for defining the class label of the movie reviews
to build with Auto ML Natural Language. You can which is divided as follows:
use Cloud Natural Language to extract information
about people, places, events and much more 1. Sentiment Label 1 belong to class 1
mentioned in text documents, news articles, or blog 2. Sentiment Label 2 belong to class 2
posts. You can use it to understand sentiment about 3. Sentiment Label 3 belong to class 3
your product on social media or parse intent from 4. Sentiment Label 4 belong to class 4
customer conversations happening in a call centre or 5. Sentiment Label 5 belong to class 5
a messaging app. You can analyse text uploaded in
For any movie finally the classification is done based
your request or integrate with your document storage
on the reviews from scale of class 1(highly disliked)
on Google Cloud Storage.
to class 5(highly liked).
D. Classification
We can see that the SentiScore(Sentiment Score) is IV.CONCLUSION
calculated from the two features(Microsoft Text
Analytics and Google Cloud Natural Language API). In this work, we extracted new features that have a
The weight of SentiScore is in the range of 0 to 1. strong impact on determining the polarity of the
Now, once we get the SentiScore[6], we determine movie reviews and applied computation linguistic
the label using the following algorithm: methods for the calculation of sentiment score for the
reviews dataset. We then performed the classification
Algorithm 1. Proposed Algorithm for determining the for the movies by using an algorithm for SentiLabel
label generation.

Initialize SentiScore = 0 In future, we would like to evaluate the effectiveness


SentiScore = value from any of the feature of the proposed sentiment classification features and
If SentiScore >= 0.8 && <=1.0, then: techniques for other tasks, such as sentiment
SentiLabel = 5 classification. We would like to apply in-depth
Elseif,SentiScore>=0.6&&<0.8,then: concepts of Natural Language Processing for better

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


149
S Jaswanth Sai Chowdary et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 147-150

prediction of the polarity of the document. We Cite this article as : S Jaswanth Sai Chowdary, Sk
would also like to extend this technique on other Waseem Farooq, Sk Abubakar Siddique, S Dinesh,
domains of opinion mining like television shows, "Sentiment Analysis of movie reviews using
political discussion forums etc. Microsoft Text Analytics and Google Cloud Natural
Language API", International Journal of Scientific
V. REFERENCES Research in Computer Science, Engineering and
Information Technology (IJSRCSEIT), ISSN : 2456-
[1] Sahu, T. P., & Ahuja, S. (2016). Sentiment 3307, Volume 5 Issue 1, pp. 147-150, January-
analysis of movie reviews: A study on feature February 2019. Available at doi :
selection & classification algorithms. 2016 https://doi.org/10.32628/CSEIT195130
International Conference on Microelectronics, Journal URL : http://ijsrcseit.com/CSEIT195130
Computing and Communications
(MicroCom).DOI:10.1109/microcom.2016.752258
3
[2] Magdum, S. S., &Megha, J. V. (2017). Mining
online reviews and tweets for predicting sales
performance and success of movies. 2017
International Conference on Intelligent
Computing and Control
Systems(ICICCS).DOI:10.1109/iccons.2017.82507
38
[3] https://developers.themoviedb.org/3/movies/get-
movie-reviews
[4] https://azure.microsoft.com/en-
in/services/cognitive-services/text-analytics/
[5] https://cloud.google.com/natural-language/
[6] Topal, K., & Ozsoyoglu, G. (2016). Movie review
analysis: Emotion analysis of IMDb movie
reviews. 2016 IEEE/ACM International
Conference on Advances in Social Networks
Analysis and Mining
(ASONAM).DOI:10.1109/asonam.2016.7752387

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


150
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195137

Blockchain and its Role in the Internet of Things (IoT)


Tanweer Alam
Department of Computer Science, Faculty of Computer and Information Systems, Islamic University of
Madinah, Saudi Arabia
Email: tanweer03@iu.edu.sa

ABSTRACT
Blockchain (BC) in the Internet of Things (IoT) is a novel technology that acts with decentralized, distributed,
public and real-time ledger to store transactions among IoT nodes. A blockchain is a series of blocks, each block
is linked to its previous blocks. Every block has the cryptographic hash code, previous block hash, and its data.
The transactions in BC are the basic units that are used to transfer data between IoT nodes. The IoT nodes are
different kind of physical but smart devices with embedded sensors, actuators, programs and able to communicate
with other IoT nodes. The role of BC in IoT is to provide a procedure to process secured records of data through
IoT nodes. BC is a secured technology that can be used publicly and openly. IoT requires this kind of technology
to allow secure communication among IoT nodes in heterogeneous environment. The transactions in BC could
be traced and explored through anyone who are authenticated to communicate within the IoT. The BC in IoT
may help to improve the communication security. In this paper, I explored this approach, its opportunities and
challenges.
Keywords : Blockchain, Internet of Things (IoT), Cryptography, Security, Communication.

I. INTRODUCTION other and every device has its previous device address.
The blockchain and IoT together work in the
The IoT is growing exponentially year by year with its framework of IoT and Cloud integration. In the future,
aim in 5G technologies, like Smart Homes and Cities, the BC would revolutionize the IoT communication
e-Health, distributed intelligence etc. but it has [1]. The goals of BC and IoT integration could be
challenges in security and privacy. The IoT devices are summarized as follows.
connected in a decentralized approach. So, it is very
complex to use the standard existing security
techniques in the communication among IoT nodes.
The BC is a technology the provide the security in
transactions among the IoT devices. It provides a
decentralize, distribute and publicly available shared
ledger to store the data of the blocks that are processed
and verified in an IoT network. The data stored in the
public ledger is managed automatically by using the
Peer-to-peer topology. The BC is a technology where
transactions fired in the form of a block in the BC
among IoT nodes. The blocks are linked with each
Figure 1 : Blockchains and IoT

CSEIT195137 | Received : 05 Jan 2019 | Accepted : 18 Jan 2019 | January-February -2019 [ 5 (1) : 151-157 ]
151
Tanweer Alam Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 151-157

i) Decentralized framework: This approach is similar in exchanging a document with privacy without storing
IoT and BC. It is removed the centralized system and any information on the time-stamping service. The
provide the facility of a decentralized system. It idea of blockchains comes from [3] but the first
improves the failure probability and performance of blockchains were presented by Satoshi Nakamoto in
the overall system. 2008. He presented a paper where the blocks were
ii) Security: In the BC, the transactions among nodes added in a chain and form a blockchain [4]. In the
are secured. It is a very novel approach for secure article [5], the authors presented the “IoTChain” for
communication. BC allows IoT devices to authentication of information exchanged between two
communicate with each other in a secure way. nodes in an IoT network. They have presented an
iii) Identification: In IoT, all the connected devices are algorithm to exchange the information in IoT and
uniquely identified with a unique ID. Every block in blockchains (fig 2) [5]. In this paper, authors are
BC is also uniquely identified. So, BC is a trusted focused on the authorization part of the security in the
technology that provides uniquely identified data IoTChain framework.
stored in public ledger.
iv) Reliability: IoT nodes in BC have the capabilities to
authenticate the information passed in the network.
The data is reliable because it is verified by the miners
before entering in BC. Only verified blocks can enter
in the BC.
v) Autonomous: In BC, all IoT nodes are free to
communicate with any node in the network without
the centralized system.
vi) Scalability: In BC, the IoT devices will Figure 2 : IoT Chain framework
communicate in high-available, a distributed
intelligence network that connects with destination In the article [6], the authors explored the cloud and
device in a real-time and exchange information. MANET framework to connect the smart devices in
the internet of things and provide communication
The rest of the paper is summarized as follows: section security. In the article [7], authors represent a very
1 represents the introduction of the paper, section 2 nice framework called internet-cloud framework, it is
represents the literature survey, section 3 introduces a good idea to provide secure communication to the
the role of BC in IoT, section 4 represents the IoT devices. In the article [8], the authors provide a
opportunities of the integrated approach, section 5 middleware framework in the cloud-MANET
represents the challenges and section 6 represents the architecture for accessing data among the IoT devices.
conclusion. Article [9,10] represents the reliability in the
communication among IoT nodes. The articles
II. LITERATURE SURVEY [11,12,13,14,15] are providing the mobility models for
communication in 5G networks. In the article [16], the
The security and privacy in the communication among fuzzy logic-based mobility framework is explained for
IoT devices paid too much attention in the year of communication security. In the article [17], a nice
2017 and 2018. Several papers are published during the survey on blockchains and IoT done by the researchers.
year 2017 and 2018. In the year of 1990, Stuart Haber They present the idea of the security in the BC-IoT to
and W. Scott Stornetta were written an article [3] on develop the IoT apps with the power of BCs.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


152
Tanweer Alam Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 151-157

III. THE ROLE OF BC IN IoT publicly accessed, privacy is low to high and the
transferable identities while in a centralized database,
The IoT enables the connected physical things to the properties are centralized trust model, low in
exchange their information in the heterogeneous security, low publicly accessed, privacy is high and
network [18]. The IoT could be divided into the non-transferable identities. From the above properties,
following sections. the blockchain is more advanced than the centralized
1. Physical Things: The IoT provide the unique id for storage.
each connected thing in the network. The physical
things are able to exchange data with other IoT nodes.
2. Gateways: The gateways are the devices work
among physical things and the cloud to ensure that the
connection is established and security provided to the
network.
3. Networking: it is used to control the flow of data and
establish the shortest route among the IoT nodes.
4. Cloud: It is used to store and compute the data.
The BC is a chain of verified and cryptographic blocks (a)
of transactions held by the device connected in a
network. The blocks data are stored in the digital
ledger that is publicly shared and distributed. The BC
provides secure communication in IoT network. The
blockchain can be a private, public or consortium with
different properties. The following table represents
the differentiation among all kind of blockchains.

Table 1 : Kinds of Blockchains and their properties (b)


immovableness
Accord growth
BC/ Properties

Decentralized

Determining
Efficiency

Reading

Private good No yes Can Can Only


BC be be one
publi industr
cly y (c)
Public worse Yes no No publi All
Figure 3 : (a) Centralized (b) Decentralized (c)
BC cly miners
Consorti good Some yes Can Can IoT Distributed
um BC times be be devices
publi The following platforms are used to develop IoT
cly
applications using blockchain technology.
The database in blockchains has the properties such as
decentralized trust model, high security, highly

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


153
Tanweer Alam Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 151-157

a. IOTA: The IOTA is the new platform for the devices can communicate in the network and every
blockchain and IoT called Next generation block of the transaction will first verify by the miners
blockchains. This platform facilitates the high data then they can enter in the BC.
integrity, high performance of transactions and high 2. Reduce the Cost: This approach will reduce the cost
validity of blocks with using fewer resources. It because it communicates directly without the third
resolves the limitations of blockchains [19]. party. It eliminates all the third-party nodes between
b. IOTIFY: It provides web-based internet of things the sender and the receiver. It provides direct
solution to minimize the limitations of blockchains communication.
technology in the form of custom apps [20].
c. iExec: It is an open source blockchain based tool. It
facilitates your apps the decentralized cloud Trust

advantages [21].
d. Xage: It is the secure blockchain platform for IoT to Risk Management Cost Reduction

increase automation and secure information [22].


e. SONM: It is a decentralized blockchain based fog
Opportunities
computing platform to provide secure cloud services. in BC-IoT

The IoT and blockchains are increasing the business Financial Services Time Reduction

opportunities and opening the new markets where


everyone or everything can communicate in a real-
time with authenticity, privacy and security in a Social Services
Security and
Privacy

decentralized approach. The integration of these novel


technologies will change the current world where the
devices will communicate without the humans in
Figure 4 : Opportunities in BC-IoT
various stages. The objective of the framework is to get
the secured data on the right location, on the right
3: Reduce Time: This approach is reduced the time a
format, at real-time. The BC could be used to track
lot. It reduces the time taken in transactions from days
billions of IoT connected things, coordinate these
to second.
things, enabling the processing of the transactions,
4: Security and Privacy: It provides security and
resolving or eliminating the failures and making the
privacy to the devices and information.
flexible ecosystem for running the physical things on
5. Social Services: This approach provides public and
it. Hashing techniques are used in blocks of data by BC
social services to the connected devices. All connected
to create information privacy for the users.
devices can communicate and exchange information
between them.
IV. OPPORTUNITIES
6. financial Services: This approach transfer funds in a
secure way without the third party. It provides fast,
The BC-IoT integration approach has a lot of
secure and private financial service. It reduced transfer
remarkable opportunities. It opens the new doors for
cost and time.
both together. Some of the opportunities are described
7. Risk management: This approach is played the
as follows.
important roles to analyze and reduce the risk of
1. Building the Trust between parties: The BC-IoT
failing the resources and transactions.
approach will build trust among the various connected
devices because of its security features. Only verified

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


154
Tanweer Alam Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 151-157

V. CHALLENGES 6. Interoperability: The BC can be public or private. So,


the interoperability between public and private
The IoT and BC could face a lot of challenges such as blockchains is also a challenge in the BC-IoT approach.
scale, store, skills, discover etc. The following are the 7. Rules and Regulation: The IoT-BC will act globally,
challenges faced by the integration approach. so it faces many rules and regulations for
1. Scalability: The BC can become hang because of its implementing this approach globally.
heavy load of the transaction. The Bitcoin storage is
becoming more than 197 GB storage in 2019 [24]. VI. CONCLUSION
Imagine if IoT integrates with BC then the load will be
heavier than the current situation. The BC and IoT is a novel approach explored in this
2. Storage: The digital ledger will be stored on every article. Many opportunities and challenges are
IoT node. By the time, it will increase in its storage size described. Also, available platforms are listed in this
that will be a challenging task and become a heavy article. This approach can be the future of the internet
load on each and every connected device. because it can overhaul the current internet system
3. Lack of Skills: The BC is a new technology. It is and change it with the new one where every smart
known by very few people in the world. So, it is also a device will connect to other devices using the peer-to-
challenge to train the people about the technology. peer network in a real-time. It can reduce the current
cost and time and provide the right information to the
Scalability right device in a real-time. So, it can be very useful in
the future.
Rules and
Storage
Regulation

VII. ACKNOWLEDGMENT

Challenges
in BC-IoT This research is supported by Deanship of Scientific
Interoperability Lack of Skills
Research, Islamic University of Madinah, Kingdom of
Saudi Arabia. The grant number is 10/40.

Discovery and
Privacy
Integration VIII. REFERENCES

Figure 5 : Challenges in BC-IoT [1]. Reyna, Ana, et al. "On blockchain and its
integration with IoT. Challenges and
4. Discovery and Integration: Actually, BC is not opportunities." Future Generation Computer
designed for IoT. It is a very challenging task for the Systems (2018). DOI:
connected devices to discover another device in BC https://doi.org/10.1016/j.future.2018.05.046
and IoT. So, IoT nodes can discover each other but [2]. Zheng, Zibin, et al. "Blockchain challenges and
they can be unable to discover and integrate the BC opportunities: A survey." International Journal
with another device. of Web and Grid Services 14.4 (2018): 352-375.
5. Privacy: The ledger is distributed publicly to every DOI:
connected node. They can see the ledger transactions. https://doi.org/10.1504/IJWGS.2018.095647
So, privacy is also a challenging task in the integrated [3]. Haber, Stuart, and W. Scott Stornetta. "How to
approach. time-stamp a digital document." Conference on

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


155
Tanweer Alam Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 151-157

the Theory and Application of Cryptography. [12]. Alam, Tanweer, Parveen Kumar, and Prabhakar
Springer, Berlin, Heidelberg, 1990. Singh. "SEARCHING MOBILE NODES USING
[4]. Nakamoto, Satoshi. "Bitcoin: A peer-to-peer MODIFIED COLUMN MOBILITY MODEL.",
electronic cash system." (2008). International Journal of Computer Science and
[5]. Alphand, Olivier, et al. "IoTChain: A blockchain Mobile Computing, (2014).
security architecture for the Internet of Things." [13]. Alam, Tanweer, and B. K. Sharma. "A New
Wireless Communications and Networking Optimistic Mobility Model for Mobile Ad Hoc
Conference (WCNC), 2018 IEEE. IEEE, 2018. Networks." International Journal of Computer
[6]. Alam T, Benaida M. The Role of Cloud-MANET Applications 8.3 (2010): 1-4. DOI:
Framework in the Internet of Things (IoT). https://doi.org/10.5120/1196-1687
International Journal of Online Engineering [14]. Singh, Parbhakar, Parveen Kumar, and Tanweer
(iJOE). 2018;14(12):97-111. DOI: Alam. "Generating Different Mobility Scenarios
https://doi.org/10.3991/ijoe.v14i12.8338 in Ad Hoc Networks.", International Journal of
[7]. Alam T, Benaida M. CICS: Cloud–Internet Electronics Communication and Computer
Communication Security Framework for the Technology, 4(2), 2014
Internet of Smart Devices. International Journal [15]. Sharma, Abhilash, Tanweer Alam, and Dimpi
of Interactive Mobile Technologies (iJIM). 2018 Srivastava. "Ad Hoc Network Architecture
Nov 1;12(6):74-84. DOI: Based on Mobile Ipv6 Development." Advances
https://doi.org/10.3991/ijim.v12i6.6776 in Computer Vision and Information
[8]. Alam, Tanweer. "Middleware Implementation Technology (2008): 224.
in Cloud-MANET Mobility Model for Internet [16]. Alam, Tanweer. "Fuzzy control based mobility
of Smart Devices", International Journal of framework for evaluating mobility models in
Computer Science and Network Security, 17(5), MANET of smart devices." ARPN Journal of
2017. Pp. 86-94 Engineering and Applied Sciences 12, no. 15
[9]. Tanweer Alam, "A Reliable Communication (2017): 4526-4538.
Framework and Its Use in Internet of Things [17]. Conoscenti, Marco, Antonio Vetro, and Juan
(IoT)", International Journal of Scientific Carlos De Martin. "Blockchain for the Internet
Research in Computer Science, Engineering and of Things: A systematic literature review."
Information Technology (IJSRCSEIT), Volume Computer Systems and Applications (AICCSA),
3, Issue 5, pp.450-456, May-June.2018 URL: 2016 IEEE/ACS 13th International Conference
http://ijsrcseit.com/CSEIT1835111 of. IEEE, 2016.
[10]. Alam, Tanweer. (2018) "A reliable framework [18]. Gubbi, Jayavardhana, et al. "Internet of Things
for communication in internet of smart devices (IoT): A vision, architectural elements, and
using IEEE 802.15.4." ARPN Journal of future directions." Future generation computer
Engineering and Applied Sciences 13(10), 3378- systems 29.7 (2013): 1645-1660. DOI:
3387 https://doi.org/10.1016/j.future.2013.01.010
[11]. Alam, Tanweer, Arun Pratap Srivastava, [19]. https://www.iota.org
Sandeep Gupta, and Raj Gaurang Tiwari. [20]. https://iotify.org
"Scanning the Node Using Modified Column [21]. https://iex.ec/overview
Mobility Model." Computer Vision and [22]. https://xage.com
Information Technology: Advances and [23]. https://www.i-scoop.eu/blockchain-distributed-
Applications 455 (2010). ledger-technology/blockchain-iot

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


156
Tanweer Alam Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 151-157

[24]. https://www.statista.com/statistics/647523/worl
dwide-bitcoin-blockchain-size

Cite this article as :


Tanweer Alam, “Blockchain and its Role in the Internet of Things (IoT)”, International Journal of Scientific Research in
Computer Science, Engineering and Information Technology, pp. 151-157, 2019. DOI:
https://doi.org/10.32628/CSEIT195137

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


157
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195138
A Review on Various Payment and Billing Techniques
Alka D. Aswar1, Diksha M. Chhipane1, Dipali C. Sontakke1, Pallavi P. Onkar1, Snehal S. Helonde1, Prof. S. V. Sonekar2
1BE Scholar, Department of Information Technology, J.D. College of Engineering & Management, Nagpur, Maharashtra,
India
2Associate Professor, Department of Information Technology, J.D. College of Engineering & Management, Nagpur,
Maharashtra, India
ABSTRACT

Organizations have been compelled to change the manner in which they lead business to react to the ongoing
changes occurring both at the worldwide, provincial and at the national dimension. The motivation behind this
examination was to explore through an overview the electronic billing and payment framework. The premise
of the issue in regards to electronic billing and payment framework that the exploration examine planned to
investigate was talked about dependent on two research destinations; to set up the advantages acknowledged
with the dispatch electronic billing and payment, and to survey difficulties in the execution of electronic billing
and payment.
Keywords : Electronic Billing, Payment System, E-Billing, Online Payment

I. INTRODUCTION change are well on the way to get by in future. As it


states, "On the off chance that you don't care for
E-billing and e-payment frameworks are a piece of e- change, you will like superfluity even less" Some of
business. Electronic Business (e-business) might be the progressions saw in various establishments
characterized comprehensively as any business program, rebuilding, cost-cutting, expansion and the
procedure that depends on a robotized data adjustment of Information and Communication
framework. Electronic business techniques empower Technologies (ICTs).
organizations to connect their inner and outside
information preparing frameworks all the more In Information and Communication Technology, the
efficient and adaptable to work all the more progressions are quick. The fast changes in payment
intimately with providers and accomplices and to frameworks are making a wide exhibit of new
better fulfill the requirements and desires for their business openings and difficulties. One of the open
clients. Today, this is for the most part accomplished doors is electronic billing (e-billing) and electronic
utilizing web-based innovations and portable payment (e-payment). The business needs of KPLC
applications. clients are changing as purchasers grasp new
advances. The Company has left on a huge extension
Organizations have been compelled to change the process through the Rural Electrification Program
manner in which they direct business so as to react to subsidized incompletely by improvement
the ongoing changes occurring both at the worldwide, accomplices, for example, IMF and World Bank. The
territorial and at the national dimension. This is extra number of shoppers has made requirements in
unavoidable as organizations that rush to receive to

CSEIT195138 | Received : 25 Jan 2019 | Accepted : 03 Feb 2019 | January-February -2019 [ 5 (1) : 158-162 ]
158
Alka D. Aswar et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 158-162

the KPLC organize prompting power outages and a portrayal of the information which is filtered and
drop in the nature of administration. deciphered for data. Scanner tag is littler, lighter and
more affordable. The standardized identification must
With the progression in innovation, there are be in the immediate viewable pathway to be perused
different new advances or frameworks that outcome by scanner tag scanner. Scanner tag have no
in the smart billing and stock administration of the perused/compose capacities thus they can't contain
obtained things to spare the time required for the data for example, item expiry date and so on. The
billing and furthermore stock or stock administration scanner tags can be effectively repeated and produced
[1] [2]. The bill is created naturally when the user those outcomes in less security. In the event that the
distinguishes item tag and the data put away with the scanner tag is tore, harmed there's no real way to
tag. The data can be the item name and the expense examine the item.
of the item. The stock administration should be
possible by creating and interfacing the product with
these innovations. At the point when the user
peruses data from the label, it shows the expense and
amount of the item which is being acquired by the
client. After checkout, the bought things are
deducted from the fundamental database server
which deals with the stock and amount. These new
advances are as a rule exceptionally valuable in
Fig (1). Barcode
today’s life by diminishing the heaps of endeavours
and human work. There are numerous such
The standardized tag based information gathering
advancements like Barcode, QR Code, RFID, OCR.
framework can be regularly included any of the
The innovations utilized in current billing
segments like Barcode scanner; scanner tag based
framework. In this investigation, we talk about the
portable PC, standardized tag printers, scanner tag
different existing system for smart or e-billing.
name, standardized tag information accumulation
programming and so on. The PC looks into the cost
II. Review of Literature
in an ace database (the cost of an item isn't in the
standardized tag!), subtracts it from the store stock,
A. Scanner Tag
and ascertains the change. The product additionally
Scanner tag is otherwise called the Auto ID
makes reports with respect to stock dimensions,
(Automatic Identification). It was designed in the
demonstrates what items are the most and least
mid-1970s. The scanner tag was made to help huge
famous, makes statistic provides details regarding
retail locations and markets process their
singular items and clients, and tracks substantially
merchandise. The standardized identification
more. The way to the entire framework is exact
innovation is very effective for the administration
revealing of the item bought. Clerks are innately
and data refreshing in the PCs [1]. Standardized tag is
questionable and moderate. Barcoding is not one or
one of the recognizable business models. It is utilized
the other. Standardized identification benefits in
for examining the things at grocery store for
information exactness, consistency and effectiveness.
checkouts [2]. The standardized tag peruser is utilized
Standardized identification is utilized in Retail
to distinguish the tag on things and recognizes the
Operations, Receiving and dispatching Operations,
item and related data. Scanner tag result in a visual

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


159
Alka D. Aswar et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 158-162

resource the executives, producing tasks, Office and At the point when QR code is checked it connects to
Customer Service Applications and Warehousing and plain content, picture, website pages and interactive
so forth. Standardized identifications are estimated by media, maps areas or anything that can be gotten to
the width of the thin bar and are recorded in mils, or over the web. QR code generator is required to make
1/1000 inch. A 15 Mil standardized identification, for a QR code in which the required data is entered. A
example, has a restricted bar that is 15/1000 inches viable QR code ought to be created utilizing 40-80
wide. Further, "calm zones," or clear spaces to one characters and least of 2cm wide. The more
side and right of standardized identification images, characters result in the debased picture. QR code
are incorporated to safeguard the scanner tag can be enables us to choose the activities we need the clients
perused. to take. The QR code is ISO standard, advertise
demonstrated and totally quantifiable. QR codes
B. QR Code enable the normal individual to decipher the QR
code by examining and perused by a camera-prepared
A QR code is Quick Response code. This code is a 2D Smartphone when you’ve downloaded a scanner
standardized identification bringing about higher application [3].
speed at disentangling substance. The Japanese
company Denso-Wave made the QR code for C. OCR (Optical Character Recognition)
following reason in 1994. QR Codes utilize a 2
dimensional network so as to store information. The OCR is an innovation that empowers you to change
more lines and sections the more information it holds. over various kinds of archives, for example, checked
It underpins the mistake rectification. The code can paper reports, PDF records or pictures caught by an
be examined utilizing cell phone camera or QR advanced camera into editable and accessible
peruser and QR code peruser programming. There information. Albeit Optical Character Recognition
are 4 diff dimension of mistake redress low mid-range (OCR) filtering innovation has expanded quickly
quality high. The most elevated amount of mistake throughout the years, there are constraints with
rectification can re-establish up to 30% of a respect to the source materials and character
degenerate document. designing. Content from a source with a text
dimension of under 12 will result in more mistakes.
OCR additionally requires a spell checking editing
and reformatting for the ideal format. There are
numerous components that will influence the OCR,
for example, shapes, little messages, hazy duplicates,
scientific recipes, shaded paper, and written by hand
massage.

D. RFID

Presently days, radio recurrence ID innovation has


picked up the position from lack of clarity to standard
applications that outcomes in the speed of taking care
Fig(2). QR Code
of fabricated products and materials. RFID can
distinguish the labels utilizing RFID peruser from

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


160
Alka D. Aswar et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 158-162

separation without the viewable pathway. RFID they need to hang tight in the line for longer time.
bolster the immense measure of one of a kind ids that The thought which is proposed utilizing RFID
helps in making the one of a kind labels for innovation will conquer the issue and it gets the
everything. It permits contactless identification of the undertaking less demanding. The joined impacts of
items utilizing the radio recurrence. RFID label simple and adaptable usage, secure transmission of
arrangement is less compelled as it doesn't require record data, and diminished debate offer the
viewable pathway. It has most extreme and more accompanying advantages for all. It will spare time,
grounded read go. RFID bolsters the read/compose vitality and labor of Customer, Owner and provider.
memory capacity and it can hold substantial measure There are numerous advances which are right now
of information under a special identifier. The being utilized for billing frameworks in general stores.
unfriendly conditions, for example, dust, substance, The choice of the innovation relies on the execution,
physical harm don't influence on RFID without a proficiency and QoS of the innovation in regards to
doubt. This innovation has three sections: a filtering specific undertaking and condition.
reception apparatus, a handset with a decoder to
translate the information and a transponder (RFID IV.REFERENCES
tag) pre-set with data. RFID labels can be perused at a
quicker rate as they roughly can peruse 40 RFID [1] T. Sriram, K. Vishwanatha Rao, S. Biswas and
labels at same time. RFID contain abnormal amounts B. Ahmed "Applications of barcode technology
of security; information can be encoded, secret key in automated storage and retrieval systems",
ensured or set to incorporate an "execute" highlight Industrial Electronics, Control, and
to evacuate information for all time [4]. Instrumentation, 1996., Proceedings of the
1996 IEEE IECON 22nd International
The Tag or Transponder gets recognized by the Conference on, vol. 1, pp.641 -646
Reader or Interrogator when it goes inside the [2] Alistair Milne “The Rise and Success of the
receiving wire extend and the distinguished novel id Barcode: some lessons for Financial Services”
is sent to the Computer or server. feb 2013
[3] Andrew Kerry-Bedell IT‟s in Conservation
III.Comparative Analysis Consultants. Mobile: 07899 741939 email:
a.kb@live.co.uk
http://www.kbstconsulting.co.uk/QR/images/Q
R.pdf
[4] Mandeep Kaur, Manjeet Sandhu, Neeraj Mohan
and Parvinder S. Sandhu “RFID Technology
Principles, Advantages, Limitations & Its
Applications” International Journal of
Computer and Electrical Engineering, Vol.3,
No.1, February, 2011
[5] Abwao, V. (2012). Information Technology
The payment of bill by remaining in long line is a
Applications in Business Management within
monotonous factor when individuals need to buy
Kenyan Companies; A Survey of Selected
items from shops. In spite of the fact that individuals
Insurance Firms in Nairobi, Unpublished MBA
can pay immediately utilizing electronic cash office,
Thesis, University of Nairobi.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


161
Alka D. Aswar et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 158-162

[6] Andrew, J, W. (2010). Electronic bill payment


and presentment: Jump-starting the next great
Web enabled consumer empowerment engine,
Robertson Stephens e-Processing Research,
April 11.
[7] Atebe, E. (2011). An Effect of Business Process
Reengineering on Business Process Cylce. The
Case of KLPC, Unpublished MBA Thesis,
University of Nairobi.

Cite this article as :

Alka D. Aswar, Diksha M. Chhipane, Dipali C.


Sontakke, Pallavi P. Onkar, Snehal S. Helonde, Prof.
S. V. Sonekar, "A Review on Various Payment and
Billing Techniques", International Journal of
Scientific Research in Computer Science, Engineering
and Information Technology (IJSRCSEIT), ISSN :
2456-3307, Volume 5 Issue 1, pp. 158-162, January-
February 2019. Available at doi :
https://doi.org/10.32628/CSEIT195138
Journal URL : http://ijsrcseit.com/CSEIT195138

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


162
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195128

Finite/Infinite Queueing Models Performance Analysis in


Optical Switching Network Nodes
Waleed M. Gaballah*
* Department of Electronics and Communication Engineering, Mansoura High Institute of Engineering and
Technology, Mansoura, Egypt

ABSTRACT

In optical switching networks, optical information (packets/bursts) that are forwarded from an optical switch to
another through the network could entering a queue of a certain length in each node and may be waiting inside
this node before it will be transmitted to the next node. Various queueing models have been widely applied as
the main tool in optical switching networks for modeling and performance evaluation analysis of the switching
nodes. In this paper, studying the performance analysis of an optical switching node is discussed at Finite and
Infinite queueing models to have the optimum queueing model for optical core node switch design. The
waiting delay time, the average expected number of optical packets and the loss probability of packets in the
optical switch are estimated at variable traffic loads and different wavelength channels.
Keywords : Queueing Models, Optical Switching Network, Finite Queue, Infinite Queue, Blocking Probability

I. INTRODUCTION time. In the switch if the wavelength channel is busy


serving other packets, the arrivals are queued in the
An optical information traffic access to the network switch queueing buffer. Therefore, the model can
with a given capacity is modeled in optical nodes as a determine the distribution of the number of packets
queue with a certain distribution of traffic arrival in the system and their waiting time.
times and a certain distribution of traffic service time.
Queueing models in the optical switching networks
Queueing in the network nodes are modeled in a have various application aspects. Queueing models
variety of network techniques and principles [1, 2]. with optical delay lines in optical packet switching
The Queueing model is a theoretical aspect of such networks are studied in [4]. The edge OBS node
the network node [3]. Typically, a queueing model queueing is modelled in [5]. An analytical model
represents the network node physical configuration optical delay line buffers in OBS networks using
by determining the number of optical packets in the queueing theory are developed in [6].
switch and how fast that switch serves the optical
packet traffic. Also, the queueing models give a In this paper, study the performance analysis of the
statistical nature of the optical switch node, by optical switching network node is performed at
specifying the variability in the arrival process to the different queueing models. Validating that the
switch and in the switch service process. In an optical various queueing model parameters and the blocking
switching, optical packets arrive at a system (switch) probability are affected by the wavelength channels
as random intervals and are served during a random number and traffic loads.

CSEIT195128 | Received : 25 Jan 2019 | Accepted : 03 Feb 2019 | January-February -2019 [ 5 (1) : 163-170 ]
163
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

For this, our work first presents some specific capacity. In finite queue, the optical arrivals that
queueing models overview that can be used in optical attempt to enter the full-occupied system are denied
switching networks, at section 2. The infinite and entry or blocked.
finite queueing models’ numerical analysis is
obtained in section 3. Finally, conclude in section 4. Table 1 represents the main differences between
Finite and Infinite queueing models.
II. Queueing Models in Optical Switches
TABLE I
FINITE VS. INFINITE QUEUEING MODELS
At the optical switches, the arriving optical
information traffic request some specific amount of Infinite Queueing
Finite Queueing Model
resources such as; circuit, bandwidth, wavelength Model
channel, etc. to be served. The most common M/M/w/m M/M/w
queueing models assume that the optical information The arrival rate The arrival rate is not
traffic inter-arrival and service times follow the depends on the number affected by the number
exponential distribution or equivalently follow a of served and waiting of packets being served
Poisson distribution process with Markovian or packets in the system. and waiting.
memoryless properties [7]. A commonly used Limited buffer capacity. Unlimited buffer
shorthand notation, called Kendall’s notation [8], for capacity.
such queue models describes the arrival process, Faster and has lower Packets waiting long
service distribution, the number of servers and the average number of times in the buffer.
buffer size (waiting line). The complete notation waiting packets in the
expressed as (a/b/c/d) where, Arrival process/service system.
distribution/ number of servers/waiting line. There is a packet loss No packet loss
probability. probability.
In optical switching networks, the commonly used
characters for the first two positions in the shorthand III. Infinite/Finite Queueing Models Performance
notations are M (Markovian – Poisson for the arrival Analysis
or Exponential for the service time). The third
positions used for the number of the output optical In this section, a numerical performance analysis of
wavelength channels w. The fourth position indicates infinite and finite queueing models in the optical
the switch queueing size m and it's usually not used switching node are represented. This analysis study
in infinity waiting room buffers. aimed to determine the suitable queuing model for
enhancement of an optical switch performance,
There are single server queueing models such as which queueing model has low loss probability,
M/M/1 and M/M/1/w, and multiple server's systems which one is faster and has lesser number of optical
such as M/M/w, M/M/w/m, and M/M/w/w systems packets in the optical switch system serviced and
[9]. There is an infinite queue system such as M/M/w, waiting for service. In our analysis taking in
where the optical traffic arrivals are hold waiting for consideration the effect of the number of optical
service and not affected by the number of packets switch wavelength channels w whether at Infinite
already on the queue because there is unlimited model M/M/w or Finite one M/M/w/m. The queueing
buffer size. In addition, there is a finite queue system models are used at suitable average arrival rate 
such as M/M/w/m, which has a limited buffer

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


164
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

packet/sec and average service rate µ packets/sec wavelengths (servers), w = 1, 2, 8, 16, 32. In Fig. 1,
values. the service rate per channel is constant µ = 1
The performance parameters that measured are: packet/time unit. While at figures 2 to 4, the entire
- The average number of optical packets resident in switching capacity wµ is kept constant by µ = 1/w
the system Ls packets. packets/time unit, which gives  = λ becomes
- The average number of optical packets waiting in independent on w.
the queue Lq packets.
- The average time of optical packets spend in the From Fig. 1 it is clear that for a lower offered load,
system (the average switch queueing delay time) the mean time in the system is very low and equal to
Ws sec. 1/µ = 1 time unit. As the load increases, in the M/M/1
- The average time of optical packets waiting in the queue, the mean total time in the optical switch
queue (the average waiting time to serviced) Wq increases greatly and increased slightly as the number
sec. of wavelengths w increased. At the M/M/32 queue,
- The blocking probability of optical packets PB (at the mean total time in the system only has a slightly
finite queueing models). increment. Therefore, as the number of wavelengths
increased the queue has a superior performance than
In this analysis, the waiting time in the system Ws
M/M/1 queue.
established at two different cases [10]. First, if it is
considered that the service rate per channel is
constant µ = 1 packet/sec, which gives  = λ/w. Later, 4
M/M/1
Average waiting time in the switch Ws time unit

the total service rate in the switch is kept constant M/M/2


wµ = 1, which gives  = λ becomes independent on w. M/M/8
3 M/M/16
At Infinite model we will study the influence of the Ws1(  )
M/M/32
Ws2(  )
wavelength channel numbers at the different two
Ws3(  )
analysis cases. At Finite model the effect of the Ws4(  )
2

switch queueing size and the number of wavelength Ws5(  )

channels are illustrated on the optical switch 1


performance taking in consideration the second
analysis case.
0
0 0.2 0.4 0.6 0.8 1
A. Infinite Queueing Models 
Offered Load 
Figure 1. Average waiting time in the system vs. the
offered load of M/M/w with different number of
With infinite queuing systems M/M/w, where w is
wavelength channels and μ = 1 packet/time unit per
the number of servers or optical wavelength channels,
channel
the queue buffer size is infinite. The system filling Ls
and the waiting times Wq and Ws, in the queue and If the switching capacity is constant, figures 2 to 4 are
the optical switch respectively, establish the main illustrated. Obviously, at low loads, less than 80%
optical switch characteristics. load or  < 0.8, the Waiting time in the optical switch
(mean flow time), Ws = Wq + 1/μ , shown is ruled by
In figures from 1 to 4, the optical switch performance the increased holding time, 1/μ = w, which results
is studied at an infinite queueing models. Ws, Wq and from decreasing the service rate for increased
Ls parameters are represented at different number of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


165
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

wavelength numbers w in order to keep the

Average number of packets in the switch Ls (packets)


50
switching capacity constant. w=1
w=2
40 w=8
At high loads the mean time in the optical switch Ls1(  ) w = 16
Ls2(  ) w = 32
increases significantly because the waiting time in 30
Ls3(  )
the queue component Wq becomes dominant. This
Ls4(  )
increase in queueing waiting time also causes the Ls5(  )
20

increase in the system filling (Number of packets in


10
the system) Ls shown in Fig. 3. However, this only
results from the longer serving interval required per 0
0 0.2 0.4 0.6 0.8 1
server and not from an increase in the mean waiting 
Offered Load 
queueing time Wq. Figure 4 shows the mean waiting
time in the switch queue Wq. It is decreased Figure 3. Average number of packets in the system vs.
dramatically with increasing wavelength numbers w, the offered load of M/M/w with different number of
and the service rate per channel be slower. So, in this wavelength channels and normalized to equal system
case, it is clear that a single wavelength queueing load  =  by setting μ = 1/w packet/time unit
model is preferred over a multi-wavelength model
Average waiting time in the queue Wq time unit

because it has less average number of packets resident 0.2

in the system and lower mean time in the system.


0.15
Wq1(  )

Wq2(  )
Average waiting time in the switch Ws time unit

M/M/1 Wq3(  )
0.1
Wq4(  )
M/M/2
40 Wq5(  )
Ws1( ) M/M/8 w =1
Ws2( )
M/M/16 0.05 w =2
30 M/M/32 w =8
Ws3( ) w = 16
w = 32
Ws4( ) 0
20 0 0.2 0.4 0.6 0.8 1
Ws5( ) 
Offered Load 
10

Figure 4. Average waiting time in the queue vs. the


0
0 0.2 0.4 0.6 0.8 1 offered load of M/M/w with different number of
Offered Load  wavelength channels and normalized to equal system
load  =  by setting μ = 1/w packet/time unit
Figure 2. Average waiting time in the system vs. the
offered load of M/M/w with different number of In a realistic view, if we consider that the switching
wavelength channels and normalized to equal system capacity is constant, the best service is provided for
load  =  by setting μ = 1/w packet/time unit per the lower number of wavelengths possible. While, if
channel the service rate per channel is constant, increasing a
system wavelength channels is preferred.

Consequently, if the costs of using more wavelengths


switch less, constant service rate per channel is more
preferred choice. However, constant switching
capacity is better at lower wavelengths number. An

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


166
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

optical switch with an adequately chosen number of Figure 5. Average number of packets in the system
wavelengths is the better choice. vs. the offered load of M/M/w/m with different
switch buffer size m and w = 8 wavelengths
B. Finite Queueing Models

Average waiting time in the switch Ws time unit


In finite queueing model systems M/M/w/m, an 80
m=8
arriving optical packet may be admitted to the free m = 16
wavelength channel immediately. It may be placed in 60
m = 32
Ws1(  )
m = 64
the queue until a wavelength channel is available, or
Ws2(  )
it may be blocked due to all wavelengths are busy Ws3(  )40

and all buffer places are occupied. Ws4(  )

20

Figures from 5 to 8 illustrate how the finite queueing


M/M/w/m performance depends on the system size m. 0
1 1.2 1.4 1.6 1.8
These figures show the mean time in the optical 
Offered Load 
switch Ws, the mean waiting time in the switch
Figure 6. Average waiting time in the switch vs. the
queue Wq, the mean optical packet number in the
offered load of M/M/w/m with different switch
switch Ls and the blocking probability PB all over the
buffer size m and w = 8 wavelengths
increasing system load  of M/M/w/m queueing
systems. The analysis is using eight wavelengths, w =
Average waiting time in the queue Wq time unit

8, and the switch buffer size m = (8, 16, 32, 64). 40


Impact of the system size m on the optical switch
declared at high traffic loads  > 1. As the system Wq1(  ) 20

buffer size increases, with a constant wavelength Wq2(  )

Wq3(  )
number, the more load can be buffered and that may 0
Wq4(  )
increase the mean optical packets number in the m=8
m = 16
system (switch filling) Ls, Fig. 5. Therefore, it is − 20
m = 32
increasing the waiting time in the queue and in the m = 64

system Wq and Ws, figures 6 and 7 respectively. 1 1.2 1.4 1.6 1.8

Offered Load 
Average number of packets in the switch Ls (packets)

100 Figure 7. Average waiting time in the queue vs. the


offered load of M/M/w/m with different switch
buffer size m and w = 8 wavelengths
Ls1( )
10
Ls2( )
However, this cause a heavily decreases in the
Ls3( )

Ls4( ) switching blocking probability PB as in Fig. 8.


1 m=8
m =16
m = 32
m = 64
0.1
1 1.2 1.4 1.6 1.8
Offered Load 

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


167
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

Figure 9. Average waiting time in the switch vs.


− 13
110 Average number of wavelengths for M/M/w/m with
Blocking Probability of the switch PB

− 27
110 different switch buffer size m at  = 
− 41
pk1(  ) 110
− 55
pk2(  ) 110
− 69
pk3(  ) 110 100

Average number of packets in the switch Ls


− 83
pk4(  ) 110
− 97 m=8
110
m = 16
− 11 1
110 m = 32 Ls1(  )
− 12 5 10
110 m = 64
Ls2(  )

(packets)
− 13 9
110
1 1.2 1.4 1.6 1.8 Ls3(  )
Offered Load 
Ls4(  )
1 m=8
Figure 8. Packets blocking probability PB in the m =16
m = 32
system vs. the offered load of M/M/w/m with m = 64
different switch buffer size m and w = 8 wavelengths 0.1
20 40 60
Average number of wavelengths w
For the special case m = w at queue model M/M/w/w,
no loads can wait for service, such that no waiting Figure 10. Average number of packets in the switch
time in the queue, Wq = 0. In this case, all loads that vs. Average number of wavelengths for M/M/w/m
cannot be served immediately become blocked, and with different switch buffer size m at  = 
consequently, these systems are commonly called loss
systems.
Average waiting time in the queue Wq time

40

To study the influence of the number of wavelength


channels on the queueing model performance, figures Wq1(  ) 20

Wq2(  )
from 9 to 12 show the performance of M/M/w/m
unit

Wq3(  )
switching system over the number of provided 0
Wq4(  )
wavelengths w. This study at high load  = 2, at m=8
different system size m values (m = 8, 16, 32, 64), and − 20
m = 16
m = 32
constant system capacity wμ = 1. m = 64
20 40 60

Average waiting time in the switch Ws time unit

100
Average number of wavelengths w

80 Figure 11. Average waiting time in the queue vs.


Ws1(  ) Average number of wavelengths for M/M/w/m with
Ws2(  ) 60
different switch buffer size m at  = 
Ws3(  )

Ws4(  ) 40
m=8
m = 16
20
m = 32
m = 64
0
0 20 40 60

Average number of wavelengths w

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


168
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

−4
110
Blocking Probability of the switch PB

− 36 0.1
110

Blocking Probability of the switch PB


−3
− 68 110
110
−5
110
pk1(  ) 110− 10 0
− 13 2 −7
110
pk2(  ) 110 pk1(  )
− 16 4 −9
pk2(  ) 110
pk3(  ) 110
− 19 6 pk3(  ) − 11
110
pk4(  ) 110
− 22 8 m=8 − 13
110 110
− 26 0
m = 16 − 15
110 110 M/M/8/8
m = 32
− 29 2 − 17 M/M/16/16
110 m = 64 110
M/M/32/32
− 19
0 110
0 20 40 60 80 0 0.5 1 1.5 2

Average numberof wavelengths w Offered Load 

Figure 12. Packets blocking probability PB in the Figure 13. Packet blocking probability PB in the
system vs. average number of wavelengths w of system vs. the offered load of M/M/w/w at different
M/M/w/m with different switch buffer size m at  =  servers and  = 

Again, for m = w no waiting space exists, and thus, It is clear that as the load increases, the blocking
the waiting time in the queue Wq equal zero in case probability also increases at low traffic and saturated
of M/M/8/8. In addition that the queue M/M/8/8 at high traffic. However, as the number of
gives the better performance, less average number of wavelength channels increases, the blocking
packets and low waiting time in the system. However, probability decreases because the incoming traffic has
the strong reduction of the blocking probability PB is a more chance to be serviced.
gained from the implicit queue size increase.
IV.CONCLUSION
Then, on finite queueing optical switches, the
blocking probability decreases at increasing the Optical switching network node performance
number of wavelengths (servers). analysis was done with Finite and Infinite queueing
models to indicate the optimum model at optical
Finally, Fig. 13 investigates the switch blocking switch design. Queue delay time, expected number of
probability PB for the M/M/w/w model at different packets in the system and blocking probability are
number of wavelength channels w = 8, 16, 32. The the main parameters in the analysis to demonstrate
switch performance illustrated at different traffic the optical switch performance. At Infinite queueing
loads  = 0 to 2. model, two different analysis cases were used to
represent the wavelength channel number impact on
the optical switch node performance. First analysis
was considered a constant service rate per channel
which is preferred if the costs of using more
wavelengths switch less. However, fixed switching
capacity analysis is better at lower wavelength
number. For Finite queueing model, impact of the
system size on the optical switch declared at high

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


169
Waleed M. Gaballah Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 163-170

traffic loads. As the system buffer size increases, with [5] Lisong Xu, Harry G. Perros and G. N. R., “A
a constant wavelength number, the more load can be Queueing Network Model of an Edge Optical
buffered and that may increase the mean optical Burst Switching Node,” in INFOCOM 2003.
packets number in the system. Therefore, it is Twenty-Second Annual Joint Conference of the
increasing the waiting time in the queue and in the IEEE Computer and Communications. IEEE
system. Also, the switching blocking probability PB is Societies, 2003, pp. 2019–2029.
heavily decreases. [6] Kumar, OS Sunish, C. S. Jaison, Indu Reena
Varughese and V. K. J., “The Analysis of M/M/r
As the number of wavelength channels increased Queuing Model for Fiber Delay Lines to
with fixed queue buffer size, the blocking probability Resolve BLP in OBS Networks,” In Emerging
increased due to the switch buffer capacity lowered Research Areas and 2013 International
until it reaches zero at M/M/w/w queueing system. Conference on Microelectronics,
At M/M/w/w queue system, it is faster and has less Communications and Renewable Energy
waiting packets in the switch, while it has high (AICERA/ICMiCR), 2013 Annual International
blocking probability due to no queue buffer. At this Conference, 2013, pp. 1–3.
model, as the number of wavelength channels [7] Bolch Gunter, Stefan Greiner, Hermann de
increases, the blocking probability decreases due to Meer, and Kishor S. Trivedi, Queueing
that the incoming traffic has a more chance to networks and Markov chains: modeling and
serviced. Therefore, using M/M/w/w queueing performance evaluation with computer science
system can be modelled well at optical switching applications, John Wiley & Sons, 2006.
nodes under a certain predefined number of [8] Harchol-Balter, Performance modeling and
wavelengths to lower the blocking probability design of computer systems: queueing theory in
problem. action, Cambridge University Press, 2013.
[9] Zukerman M., “Introduction to queueing
V. REFERENCES theory and stochastic teletraffic models,”
arXiv:1307.2968, 2018.
[1] Van Do, Tien, et al., eds., Queueing Theory and [10] Bhat, U. Narayan. An introduction to queueing
Network Applications, Springer International theory: modeling and analysis in applications,
Publishing, 2016. Birkhäuser, 2015.
[2] E. Yue, Wuyi, Quan-Lin Li, Shunfu Jin, and
Zhanyou Ma, “Queueing Theory and Network Cite this article as :
Applications,” in 12th International Waleed M. Gaballah, "Finite/Infinite Queueing
Conference, QTNA 2017, 2017. Models Performance Analysis in Optical Switching
[3] Thomopoulos, Nick T., Fundamentals of Network Nodes", International Journal of Scientific
Queuing Systems: Statistical Methods for Research in Computer Science, Engineering and
Analyzing Queuing Models, Springer Science & Information Technology (IJSRCSEIT), ISSN : 2456-
Business Media, 2012. 3307, Volume 5 Issue 1, pp. 163-170, January-
[4] Zhu, Xiaoming, and Joseph M. Kahn, February 2019. Available at doi :
"Queueing models of optical delay lines in https://doi.org/10.32628/CSEIT195128
synchronous and asynchronous optical packet- Journal URL : http://ijsrcseit.com/CSEIT195128
switching networks," Optical Engineering, vol.
42, no. 6, pp. 1741-1748, 2003.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


170
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195118
Improving Central and Remote Access Units in Cloud Based Wi-
Fi Network for Better Client and Service
Kavitha-R1, Sathiyabama-G2
1Department of Computer Science and Engineering, Velalar College of Engineering &Technology, Erode,
Tamilnadu, India
2 ME, Department of Computer Science and Engineering, Velalar College of Engineering &Technology, Erode,
Tamilnadu, India
ABSTRACT

People show attention to cloud computing since it is efficient and scalable. But maintaining the stability of
processing so many jobs in the cloud computing environment with load balancing is a very difficult problem
and receives much attention by researchers. Load balancing in cloud computing platform has performance
impact also. Good load balancing makes cloud computing more efficient and increases user satisfaction. This
paper introduces a better load balance model for public cloud based on cloud partitioning concept with a switch
mechanism to choose various strategies in different situations. The algorithm applies game theory to load
balancing strategy and improves the efficiency in the public cloud environment. MCC is a concept which aims
to mitigate these limitations by extending the capabilities of smart devices by employing cloud services, as
required. In MCC both the data storage and processing occur external to the mobile device, while in cloud
computing it is usually only the data storage which is external. In this following a some of these architectures as
categorized in augmented execution, elastic partitioned/modularized applications, application mobility, ad-hoc
mobile cloud and add a fifth category; cyber foraging. This system addresses whether MCC techniques can be
used to extend the capabilities of resource-constrained mobile-devices to provide the illusion of infinite, elastic
resources on demand. The existing system limitations of mobile-devices, and identified five key limited
resources as being CPU, memory, battery, data usage and time. In this research explored existing solutions for
these limitations and identified offloading computation and storage from the device as a possible solution.
Keywords : Mobility, Load Balancing, MCC, Cyber Foraging

I. INTRODUCTION The most sanctioning technology for cloud


computing is virtualization. Virtualization computer
Cloud computing is internet-based computing during code permits a physical computer to be electronically
which massive teams of remote servers area unit separated into one or additional "virtual" devices,
networked to permit sharing of data-processing tasks, every of which may be simply used and managed to
centralized information storage, and on-line access to perform computing tasks. Cloud computing adopts
pc services or resources. Clouds are classified as ideas from Service directed design (SOA) that may
public, non-public or hybrid. Cloud computing may facilitate the user break these issues into services that
be a form of computing that depends on sharing may be integrated to supply an answer. Cloud
computing resources instead of having native servers computing provides all of its resources as services,
or personal devices to handle applications. and makes use of the well-established standards and

CSEIT195118 | Received : 25 Jan 2019 | Accepted : 05 Feb 2019 | January-February -2019 [ 5 (1) : 171-179 ]
171
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

best practices gained within the domain of SOA to 3. Microsoft Azure


permit world and simple access to cloud services in 4. There are many small and medium scale cloud
an exceedingly standardized method. operators that include IBM, Oracle etc.

The internet of Things is setting out to rework daily


tasks area unit completed. the net of Things (IoT)
consists of everyday objects – physical devices,
vehicles, buildings etc. with embedded natural
philosophy, software, sensors, and network property,
permitting them to gather, send and receive
knowledge. The IoT generates a massive quantity of
huge knowledge and this successively puts a large
strain on web Infrastructure. As a result, this forces
corporations to search out solutions to minimize the
pressure and solve their drawback of transferring
massive amounts of knowledge. Cloud computing has
entered the thought of knowledge technology,
Cloud computing may be a quite grid computing; it's
providing quantifiability in delivery of enterprise
evolved by addressing the QoS (quality of service)
applications and software system as a Service (SaaS).
and responsibleness issues. Cloud computing provides
Corporation area unit currently migrating their data
the tools and technologies to make data/compute
operations to the cloud. Several cloud suppliers will
intensive parallel applications with rather more
yield your knowledge to be either transferred via
reasonable costs compared to ancient parallel
your ancient web association or via a passionate
computing techniques.
direct link. The advantage of an on the spot link into
the cloud can make sure that your knowledge is
Characteristics
uncontended which the traffic isn't crossing the net
Cloud computing exhibits the following key
and also the Quality of Service are often controlled.
characteristics
• Agility
II. RELATED WORKS
• Application programming interface
• Cost reductions
1. Niroshinie Fernando, Seng W. Loke, Wenny
• Device and location independence
Rahayu
• Maintenance
• Multitenancy
Despite increasing usage of mobile computing,
• Performance
exploiting its full potential is troublesome thanks to
• Productivity
its inherent issues like resource scarceness, frequent
• Reliability and Scalability
disconnections, and quality. Mobile cloud computing
• Security
will address these issues by corporal punishment
mobile applications on resource suppliers external to
Cloud computing is dominated by
the mobile device. during this paper [1], to produce
1. Amazon Web Services (AWS),
an intensive survey of mobile cloud computing
2. Google Compute Engine, Google App Engine
analysis, whereas lightness the precise considerations

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


172
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

in mobile cloud computing. this taxonomy supported facilitate building a lot of powerful mobile
the key problems during this space, and discusses the applications.
various approaches taken to tackle these problems. To
conclude the paper with a assessment and challenges 3. HanQi , Abdullah Gani
that haven't nevertheless been absolutely met, and Mobile Cloud Computing (MCC) which mixes mobile
highlight directions for future work. computing and cloud computing, has become one
The increasing usage of mobile computing is clear by amongst the trade buzz words and a serious
the study by Juniper analysis, that states that the discussion thread within the IT world since 2009. As
buyer and enterprise marketplace for cloud-based MCC continues to be at the first stage of development,
mobile applications is anticipated to rise to $9.5 it's necessary to understand a radical understanding
billion. In recent years, applications targeted at of the technology so as to entails the direction of
mobile devices have started changing into plentiful future analysis. With the latter aim, this paper
with applications in numerous classes like presents a review on the background and principle of
amusement, health, games, business, social MCC, characteristics, recent analysis work, and
networking, travel and news. The recognition of future analysis trends. a quick account on the
those ar evident by browsing through mobile app background of MCC: from mobile computing to
transfer centers like Apple’s iTunes or Nokia’s Ovi cloud computing is given and so followed with a
suite. discussion on characteristics and up to date analysis
work. It then analyses the options and infrastructure
2. Dejan Kovachev, Yiwei Cao and Ralf Klamma of mobile cloud computing.
Cloud computing is an rising construct combining
several fields of computing. the inspiration of cloud 4.Preeti Garg, Dr. Vineet Sharma
computing is that the delivery of services, computer Describes a cloud computing extremely climbable
code and process capability over the net, reducing computing resources are equipped as associate outer
price, increasing storage, automating systems, service through net on pay-as-usability basis. Portio
decoupling of service delivery from underlying analysis estimates that mobile subscribers can reach
technology, and providing flexibility and quality of half dozen.5 billion by the tip of 2012, 6.9 billion by
knowledge. However, the particular realization of the tip of 201 three. thanks to increasing use of
those edges is much from being achieved for mobile mobile devices the necessity of cloud computing in
applications and open several new analysis queries. In mobile devices arise, that gave birth to Mobile Cloud
order to higher perceive the way to facilitate the Computing (MCC). Mobile Cloud Computing refers
building of mobile cloud-based applications, they to associate infrastructure wherever processing and
need surveyed existing add mobile computing storage will happen faraway from mobile device.
through the prism of cloud computing principles. The Mobile devices don't have to have massive storage
given definition of mobile cloud computing and capability and powerful electronic equipment speed.
supply an outline of the results from this review, Thanks to storing information on cloud there's
specially, models of mobile cloud applications. To associate issue of information security. as a result of
focus on analysis challenges within the space of the chance related to information storage several IT
mobile cloud computing. To conclude with professionals aren't showing their interest towards
recommendations for a way this higher Mobile Cloud Computing. This paper[10] explores: (i)
understanding of mobile cloud computing will The thought of Mobile Cloud Computing and
problems associated in it (ii) Security of information

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


173
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

hold on in cloud with varied mechanisms (iii) resource utilization while maintaining affordable
projected a attainable answer to supply quality of service. However, such a goal cannot be
confidentiality, access management similarly as achieved while not economical resource management.
integrity of information. The most physical resources, like processor cores and
I/O devices, are shared among virtual machines
5.Byung-Gon Chun, Sunghwan Ihm, Petros Maniatis exploitation time slicing and are often scheduled
Describe a mobile applications are becoming flexibly supported priority, allocating an acceptable
increasingly ubiquitous and provide ever richer quantity of main memory to virtual machines is
functionality on mobile devices. Meanwhile, such tougher. Totally different applications have different
devices often enjoy strong connectivity with more memory needs. Even one application shows varied
powerful machines ranging from laptops and operating set sizes throughout its execution. A best
desktops to commercial clouds. This paper presents memory management strategy below a virtualized
the design and implementation of Clone Cloud, a environment therefore has to dynamically alter
system that automatically transforms mobile memory allocation for every virtual machine that
applications to benefit from the cloud. The system is more needs a prediction model that forecasts its host
a flexible application partitioned and execution physical memory wants on the fly. This paper
runtime that enables unmodified mobile applications introduces Memory Balancer (MEB) that dynamically
running in an application-level virtual machine to monitors the memory usage of every virtual machine,
seamlessly off-load part of their execution from accurately predicts its memory wants, and
mobile devices into device clones operating in a sporadically reallocates host memory. MEB uses 2
computational cloud. Clone Cloud uses a effective memory predictors that, severally, estimate
combination of static analysis and dynamic profiling the number of memory accessible for reclaiming
to partition applications automatically at a fine while not a notable performance drop, and extra
granularity while optimizing execution time and memory needed for reducing the virtual machine
energy use for a target computation and paging penalty. The advantage of this study is that
communication environment. During runtime, the swap usage from the OS performance statistics
application partitioning is achieved by migrating a and posts the swap usage statistics to the central
thread from the mobile device at a chosen point to information storage. The limitation of this study is
the clone in the cloud, executing there for the residue reducing overhead while up prediction accuracy.
of the partition, and re-integrating the migrated
thread back to mobile. Their evaluation shows that 7.Weizhe Zhang et al
Clone Cloud can adapt application partitioning to Cloud computing emerges as a brand new computing
different environments, and can help some paradigm involved by each domain and trade.
applications achieve as much as a 20x execution Resource management of multiple virtual machines is
speed- up and a 20-fold reduce of energy spent on that the core of Infrastructure as a Service.
mobile. Specializing in the mainframe resources, the aim of
this paper is to extend the QoS of net service by
6.Weiming Zhao properly planning the mainframe resource across the
Virtualization primarily enables multiple operating virtual machines. The authors of this study formulate
systems and applications to run on one physical pc by the mainframe planning of multiple virtual machines
multiplexing hardware resources. A key motivation into an whole number programming drawback. Then,
for applying virtualization is to enhance hardware a worldwide regulation algorithmic rule supported

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


174
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

utility optimization theory is planned. It principally The administrator selects a minimum and a most
consists of 4 parts: the monitor, the regulator, the information measure limit. The primary pre-copy
native effectiveness of management and therefore the spherical transfers pages at the minimum information
international effectiveness of management. The measure. Every resultant spherical counts the
regulator is within the Xen virtual machines’ monitor. quantity of pages dirtied within the previous
The monitor, native effectiveness of management and spherical, and divides this by the length of the
therefore the net server are settled in DomainU’s previous spherical to calculate the soiling rate. The
interior. The international effectiveness of information measure limit for consecutive spherical is
management is created of the worldwide utility then determined by adding a relentless increment to
generating module and global optimizing and finding the previous round’s soiling rate have by trial and
module, realized in the Domain0. Monitor: the most error determined that 50Mbit/sec may be a
role of monitor is to discover an internet server QoS appropriate worth. It terminates pre-copying once
metrics, in the text it means that the typical latency the calculated rate is bigger than the administrator’s
of net. In their experimental environment Apache chosen most, or once but 256KB remains to be
module was developed for the Apache server to transferred. By integration live OS migration into the
record period of time server latency. Additionally, we Xen virtual machine monitor we tend to alter speedy
have a tendency to monitor the changes of the system movement of interactive workloads at intervals
resources. It’s principally to try to do some tests for clusters and knowledge centers. The dynamic
proc CPUinfo, and procure the condition of the network-bandwidth adaptation permits migration to
system mainframe etc. proceed with stripped-down impact on running
services, whereas reducing total time period to below
8. Christopher Clark et al discernible thresholds. Feature of this study is
Migrating software package instances across distinct moving the contents of a VM’s memory from one
physical hosts may be a useful gizmo for directors of physical host to a different is approached in any
knowledge centers and clusters: It permits a clean range of how. And also the limitation may be a live
separation between hardware and computer code, service it's vital that this transfer happens in a very
and facilitates fault management, load equalization, manner that balances the necessities of minimizing.
and low-level system maintenance. By finishing up
the bulk of migration whereas OSes still run, atuhors 9. Albert M.K. Cheng et al
reach spectacular performance with stripped-down Describe a Virtualization technology recently
service downtimes; the study demonstrate the becomes a hot analysis topic once more in each
migration of entire OS instances on a artifact cluster, business and teachers. Some physical hardware like
recording service downtimes as low as 60ms. And processors and I/O devices area unit shared among
also the author shows that that our performance is virtual machines victimization time slicing, however
adequate to create live migration a sensible tool even the memory resource management is comparatively
for servers running interactive hundreds. During this sophisticated. Light-weight memory management
paper, the look choices for migrating OSes running design for multiple virtual machines is projected, and
services with aliveness constraints, specializing in it includes a mix of self adjustment and global-
knowledge center and cluster environments is taken adjustment policies, each of that collaborate with one
into account. another to boost the memory potency. IC module: it
consists of 2 parts: Virtual Machine Collector (VMC)
and Physical Machine Collector (PMC). VMC is

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


175
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

found in DomainU, accountable for aggregation every PROPOSED WORK


domain’s memory usage and disc space usage statistics.
PMC is put in at Domain0 to gather the scale of total MCC is a concept which aims to mitigate these
and idle memory within the physical machine, etc. limitations by extending the capabilities of smart
Regulator module: it's accountable to control devices by employing cloud services, as required. In
resources. The authors used Balloon-Driver as a MCC both the data storage and processing occur
restrictive mechanism of memory to trigger the external to the mobile device, while in cloud
memory rescaling and XenStore to exchange computing it is usually only the data storage which is
knowledge between varied domains. external

PM module: it's the core of the system. If the These Paper describes the choices to synchronize the
resource is decent, memory sizes area unit adjusted offline folder contents to the online. Word
by the VMs themselves that is named self-adjustment. documents, surpass worksheets and alternative files is
Otherwise, the global-adjustment is activated for the designated and uploaded to the online application.
complete memory management consistently. Additionally, they'll be reborn into bytes and saved
Following area unit the benefits of this study: Self- within the info provided within the net server
adjustment is running as a daemon method once the location. In anyplace, the administrator will read the
OS of a particular domain starts. contents within the server choose a file and transfer
to the system where the administrator is functioning.
10. Tudor-Ioan Salomie, Gustavo dancer, Timothy Likewise, the files are uploaded from the online page
Roscoe additionally. The windows application is wont to
Describes a program like databases and language transfer the content from the online website to the
runtimes usually manage memory themselves to use native machine. So the applying synchronizes the
application information out of stock to the OS. online content and windows content. The applying
Historically deployed on dedicated machines, they're acts like content storage also as document version
designed to be statically designed with memory manager.
comfortable for peak load. In virtualization
eventualities (cloud computing, server consolidation), Likewise, all the documents square measure keep in
however, static peak provisioning of RAM to info additionally with the name, the system from that
applications dramatically reduces the potency and it's uploaded, the information and time of uploading
cost-saving edges of virtualization. also because the original file copy is additionally
unbroken within the server. The files uploaded from
III. METHODOLOGY the system in foreign places apart from the native
workplace if presents within the information
EXISTING WORK processing system, then the administrator are
intimated specified some files gift within the
The existing system introduces a load balance model information processing system isn't downloaded to
for the public cloud based on the cloud partitioning the native system. This helps the administrator to
concept with a switch mechanism to choose different synchronize the document expeditiously. The system
strategies for different situations, then after creating is implemented with the following methods:
all partitioning load balancing will starts to the all
node, these are main drawback.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


176
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

o Augmented Execution The applications area unit partitioned off into


o Elastic/Partitioned Modularized Applications elements such these may be remotely dead severally
o Ad-hoc Mobile Cloud on a cloud. With static partitioning, applications area
o Application Mobility unit partitioned off into mounted elements
o Cyber Foraging throughout compile time or runtime. In distinction,
Dynamic partitioning is that the partitioning of
Resource-intensive tasks can be relegated to the applications at runtime, sporadically (on interval
cloud or other resources such that the mobile device basis) or nonchalantly. Dynamic configuration
constraints are mitigated. In MCC both the data supported factors like CPU standing, memory, power,
storage and processing occur external to the mobile bandwidth and user preferences determines
device, while in cloud computing it is usually only wherever an application is run; whether or not on
the data storage which is external. the device or within the cloud server.

3.1 AUGUMENTED EXECUTION 3.3 AD-HOC MOBILE CLOUD

Processes on mobile devices area unit offloaded to the In Ad-hoc Mobile Cloud, the given task by a mobile
cloud so as to beat good device limitations, like node, the method execution is happened within the
process power, battery life and memory. During this shopper if the server is unavailable. Ad-hoc mobile
module thought-about the restricted Memory-Client/ cloud may be a set of mobile devices that replace the
Excessive Memory-Server. the first practicality cloud infrastructure by providing their computation
outsourcing (intensive computation is offloaded), resources to different devices. Mobile devices have
background augmentation (processes that don't need storage, procedure power, battery power and sensing
user interaction area unit offloaded), inject capabilities, that provides a chance for exploiting the
augmentation (changes within the execution path collective power of those devices. This is often
may be offloaded), hardware augmentation particularly helpful in things wherever there's very
(augmentation of underlying platform) Association little or no affiliation to the network, restricted
augmentation through multiplicity (running multiple power (short vary communications consume less
parallel copies of an application to create optimum energy).
decisions).
3.4 APPLICATION MOBILITY
The computationally overpriced tasks area unit full to
an online service containing image of the device, Application quality relies on the idea of method
whereas straightforward tasks area unit unbroken on migration, whereby processes may be paused,
the device. Once the net service in cloud task transferred to a distinct machine and seamlessly
completes the execution, results area unit integrated resumed. The distinction between this and increased
back to the mobile node. Execution is that in application quality, migration
will occur between completely different underlying
3.2 ELASTIC/PARTIONED MODULARIZED mobile architectures. The applying quality, one
APPLICATIONS method is dead in client/ different method is
processed within the server that is Cloud choice
Some of the processes are dealt in shopper and based mostly. it's engineered on a distributed filing
remaining process area unit are dealt from the server. system, that support the transmission of state from

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


177
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

one web site to a different, and also the modification to figure effectively in each non-public and public
in state from suspend to resume or contrariwise. The cloud.
purchaser’s area unit that internets suspend / resume
have association asynchronous dependence on the Furthermore, to gauge a best help in document
network, whereas purchasers would historically be management, the application become helpful if the
synchronously captivated with the network. below enhancements square measure created in
future. If the appliance is meant as internet service, it
3.5 CYBER FORAGING may be integrated in several internet sites. Non-
synchronized documents may be alerted to
The process or the given task of a node is partly dead synchronize. the appliance is developed specified on
in an un-trusted server. It's a way whereby proximate top of aforementioned enhancements may be
offered resources (called surrogates) area unit used integrated with current method.
for remote execution of applications. These resources
area unit dynamically discovered and used V. REFERENCES
notwithstanding the servers don't seem to be
expressly trustworthy by the user. Surrogates area [1]. N. Fernando, S. W. Loke, and W. Rahayu,
unit supported 2 main premises; un-trusted and "Mobile cloud computing: A survey," Future
unmanaged. This reduces the overall price of Generation Computer Systems, vol. 29, no. 1,
possession and maintenance and so encourages their pp. 84-106, 2015.
broad adoption. The offload tasks to a distant server [2]. D. Kovachev, Y. Cao, and R. Klamma, "Mobile
area unit supported many factors together with cloud computing: a comparison of application
computation quality, network strength, and device models," arXiv preprint arXiv:1107.4940, 2012.
capability and information neighborhood. [3]. [H. Qi and A. Gani, "Research on mobile cloud
computing: Review, trend and perspectives," in
IV. CONCLUSION Digital Information and Communication
Technology and it’s Applications (DICTAP),
Through this analysis, the information management 2012 Second International Conference on .
method becomes straightforward. All the regular IEEE, 2012, pp. 195-202
activities square measure assigned to them through [4]. Preeti Garg, Dr. Vineet Sharma, "Secure Data
browser interface. The administrator will read the Storage In Mobile Cloud Computing ," in
contents within the server choose a file and transfer Computer Communications Workshops
to the system where the administrator is functioning. (INFOCOM WKSHPS), 2011 IEEE Conference
Likewise, the files may be uploaded from the online on . IEEE, 2014, pp. 1060-1065.
page additionally terribly straightforward manner. [5]. Byung-Gon Chun , Sunghwan Ihm, Petros
The new system eliminates the difficulties within the Maniatis, "Securing elastic applications on
existing system. It’s developed in a very easy manner. mobile devices for cloud computing," in
The system is incredibly quick and any dealings may Proceedings of the 2009 ACM workshop on
be viewed or retaken at any level. Error messages Cloud computing security . ACM, 2016, pp.
square measure given at every level of input of 127-134
individual stages. Several of the tested practices and [6]. Weiming Zhao, "Dynamic Memory Balancing
technologies for managing trust relationships in for Virtual Machines ", 2016 14th International
ancient enterprise IT environments may be extended Conference on Modelling and Simulation

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


178
R. Kavitha, G. Sathiyabama Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 171-179

[7]. Weizhe Zhang C. Wang, K.Q. Yan, W.P. Liao


and S.S. Wang, Multiple Virtual Machines
Resource Scheduling , Proceedings of the 3rd
IEEE International Conference on Computer
Science and Information Technology, pp. 108-
113, 2015.
[8]. Christopher Clark , H. Jamal, A. Nasir, K.
Ruhana, K. Mahamud and A.M. Din, Live
Migration of Virtual Machines, Proceedings of
the Second International Conference on
Computational Intelligence, Modelling and
Simulation, pp. 160-165, 2014.
[9]. Albert M.K. Cheng, R. Subrata, and A.Y.
Zomaya, "LVMM: A lightweight virtual
machine memory management architecture for
virtual computing environment," Proc. 25th
IEEE Int’l Performance Computing and Comm.
Conf. (IPCCC ’06), 2016.
[10]. Tudor-Ioan Salomie, Gustavo Alonso, Timothy
Roscoe " Game-Theoretic Approach F Or Load
Balancing In Computational Grids ",IEEE
transactions on parallel and distributed systems,
vol. 19, no. 2, February y 2018.

Cite this article as :

R. Kavitha, G.Sathiyabama, "Improving Central and


Remote Access Units in Cloud Based Wi-Fi Network
for Better Client and Service ", International Journal of
Scientific Research in Computer Science, Engineering
and Information Technology (IJSRCSEIT), ISSN :
2456-3307, Volume 5 Issue 1, pp. 171-179, January-
February 2019. Available at doi :
https://doi.org/10.32628/CSEIT195118
Journal URL : http://ijsrcseit.com/CSEIT195118

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


179
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195122
Effective Prediction for Rock Burst Dataset Using Classification
Algorithms with Particle Swarm
Sadesh-S1, Banupriya-GK2
1Department of Computer Science and Engineering, Velalar College of Engineering &Technology, Erode,
Tamilnadu, India
2 ME, Department of Computer Science and Engineering, Velalar College of Engineering &Technology, Erode,
Tamilnadu, India

ABSTRACT

Rock burst and Slope Stability is one of the common failures in hard rock mining and civil construction. This
study focuses on the prediction of rock burst and Slope Stability classification with case instances using cloud
models and attribution weight. First, cloud models are introduced briefly related to the rock burst and Slope
Stability classification problem. Then, the attribution weight method is presented to quantify the contribution
of each rock burst and Slope Stability indicator for classification. In addition, analysis and prediction of slope
stability is of great importance in geotechnical engineering. With the development of economics, the number
of slopes is increasing and landslides caused by slope instability have become one of the three major geological
disasters in the world along with earthquakes and volcanoes. To reduce or prevent landslide damage, slope
stability analysis and stabilization are required. However, accurately predicting slope stability is challenging
because of the complexity of slope structures and the difficulty to determine the precise input data associated
with key geotechnical parameters the proposed methodology PSO feature extraction preserves important
distance relationships, such as : The Random forest, Naive Bayes of each object of the original dataset. This
leads to preservation of any mining operation that depends on the ordering of distances between objects, such
as Random forest, Naive Bays -search, SVM, J.48 and MLP classification, as well as many visualization
techniques. In particular, it establishes a restricted isometric property, i.e., tight bounds on the
contraction/expansion of the original distances.
Keywords: Raspberry Pi, USB Camera, Motor Driver, Motor, LCD Display, Road Side Speed Sign Controlling.

I. INTRODUCTION correlation or patterns among dozens of fields in the


large relational database. [14]
Data mining is analyzing data from different
perspective and summarizing it into useful ISSUSES IN DATA MINING
information. The information can be used to increase
revenue, cut cost, or both. Data mining allows the Mining different kinds of knowledge in databases
user to analyze data from many different dimensions Data mining is used to cover a wide spectrum of data
or angles, categorize it, and summarize the analysis and knowledge discovery task. Different user
relationships identified. It is the process of finding can use different kinds of knowledge.

CSEIT195122 | Received : 25 Jan 2019 | Accepted : 05 Feb 2019 | January-February -2019 [ 5 (1) : 180-184 ]
180
Sadesh S, Banupriya GK Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 180-184

Interactive mining at multiple levels of abstraction XIAO FAN and KAN WANG [2] In the field of
It is difficult to know what can be discovered within nuclear engineering, deterministic and stochastic
database. The Data mining process should be methods are used to solve radiation transport
interactive. Database containing a huge amount of problems. To solve the transport equation for the
data, appropriate sampling techniques can be used to average particle behavior and also contain
facilitate interactive data exploration. [9] uncertainties associated with the discretization of the
independent variables such as space, energy and
Pattern evaluation angle of the transport equation and can admit
It can uncover thousands of patterns. Many of the solutions that exhibit non - physics features. The
patterns discovered may be uninteresting to the given Monte Carlo method obtains s results by simulating
user they present common knowledge or lacking individual particles and recording some aspects of
novelty. their average behavior. This method enables detailed,
explicit geometrical, energy and angular
Handling noisy or incomplete data representations and hence is considered the most
The data stored in a database may reflect noise such accurate method presently available for solving
as exceptional cases or incomplete data objects. The complex radiation transport problems. One of the
accuracy can be poor. Data cleaning and data analysis difficulties associated with Monte Carlo method is
method is used to handle the noisy data. the amount of computer time required to obtain
sufficient precision in the simulations.
II. LITERATURE SURVEY
QIANG YANG AND PING- AN DU [3] To present a
CHUN'AN TANG AND JOHN A. HUDSON [1] he novel approach for determining the weights of
subject of rock failure has been studied in a decision makers (DMs) based on rough group
coordinated way since the 1960s.It can be studied by decision in multiple attribute group decision-making
examination of natural rock formations that have (MAGDM) problems. At first, we construct a rough
been stressed and strained over geological time, by group decision matrix from all DMs’ decision
laboratory experiments on the rock samples in situ matrixes on the basis of rough set theory. The second
experiments takes place, and by observing the results process we derive a positive ideal solution (PIS)
of rock excavation and loading during engineering founded on the average matrix of rough group
construction. In this book, rock failure mechanisms decision, and negative ideal solutions (NISs) founded
explanations and illustrations are given. Over the on the lower and upper limit matrixes of rough group
years, there have been three main developmental decision. After that we obtain the weight of each
phases supporting rock engineering design: analysis group member and priority order of alternatives by
based on elasticity theory; the use of rock mass using relative closeness method, which depends on
classification systems; and computer modeling. The the distances from each individual group member’
elasticity theory approach is more useful because it decision to the PIS and NISs. Comparisons with
enables the stresses around circular and elliptical existing methods and an on-line business manager
holes to be determined, although the approach is with the selection example, the proposed method
most useful for deep excavations where the rock show that it can provide more insights into the
behavior is essentially elastic. subjectivity and vagueness of DMs’ evaluations and
selections.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


181
Sadesh S, Banupriya GK Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 180-184

MATTHIJS J. WARRENS [4] The kappa coefficient, ZONG- XIAN ZHANG [7] are two missions for rock
denoted by κ, is widely used as a descriptive statistic mechanics to accomplish in mining engineering: (1)
for summarizing the cross-classification of two to destroy rock efficiently; (2) to make rock
variables with the same unordered categories. structures safe. The missions are completed and
Originally proposed as a measure of agreement mining operations are well managed, best mining
between two raters classifying subjects into mutually results should be achieved. For accomplishments of
exclusive categories, Cohen’s κ has been applied to the two missions, rock mechanics faces following
square cross-classifications encountered in challenges: (1) how to make drilling, crushing and
psychometrics, educational measurement, grinding more efficiently, in particular for grinding
epidemiology, diagnostic imaging, map comparison, whose energy efficiency is less than 1%; (2)now how
and content analysis. The popularity of Co- hen’s κ to make full use of explosive energy and destroy rock
has led to the development of many extensions, effectively; (3) how to manage, reduce and finally
including multi-rater kappas, kappas for groups of predict seismic events and rock bursts; (4) how to the
raters, and weighted kappas. The value of κ is 1 when develop various mining methods; (5) how to reduce
perfect agreement between the two observers occurs, borehole damage in deep mines or in the mines with
0 when agreement is equal to that expected under high in - situ stresses; (6) how to increase ore
independence, and negative when agreement is less recovery and decrease dilution; (7) how to improve
than expected by chance. mining safety; (8) how to make rock support de signs
more scientifically. All these challenges will be
YU ZHOU AND TINGLING WANG [5] Rock burst is analyzed in this paper. In addition, few topics such as
one of main engineering geological problems greatly rock mass classification, environment protection and
threatening the safety of construction. Prediction of the effects of loading rates, temperatures, and cyclic
rock burst is always to important for the safety of loading on mining engineering will be discussed.
workers and equipments in tunnels. Here in this
paper, PNN-based rock burst prediction model is III. METHODOLOGY
proposed to determine whether rock burst will
happen in the underground rock projects and how NORMALIZATION PROCESS
much the intensity of rock burst.
It is the process of classify data into an associated
ZAOBAO LIU and JIANFU SHAO [6] Rock burst is table it also eliminates redundancy and increases the
one of the common failures in hard rock mining and reliability which improves output of the query. For
civil construction. It focuses on the prediction of rock normalizing a database, we divide the dataset into
burst classification with the case instances using tables and establish relationships between the
cloud models and attribution weight. At first, the tables.The dataset normalization can essentially be
cloud models are introduced briefly related to the defined as the practice of optimizing table structures.
rock burst classification problem. Now, the Optimization is accomplished with the result of a
attribution weight method is presented to quantify thorough investigation of the various pieces of data
the contribution of each rock burst indicator for that will be stored within the database, in particular
classification. This approach is implemented to concentrating upon how this data is interrelated
predict the classes of rock burst intensity for the 164
rock burst instances collected.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


182
Sadesh S, Banupriya GK Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 180-184

Min Max Normalization


SYSTEM FLOW DIAGRAM
Min max normalization is a normalization strategy
which linearly transforms x to y= (x-min)/(max-min),
where min and max are the minimum and maximum
values in X, where X is the set of observed values of x.
It can be easily visible when x=min, then y=0.

y=x-(min(x)) / (max(x)-min(x))

FEATURE SELECTION

HCR-PSO feature extraction model for Rock Burst


dataset and applied an improve probability in many
Geo-graphical application such as training artificial
neural networks, linear constrained function
optimization, wireless network optimization, data
classification, and in many other areas where GA is
applied. Computation in HCR-PSO is based on a
swarm of processing elements called particles in
which each particle represent a candidate solution.
The system is initialized with a Rock Burst dataset
swarm of random solutions and searches for optima
by updating Rock Burst dataset generations. The
search process utilizes the combination of
deterministic and also tbe probabilistic rules that
depend on information sharing among their
population members to enhance their search
IV.CONCLUSION
processes. Rock Burst dataset prediction system
sharing mechanism in HCR-PSO is considerably
This project analyzed the Rock Burst dataset using
different.
algorithms such as J48, MLP, SVM, Random Forest,
and Bayesian Classification. These algorithm gives
In GAs, chromosomes share information with each
various result based on HCR-PSO feature extraction
other, so the whole Rock Burst dataset moves like
model. It has been seen that MLP and J48
one group towards a selected area. In HCR-PSO, the
Classification gives better results compare to other
global best swam particle found among the swarm is
classification algorithms. There are many criterions
the only Rock Burst dataset shared among particles. It
for evaluating the selected feature subset; here this
is a one - way Rock Burst dataset prediction sharing
project used features to evaluate the performance of
mechanism. The Rock Burst dataset prediction
different classification algorithm. In future, we have
computation time in HCR-PSO is much less than in
attempted to classify different feature selection
GAs because all swam particles in HCR-PSO tend to
algorithms into four groups: complete search,
meet to the best solution fast.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


183
Sadesh S, Banupriya GK Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 180-184

heuristic search, meta-heuristic methods and [8]. Tan P.-N., Steinbach M. and Kumar V.,
methods that use artificial neural network. The "Introduction to Data Mining", Addison
future methodology is used to analyze the Rock Burst Wesley, 2006
dataset region into separable compartments i.e. Rock [9]. Weber.M, Welling.M, and Perona.P.
Burst dataset etc. However, the method requires "Unsupervised Learning of Models for
further improvement mostly regarding feature Recognition
selection of the Rock Burst dataset into multiple
components: renal cortex, renal column, renal Cite this article as :
medulla and renal pelvis. Apart from that, this is
planned to expand the database on which the system Sadesh S, Banupriya GK, "Effective Prediction for
will be tested. And also, the proposed method in this Rock Burst Dataset Using Classification Algorithms
project can be employed for detecting the Rock Burst with Particle Swarm", International Journal of
dataset Geo-graphical in future with the Rock Burst Scientific Research in Computer Science, Engineering
dataset and classification of the Geo-graphical. and Information Technology (IJSRCSEIT), ISSN :
2456-3307, Volume 5 Issue 1, pp. 180-184, January-
V. REFERENCES February 2019. Available at doi :
https://doi.org/10.32628/CSEIT195122
[1]. Akay M.F. "Support vector machines combined Journal URL : http://ijsrcseit.com/CSEIT195122
with feature selection for breast cancer
diagnosis" (2009) Expert Systems with
Applications, 36 (2 PART 2) , pp. 3240-3247.
[2]. Cristianini N and Shawe-Taylor J., "An
Introduction to Support Vector Machines and
Other Kernel-based Learning Methods",
Cambridge University Press, 2000.
[3]. Fergus.R, Fei-Fei.L, Perona.P, and Zisserman.A.
"Learning Object Categories from Google’s
Image Search".
[4]. Han, Jiawei; Kamber, Micheline (2001). Data
mining: concepts and techniques. Morgan
Kaufmann. p. 5. ISBN 978-1-55860-489-6.
[5]. Han, Kamber, Pei, Jaiwei, Micheline, Jian (June
9, 2011). Data Mining: Concepts and
Techniques (3rd ed.). Morgan Kaufmann. ISBN
978-0-12-381479-1.
[6]. Pal S.K. and Mitra P, "Pattern Recognition
Algorithms for Data Mining", CRC Press, 2004.
[7]. R language and environment Hornik, Kurt
(November 26, 2015). "R FAQ". The
Comprehensive R Archive Network. 2.1 What
is R. Retrieved 2015-12-06.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


184
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195140
Enhancement in the Lifetime of Wireless Sensor Network
Bhumika Yadav1, Ram Neresh Sharma2
1Computer Science Engineering Department, Maharana Pratap College of Technology, Madhya Pradesh, India
Assistant Professor, Computer Science Engineering Department, Maharana Pratap College of Technology,
2

Madhya Pradesh, India


ABSTRACT

Technology changes every day and everyone wants to add more comfort in our life but some technologies
increase the safety. Wireless Sensors are one of the technologies which increase the safety in our life in some
manner. Wireless sensors are used in remote areas where human cannot survive to collect results of atmosphere,
temperature and pressure. All the technical devices are work on electricity/battery and in remote area it is
impossible to change batteries or provide continues power supply to sensors that is why we want to increase
the lifetime of wireless sensor network in this paper. In this paper we are explaining our approach which is
developed by us to increase the lifetime of wireless sensor network as well as we are comparing its results to
some other technologies which are already present now a days.
Keywords : EES Scheme, LBM, REAR, Hybrid, Network Topology, Routing, Random Event Model

I. INTRODUCTION To validate our proposed schemes the simulation are


running on 3 different network with different size
In this paper I am evaluating the performance of my and also with randomly distributed sensor nodes:
proposed approach and compare the sink lifetime
achieved with existing EES schemes to place sink. 1. 160mX160m square sensing area with 100 sensors
Firstly I will describe my experimental setup to 2. 220mX220m square sensing area with 200 sensors
calculate the performance and later we discuss the 3. 360mX360m square sensing area with 500 sensors
results.
I also consider a static or uniform grid topology in
addition to a random topologies, so that all the
Experimental Setup
sensors are evenly distributed in the network area.

In this experiment I am going to measure how the


network lifetime improves which is under a set of
random events. The Coverage area of each event,
start time, location and duration is randomly
generated. Simulation is programmed with C
language and running on Xcode in Mac OS C Lion. I
have also used a number of different-different size
networks and for each network size I am randomly
generate several sensor distributions.

CSEIT195140 | Received : 25 Jan 2019 | Accepted : 05 Feb 2019 | January-February -2019 [ 5 (1) : 185-190 ]
185
Bhumika Yadav, Ram Neresh Sharma Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 185-190

Below mentioned grid topologies are used in our


simulation

1. 160mX160m square sensing area with 100 sensors


2. 220mX220m square sensing area with 225 sensors
3. 360mX360m square sensing area with 400 sensors

Random Event Setup

For every set of simulation I consider 500 random Random event launch sequence
events. The attributes of events such as start and end
time etc are generated randomly and stored in a file Other Parameters
as an input to simulator. In performance evaluation
comparison in EES with 50 events in the simulation,
my experiment has higher occurrence rate because I
am considering the 500 random events are put into
launce sequence during one simulation. The coverage
are for each event varies from 6m to 30m in the
network and the duration of the event varies from 1
unit time to maximum 3000 unit time in the system.
The actual duration of an event will depend on Algorithm
specific application and the time may vary in
millisecond, or in several minutes or in several hours. LBM can be represented as
• S = the set of all sensors in the network
In each time interval many events can start, many • pfinal = (Xs, Ys) the final position of the sink node
can end or many can continue. To define each time • ptentative = (X’s, Y’s) the tentative position of the
unit of the simulation the simulator uses a timeline, sink node (for finding routes)
event file access, read Startn and Endn of every event • pi = (xi, yi) the position of sensor node 𝑠𝑠𝑖𝑖∈𝑆𝑆.
and place them into corresponding sequence of • bi = Total number of bits transmitted by sensor node
launch for every time interval. When the simulation 𝑠𝑠𝑖𝑖∈𝑆𝑆.
start simulation process and the simulation reach the • wi = Total number of bits generated by sensor node
time interval then it will generate pop event in the 𝑠𝑠𝑖𝑖∈𝑆𝑆 .
sequence and perform start or end activities. • dij = distance from node si to node sj
• cij = cost of edge e(i& j)
• dmax = maximum transmission range of a sensor
node
• rij = 1 if and only if node si transmits its data
directly to node sj
Algorithm 1:-
Step 1: Calculate tentative sink position ptentative =
(X’s, Y’s

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


186
Bhumika Yadav, Ram Neresh Sharma Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 185-190

Based on EES [51], use equations (3) and (4) to • Metric 1: First sensor node depletes energy
compute values for X’s and Y’s • Metric 2: 10% of total sensors depletes energy
Step 2: Calculate paths • Metric 3: Network becomes disconnected, i.e.
Use Dijkstra’s algorithm [52] to compute the path at least one sensor node cannot find a valid
from each sensor 𝑠𝑠𝑖𝑖∈𝑆𝑆 to sink node at (X’s, Y’s) path to route data back to the sink
Step 3: Determine load on each sensor node

Load on node-
Step 4: Calculate final sink position

For the LBM algorithm

Fig 2 – Network Lifetime in network with 100


Finally the new location of the sink is determined in sensors
step 4

Result Analysis and Performance Fig 3 – Network Lifetime in network with 200
sensors
I am comparing 4 different strategies in our
comparisons
1. Location based sink placement (EES)
2. Load based sink placement (LBM)
3. Residual Energy aware routing (REAR)
4. Hybrid approach that combines both LBM
and REAR (Hybird).

For each network size, we generated 5 different Fig 4 – Network Lifetime in network with 500
random topologies, and tested each topology with 5 sensors
sets of random events sequences. So, the results Comparison of lifetimes for grid topologies
reported for the random networks are averages from
the 25 runs. As mentioned earlier, different
researchers have used different metrics to measure
the lifetime of a wireless sensor network. In our
simulations, we have taken the following commonly
used metrics to measure network lifetime:

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


187
Bhumika Yadav, Ram Neresh Sharma Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 185-190

Fig 5 – Network Lifetime in grid with 100 sensors Fig 8 – Network lifetime under different maximum
transmission range in network with 200 sensors

Fig 6 – Network Lifetime in grid with 225 sensors


Fig 9 – Metric 1 under different maximum
transmission range in network with 100 sensors

Fig 6 – Network Lifetime in grid with 400 sensors

Fig 10 – Metric 1 under different maximum


Effect of transmission range
transmission range in network with 200 sensors

Fig 7 – Network lifetime under different maximum


transmission range in network with 100 sensors
Fig 11 – Metric 1 with no maximum transmission
range.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


188
Bhumika Yadav, Ram Neresh Sharma Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 185-190

II. CONCLUSION [5]. Luo.F. Relay node placement and trajectory


computation of mobile data collectors in
In this paper I have created an algorithm in which I wireless sensor networks. University of
have introduce two new methods in network lifetime Windsor, Windsor, Master Thesis 2011
in random event and time-based methods. In these [6]. Akyildiz, I.F., Su, W.,Sankarasubramaniam,
kind of cases rand event happens regularly and end Y.,Cayirci, E. Wireless sensor networks: a
for a specific time slot. In the proposed scheme it is a survey. Computer Networks, (2002), 393-422.
mixture of LBM and REAR Technology which can [7]. W. Heinzelman, A. Chandrakasan and H.
balance the lifetime of wireless sensor network in Balakrishnan.Energy efficient communication
much more efficient manner. In LBM method, it protocol for wireless micro-sensor networks.
takes load of transmission and adjust the sink node HICSS, (2000) 3005- 3014
position when an event occurs and the sink node is [8]. J. Pan, Y. T. Hou, L. Cai, Y. Shi, and S. X. Shen.
very heavily loaded. REAR technology is very helpful Topology Control for Wireless Sensor
in the calculation to find routing path in the network. Networks. International Conference on Mobile
Computing and Networking, (2003) 286-299.
We have used a simulator or we can call it custom [9]. ZhiAngEu; Seah, W.K.G. Impact of
simulator to calculate the performance of the Transmission Power And Routing Algorithms
proposed algorithms and also compare our results to in Designing Robust Wireless Sensor Networks.
EES technology. In the above-mentioned result, all Personal, Indoor and Mobile Radio
the graph shows that our method performs better Communications, (2007) 1-5 56
then EES technology continuously when only sink [10]. Y. Bi, L. Sun, N. Li. 2009. BoSS: a moving
node move from one place to another. We have also strategy for mobile sinks in wireless sensor
found that a particular matric is calculating the networks. International Journal of Sensor
lifetime has very significant effect on performance. Networks, 5(2009), 173-184
[11]. Nazir, B. and Hasbullah, H. Mobile Sink based
III. REFERENCES Routing Protocol (MSRP) for Prolonging
Network Lifetime in Clustered Wireless Sensor
[1]. Romer, K and Mattern, F. The Design Space of Network. Computer Applications and Industrial
Wireless Sensor Networks. IEEE Wireless Electronics, (2010), 624 - 629.
Communications, 11(6), (2004) 54-61 [12]. G. Gupta, M. Younis. Fault-tolerant clustering
[2]. D. Estrin, R. Govindan, J. Heidemann, S. of wireless sensor networks. Proceedings of
Kumar, Next century challenges: scalable IEEEWCNC, (2003), 1579 - 1584
coordination in sensor networks, ACM [13]. G. Gupta, M. Younis. 2003. Performance
MobiCom’99, (Washington, USA, 1999), 263- evaluation of load-balanced clustering of
270. wireless sensor networks. 10th International
[3]. Bari. A. Relay node in Wireless Sensor Conference on Telecommunications, (2003),
Network: a survey. Computer Science, 1577-1583
University of Windsor, Windsor, (2005) [14]. Bari, A and Jaekel, A, Techiques for Exploiting
[4]. Chong.C.Y. andKumar.S.P. Sensor Networks: Mobility in Wireless Sensor Networks.
Evolution, Opportunities and Challenges. Handbook of Research on Mobile Business:
Proceeding of the IEEE, 91(8), (2003) 1247- Technical, Methodological and Social
1256 Perspective, (2008), 445-455

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


189
Bhumika Yadav, Ram Neresh Sharma Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 185-190

[15]. Tian, K., Zhang, B., Huang, K. and Ma, J. Data


Gathering Protocols for Wireless Sensor
Networks with Mobile Sinks. IEEE Global
Telecommunications Conference, (2010), 1 - 6.
[16]. Wang, B., Xie, D., Chen, C., Ma, J. and Cheng,
S. Deploying Multiple Mobile Sinks in Event-
Driven WSNs. Communications, (2008), 2293 -
2297.

Cite this article as :

Bhumika Yadav, Ram Neresh Sharma, "Enhancement


in the Lifetime of Wireless Sensor Network",
International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 185-190, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195140
Journal URL : http://ijsrcseit.com/CSEIT195140

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


190
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195155

A Review Paper on IoT Based Smart Irrigation System


Anjali Dokhande1, Chetna Bomble1, Rakshanda Patil1, Puja Khandekar1, Nayan Dhone1,
Prof. Chandrashekhar Gode2
1BE Students, Department Computer Science and Engineering, Priyadarshini J.L. College of Engineering,
Nagpur, Maharashtra, India
2 Assistant Professor, Department Computer Science and Engineering, Priyadarshini J.L. College of Engineering,
Nagpur, Maharashtra, India

ABSTRACT

Automation of farm activities can transform agricultural domain from being manual and static to intelligent and
dynamic leading to higher production with lesser human supervision. This paper proposes an automated
irrigation system which monitors and maintains the desired soil moisture content via automatic watering.
Microcontroller ATMEGA328P on arduino uno platform is used to implement the control unit. The setup uses
soil moisture sensors which measure the exact moisture level in soil. This value enables the system to use
appropriate quantity of water which avoids over/under irrigation. IOT is used to keep the farmers updated
about the status of sprinklers. Information from the sensors is regularly updated on a App using WiFi modem
through which a farmer can check whether the water sprinklers are ON/OFF at any given time. Also, the
sensor readings are transmitted to a Thing speak channel to generate graphs for analysis.
Keywords : Irrigation System, Soil Moisture Sensor, Temperature Sensor, Humidity Sensor, IOT, Android

I. INTRODUCTION promising solution to farmers where the presence of


farmer in the field is not compulsory. Now a day's
In our country, agriculture is a major source of food internet is widely used. Using internet farmer know
production to the growing demand of the human about the agriculture field irrigation status. Thus,
population. In agriculture, irrigation is an essential mobile applications will be helpful in fulfilling this
process that influences crop production by supplying purpose. It helps farmers to know the status of farm
water to the needed land. Farmers have to visit their field moisture and temperature through a mobile app
land to check how much amount of water is required and it will make the farmer at some remote places to
for their field. This irrigation method takes a lot of decide whether he needs to water the field or not. The
time and effort particularly when a farmer needs to soil moisture sensor is used for sensing moisture level
irrigate multiple agriculture fields distributed in to find out whether the soil is dry or wet. The
different geographical areas. Traditionally farmers will moisture sensor and temperature sensor are interfaced
present in their fields to do irrigation process. But with Arduino microcontroller. Based on that sensor
nowadays farmers need to manage their agricultural values, the controller will activate the DC motor or
activity along with other occupations. Automation in pump by sending the commands from the mobile
irrigation system makes farmer work much easier. phone that will operate the pump through the relay.
Sensor-based automated irrigation system provides a Water is a very precious resource and must be

CSEIT195155 | Received : 25 Jan 2019 | Accepted : 05 Feb 2019 | January-February -2019 [ 5 (1) : 191-196 ]
191
Anjali Dokhande et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 191-196

properly utilized. Agriculture is one of those areas


which consume a lot of water. The aim of the paper is
to develop an intelligent irrigation system controller
which measures the moisture of the soil and
temperature of the atmosphere and helps to take
decision accordingly to turns on or off the water
supply system. It is important to know the amount of
water that has been used in the irrigation process
during agriculture. This paper describes a simple
system by using Arduino microcontroller and Android
software to automate the existing manual irrigation
system.

II. RELATED WORK Fig 1. water holding property of soil

A. Moisture Measurement This stage is called as the field limit. Void spaces are
currently loaded with water and air parcels (Fig. 1(b)).
Soil is comprised of blend of segments including Each harvest has a Critical Soil Moisture Deficit level,
mineral and natural particles, with water and air enabling soil to dry out past this level, water pulling
making up the spaces in the middle. Soil can be from edit can't occur, diminishing the yield. Advance
essentially characterized into following 4 classes: more water expulsion from soil lead little holding of
water by soil particles much firmly on account of
surface strain impact for the yield to remove this is
said to be as withering point as appeared in figure 1(c).
The accessible water limit is the measure of water a
dirt can make accessible to plants, by and large
It is encouraged to break down soil to reason its
characterized as the contrast between measure of
classification. Every classification has distinctive
water put away in a dirt at field limit and the measure
properties henceforth their water holding limit
of water put away in the dirt at the lasting shrinking
changes from one sort to the next. As water invades
point. Plants get the vast majority of water from the
soil, it begins to fill the hole between the void spaces
upper part of the root zone.
in the middle of soil particles (Fig. 1(a)), when every
one of the spaces are totally immersed with the water,
The term viable root zone alludes to about the upper
the state is known as immersion point. This state goes
portion of the root zone profundity, where around 70%
on for brief time. With time overabundance water
of harvests water is taken up.
permeates descending through water profile because
of gravitational power. At same time narrow activity
give inverse power to gravity and give adjusted
condition so descending development of water is
ruined.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


192
Anjali Dokhande et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 191-196

changes relying upon the air temperature. More


sizzling air, for instance, can hold more moisture.

III. LITERATURE REVIEW

For the development of automated irrigation structure,


soil moisture content is increasingly basic parameter
when stood out from others as it has a significant part
in plant improvement segment and availability of
water for irrigation is genuine stress for the farmers
uncommonly the ones who are dependent on rain. In
this manner, water organization has a high need while
arranging an automated irrigation system as saw in
most of the composition. The experts have used
Fig 2. Effective root zone
distinctive systems to evaluate moisture substance
precisely. The electrical conductivity estimation is the
Numbers of techniques are produced for soil moisture
most fundamental, sagacious and control capable
estimation from straightforward feel strategy to most
strategy for all. In any case, it isn't correct and its
progress electronic ones.
results vacillate after some time. In spite of its burdens,
it is comprehensively supported by various
B. Temperature Measurement
investigators. [5] Have used this technique to execute
Temperature monitoring is key in numerous modern
an automated irrigation control structure using spill
situations. It likewise assumes imperative part in plant
irrigation framework. The data was accumulated and
development subsequently monitoring temperature is
arranged by the ARM7 board. To give UI GSM
basic for good farming practice. Numerous standard
technique was used and at the customer end android
methods exist which relies on estimation of physical
based application was created to demonstrate data and
properties of the working material that differs with
with that information, the customer can pick what
temperature. Thermocouple, thermistor, RTD,
move to make. According to the call given by the
pyrometer, Langmuir tests, infrared, and so forth are a
customer, the solenoid valve was attempted to control
portion of the cases.
the irrigation. An abnormal state of flexibility was
obtained with this execution. The remaining burden
C. Humidity Measurement
for the farmer was decreased and moreover increase in
There are three approaches to speak to Humidity. It is
the benefit of the estate was viewed.
the measure of water vapour (water that has
abandoned a fluid to an undetectable gas) noticeable
In [6] acoustic methodology has been used to measure
all around. Outright humidity is the genuine measure
water substance of the earth in transit that
of water vapour in a predetermined volume of air.
development time of sound wave is assorted in dry
Relative humidity is the proportion of moisture
and wet soil. In light of the discernment, the chart
noticeable all around when contrasted with the most
was set up for development time versus moisture
extreme measure of moisture the air can hold, which
content from which moisture can be found. Another

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


193
Anjali Dokhande et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 191-196

methodology for moisture estimation is assessing moisture sensor (VH400) in the perspective of
assortment in dielectric steady [7]. Dielectric electromagnetic estimation and second is soil
unfaltering of water is (~80) high when appeared temperature sensor (DS1822). The recorded data was
differently in relation to dry soil (2-3). In [8] secured locally in the memory chip and was moreover
assortment of the dielectric was assessed as an transmitted to the web. The system is controlled
assortment in the capacitor using capacitor and capable and thoroughly wears down sun controlled
resistor interface. The response is generally immediate essentialness. The promising delayed consequence of
which is required in case of exactness agriculture. S. S. around 90% water saving was viewed. P. Bhosale and
Mathurkar and D. S. Chaudhari [9] showed a model in V. Dixit made atmosphere monitoring structure. They
perspective of moisture, temperature and humidity used a broad assortment of sensors for monitoring
sensor. All of the sensors were balanced for the specifically ecological temperature and humidity
straight response. The standard purpose of the (SHT1x), Soil temperature (LM35), Radiation and
structure is to develop an exact system which can be sunshine, soil moisture (gypsum square in perspective
utilized in real estate and offer preferences to the of the resistive framework), wind speed and course
farmer. (anemometer) and rainfall. In control board, PIC
microcontroller was used. The total data was secured
An epic method to manage design automated in memory card amassing and was also sent to the
irrigation system is the use of Plant water weight remote customer by methods for the GSM module.
examination. It is overcome selection of optical and IR Powerful usage of water was practiced through this
pictures of plant overhang. This particular enrollment structure [12].
speaks to a couple of challenges as no anticipated
essential segment or right organizing can be found IV. PROPOSED METHODOLOGY
from the data pictures. X. Wang et al [10] developed
an Automatic Cross-Correlation course of action Nowadays agricultural field is facing lot of problems
count which uses the information of smart picture due to lack of water resources. In order to help the
structure yet clears out the effect of picture shading farmers to overcome the difficulties, smart irrigation
and power in the relationship method and thusly system has been used. In this system, various sensors
achieves a pleasing enlistment result. They similarly such as pH, soil moisture and DHT11 are connected to
realized capable computation approach which can the input pins of Arduino microcontroller. The sensed
fundamentally diminish the estimation eccentrics of values from the sensors are displayed in LCD. If the
the ACC count while keeping up needed exactness. sensed value goes beyond the threshold values set in
Game plan control was also improved by accepting N- the program, the pump will be automatically switched
maxima procedure in the control point count. ON/OFF by the relay circuit and it is connected to the
Preliminary comes about showed that the proposed driver circuit which helps to switch the voltage. The
system defeated each and every other procedure for farmer will be intimated about the current field
region-based methods. J. Gutierrez et al [11] executed condition through Wi-Fi module and also updated in
amazingly capable automated system with remote the web page. By using this system, the farmer can
sensor arrange. The remote interface was given access the details about the condition of the field
through the GPRS module (MTSMC-G2-SP). The anywhere at any time.
structure has two information parameters at first is

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


194
Anjali Dokhande et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 191-196

V. COMPONENTS USED
• Wi-Fi Module
• Soil Moisture Sensor The ESP8266 Wi-Fi module is a self-contained SOC
Soil Moisture sensor is used to measure the moisture (System on Chip) with integrated TCP/IP
content present in the soil. When the soil moisture (Transmission Control Protocol/Internet Protocol)
value read by the sensor is above the threshold value, protocol stack that can give any microcontroller
low level (0V) will be the digital output and if it is access to any Wi-Fi network. Each ESP8266 module
below the threshold level, high level (5V) will be the comes preprogrammed meaning, it can be simply
digital output. The digital pin is used to directly read hooked up to Arduino device to get Wi-Fi ability. This
current soil moisture value to see if it is above module has a powerful enough on-boarding process
threshold or not. The threshold voltage can be and high storage capacity that allows it to be
regulated with help of potentiometer. integrated with the sensors and other application
specific devices.

VI. APPLICATION

• On different region of land: It can be used on


shallow soil-hilltops, hillsides and throughout
mountain regions.
• For Greenhouse: Smart Greenhouse parameter-
watering of crops, on fertilizers, temperature and
moisture level.
• Very Efficient for Paddy, Rice Fields-This plant
Fig 3. Moisture Sensor requires much water that’s why apply system on
their farm.
• DHT11 SENSOR • Irrigation in Fields: This is used to give watering
DHT11 sensor is used for measuring temperature and to the crop fields.
humidity. It uses a capacitive humidity sensor and a • Food production and safety etc.
thermistor to measure the surrounding air. This sensor
is cost effective, provides low power consumption and VII.ADVANTAGES
up-to 20 meter signal transmission is possible.
• Increase the productivity: Productivity on
farmland is going too increased.
• Reduce water consumption.
• No manpower required.
• Reduce soil erosion and nutrient leaching.
• Cost effective method.
• High quality crop production.
• System not damage by weathers and birds.
Fig 4. DHT11 Sensor
• Efficient use of water.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


195
Anjali Dokhande et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 191-196

VIII. CONCLUSIONS Instrumentation and Measurement, Vol. 53, No.


4, pp. 891-898, May 2004.
In light of all above talk it can be reason that [7]. Darold Wobschall, A frequency shift dielectric
programmed irrigation framework utilizing remote soil moisture sensor, IEEE Transactions on
innovation can give effective framework equipped for Geoscience Electronics, Vol. 16, No. 2, pp. 112-
moderating assets and human exertion. The 118, Apr. 1978.
framework likewise encourages ongoing remote [8]. S. Saxena and G. M. Tayal, Capacitive moisture
monitoring of the current ecological state of field. meter, IEEE Transaction on Industrial
Present day innovation can be consolidated to let Electronics and Control Instrumentation, Vol.
down the cost. These electronic frameworks are 28, No. 1, pp. 37-39, Feb. 1981.
control proficient thus devours less power and depend [9]. S. Mathurkar and D. Chaudhari, Smart Sensors
on auxiliary sources such sunlight based vitality for Based Monitoring System for Agriculture using
finish independence. Field Programmable Gate Array”, International
Journal of Innovative Technology and Exploring
IX. REFERENCES Engineering (IJITEE), Vol. 3, Issue 4, May 2013.
[10]. X. Wang, W. Yang, A. Wheaton, N. Cooley, and
[1]. Plant Growth Factors: Temperature, Colorado B. Moran, Efficient registration of optical and IR
State University, And Available (as on 14-09- images for automatic plant water stress
2015) at: assessment, Comput. Electron. Agricult., vol. 74,
http://www.ext.colostate.edu/mg/Gardenno- no. 2, pp. 230–237, Nov. 2010.
tes/143.html#heat. [11]. J. Gutiérrez, J. F. Villa-Medina, A. Nieto-
[2]. Plant Growth Factors: Water, Colorado State Garibay, and M. Á. Porta-Gándara, Automated
University, And Available (as on 14-09-2015) at: irrigation system using a wireless sensor
http://www.ext.colostate.edu/- network and GPRS module, IEEE transactions
mg/Gardennotes/144.html. on instrumentation and measurement, Vol. 16,
[3]. Harriot Bigas (Ed.), The Global Water Crisis: Issue 1, pp. 166-176, 2013.
addressing an urgent security issue‖, Paper for [12]. P. Bhosale and V. Dixit, Water saving-irrigation
InterAction Council, Hamilton, Canada: UNO- automatic agricultural controller, International
INWEH, 2011-12. Journal of Scientific and Technology Research,
[4]. A. I. Johnson, Methods of measuring soil Vol. 1, Issue 11, Dec. 2012.
moisture in the field‖, U. S. Geological Survey,
1992 (Third Reprint). Cite this article as :
[5]. V. Divya, A. Akhouri, C. Kumar, R. Rishabh, R. Anjali Dokhande, Chetna Bomble, Rakshanda Patil,
Bagla, A Real time implementation of a GSM Puja Khandekar, Nayan Dhone, Prof. Chandrashekhar
based automated irrigation control system using Gode , "A Review Paper on IoT Based Smart Irrigation
drip irrigation methodology, International System", International Journal of Scientific Research
Journal of Scientific and Engineering Research, in Computer Science, Engineering and Information
Vol. 4, Issue 5, pp. 146-151, May 2013. Technology (IJSRCSEIT), ISSN : 2456-3307, Volume 5
[6]. F. Adamo, An acoustic method for soil moisture Issue 1, pp. 191-196, January-February 2019.
measurement, IEEE transactions on Journal URL : http://ijsrcseit.com/CSEIT195155

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


196
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/ CSEIT195153

Marg-Sudhar : Pothole Reporting System Using Android Smart


Phone
Riya Agrawal, Payal Patel, Yash Telang, Ankita Bhandarkar, Irfan Mushtaq, Dr. S. V. Sonekar
Department of Computer Engineering, JD College of Engineering and Management, Nagpur, India

ABSTRACT

Marg-Sudhar: Pothole Reporting System is a mobile application for detecting and reporting road condition.
Dangerous road surface conditions are major distractions for safe and comfortable transportation and therefore
road damage needs to be reported by locality. Data can be gathered by clicking image to identify potholes and
other severe road surface anomalies using geotagging technique and upload on the server.
Keywords : Potholes, GPRS, Geotagging, Road Maintainance, Strategy i.e. Click, "Report" and "Fix".

I. INTRODUCTION caused due to potholes and government authorities


try hard to fix each and every pothole but due to
One major problem in developing countries is large amount of roads in India it has become rather
maintainence of road. Since India is a developing difficult to locate each and every pothole and get it
nation there is a constant demand for good quality fixed.
infrastructure, transportation and services. But since
India is a huge country with quite a sizable So to reduce time consumption in visiting the sites
population this problem still has not been addressed the describe system can be use to directly report to
in totality. the highway Authority about the potholes with the
help of android based application. By this a lot of
In order to improve the conditions of roads efforts time gets reduced of authorities and they get
began way back in the 1980s. It is during this time information about the where about of potholes and
that roads were built to link major highways, to the preventive measures can be taken on time and a
expand the width of existing roads and to construct lot of casualties can be reduced from occurring.
important bridges.

Most roads in India are narrow and congested with


poor surface quality and road maintenance need are
not satisfactorily meet.

Our topic basically concerns about the safety of road


users. Potholes, formed due to heavy rains and
movement of heavy vehicles, also become a major
reason for traumatic accidents and loss of human lives.
As it is seen that many disasters occurring on road are Fig. 1. Condition of roads with potholes

CSEIT195153 | Received : 25 Jan 2019 | Accepted : 05 Feb 2019 | January-February -2019 [ 5 (1) : 197-202 ]
197
Riya Agrawal et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 197-202

II. PROPOSED SYSTEM

Marg-Sudhar is a android based mobile application


for reporting potholes and various road anomalies.

It works on strategy that is “Click”, “Report” and


“Fix”. Using app Marg-Sudhar users can click picture
which use geotagging technique along with
description and this information gets uploaded on
server.

The pothole information obtained from pothole Fig. 2. Pothole detection system and its application
reporting system is sent to server and can be applied
to a pothole alert service and road management Application has been implemented on following
system. All uploaded information are shown as a list android features:
view on client side. Also, the obtained pothole
information is provided to Road Management System. • For designing part various activities and their
components are used.
III. IMPLEMENTATION • Communication between activities has been
implemented using Intent object. Intent is an
Marg-Sudhar has been implemented using following messaging object use to request an action from
technologies (1) Android (2) Java (3) SQL (4) another app component.
Geotagging (5) Web Technologies. • Components registration, is an entry point
through which the system or a user can enter into
(1) Android: app.
• Also, various background services rights has been
Android studio is the IDE for google’s Android OS used.
build on JetBrains’ InteliJ IDEA software and • This application needs various build app access
designed specifically for android development. The rights such as: Storage, Camera, Data access,
following features are provided in the current stable Internet access, GPRS access.
version: Gradel-Based build support, Android-specific • Data is Authenticated due to security purpose at
refactoring and quick fixes, template-based vizards to various point such as at time of login.
create common android designs and components, a
rich layout editor that allows user to drag-and-drop (2) Java:
UI components, build-in support for google cloud
platform, enabling integration with firebase cloud Java is an object-oriented, class-based, concurrent,
messaging and google app engine. secured and general-purpose computer-programming
language.

After front-end elements are finalized java classes and


packages are use to make them work.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


198
Riya Agrawal et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 197-202

(3) Database: (1) Client Side:

A database is an organised collection of data,


generally stored and accessed electronically from a
computer system where databases are more complex
they are often develop using formal design and
modelling techniques.

Android comes with in-built SQLite database. SQLite


supports all the relational database features.

(4) Geotagging:

Geotagging or GeoTagging, is the process of adding


geographical identification metadata to various media Fig. 3. Client Side Working
such as a geotagged photographh or video, websites, Client side involves user interaction. Using app user
SMS messages, QR Codes. This data usually consists can login/ register. If user gets successfully login or
of latitude and longitude coordinates, though they register they can click image of pothole. The snapped
can also include altitude, bearing, distance, accuracy image is automatically geotagged with latitude &
data, and place names, and perhaps a time stamp. longitude using GPS and GPRS. This image is sent on
There are two main options for geotagging photos: server along with some description added by user.
Additionally, user can also see lists of complaints
capturing GPS information at the time the photo is
added by other users and their status.
taken or "attaching" the photograph to a map after
the picture is taken. In order to capture GPS data at (2) Server Side:
the time the photograph is captured, the user must
have a camera with built in GPS or a standalone GPS
along with a digital camera.

IV. WORKING

Marg-Sudhar: Pothole reporting system is


implemented into two modules: (1) Client side, (2)
Server Side.
Fig. 4. Server Side Working

Image via app is uploaded to web server and


complaint is stored on server for further processing
and authenticity. Also admins get all the privileges
through this server. Confirmation SMS and
continuous status for these complaint are send to
uploader.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


199
Riya Agrawal et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 197-202

V. APPLICATION LAYOUT (3) Launch Complaint:

(1) Login/ Registration screen:

(2) Home Screen:

(4) List View:

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


200
Riya Agrawal et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 197-202

(5) Server: VIII. CONCLUSION

Marg-Sudhar will enable road maintainers to obtain


information regarding road condition easily with the
help of locality. Through this it will reduce time of
Municipal corporation in gathering information and
locality will get better roads in less time.

IX. ACKNOWLEDGEMENT

First, and foremost we would like to thank God for


the wonderful opportunities and challenges he has
given to us. We express our sincere gratitude to our
VI. ADVANTAGES & LIMITATIONS guide Dr. S.V. Sonekar for providing his valuable
guidance, patience and for encouraging us to do our
(A) Advantages: best. We wish to thank Prof. P. Lohe, Head,
• It becomes easy for road maintainers as well as for Department of Computer Science and Engineering
locality to monitor the road condition. for his valuable contribution in fulfilling the
• Smooth roads will lead to less vehicle damage and requirement related to the project.
government investment.
• Rapid increase in the rate of Smartphone, users X. REFRENCES
can help to identify Potholes easily.
[1]. Pavan Gorade, Digambar Karde, Poonam
(B) Limitations: Khilare, Abhishek Sontakke, Prof.Ranjana
• To complain on real time basis an android mobile M.Kedar, “Real Time Pothole Tracking System
with GPRS connection is required. Using Android Smart Phone”, International
• It is completely based on user interaction and Journal of Advanced Research in Computer
requires good internet connectivity. Engineering & Technology (IJARCET) Volume
3 Issue 5, May 2014.
VII.FUTURE SCOPES [2]. Salman, Waheed Ahmad, Riaz Ali, Salman
Saleem, “a research on mobile applications for
• Can be implemented in cities under Municipal location tracking through web server and short
Corporation. messages services (sms) “,vfast Transitions on
• It can also be implemented in automated manner Software Engineering Volume 7, Number 02,
using GPS sensors on vehicle July-August, 2015.
• Further research is necessary for improvement in [3]. Lizeth Chandi, Catarina Silva, “Mobile
image processing. application development process: A practical
experience”.
[4]. Seung-Ki Ryu,1 Taehyeong Kim,1 and Young-
Ro Kim2, “Image-Based Pothole Detection
System for ITS Service and Road Management
System”, Hindawi Publishing Corporation
Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com
201
Riya Agrawal et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 197-202

Mathematical Problems in Engineering Volume


2015.

Cite this article as :

Riya Agrawal, Payal Patel, Yash Telang, Ankita


Bhandarkar, Irfan Mushtaq, Dr. S. V. Sonekar, "Marg-
Sudhar : Pothole Reporting System Using Android
Smart Phone", International Journal of Scientific
Research in Computer Science, Engineering and
Information Technology (IJSRCSEIT), ISSN : 2456-
3307, Volume 5 Issue 1, pp. 197-202, January-
February 2019.
Journal URL : http://ijsrcseit.com/CSEIT195153

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


202
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195152

Smart Meals
Shivali Singh1, Aishwarya Narnaware1, Kirtikumar Jain1, Kash Patel1, Prof. S Dhawale2
1Department of Computer Science Engineering, Rashtrasant Tukadoji Maharaj Nagpur University,
Nagpur, Maharashtra, India
2 Professor, Department of Computer Science Engineering, J D College of Engineering and Management,
Nagpur, Maharashtra, India
ABSTRACT

Online business, also known as e-commerce or electronic commerce, refers to transaction and the activities that
have been performed by the customers and seller over the business is facilitated by the web based technologies.
Online food delivery system is one of the most popular online business products. Customers are able to view
and select their favourite food from the list, add to the cart, choose the delivery types, make payments and the
order is complete.
Keywords : Automated Food Ordering System, Dynamic Database Management, Global Positioning System

I. INTRODUCTION application front end is developed using .NET


technology and the back end is developed by using
With the online food ordering system, food is MySQL.
delivered online to the customer. This is possible
using electronic gadgets. Due to rapid growth in the Thus time as well as cost of the developers will be
use of internet and the technologies associated with it, reduced. In our system the user will first register and
several opportunities are coming up on the web. So login themselves into the database. So that he/she can
many business and companies are undertaking into then order their food and make their transaction
their business because of internet. happen.
II. PROBLEM STATEMENT
In this project, an efficient online food ordering
system is proposed. By using the shortest path To create a database of the eating joints falling within
algorithm, we have established a system to reduce the the origin and destination points of the journey and
time consumption of the customers. Also we have placing the orders as per the available menu at
used Heuristic algorithm for the same. In today’s particulars eating joint with the time of service. Our
world, Android and IOS operating system are more idea is to develop an application that provides the
popular. travellers, the meal without any time consumption.
Basically the application will contain the information
Different types of programming language is used for about the hotels i.e. the eating joints that will provide
such operating system. So, in order to mitigate these the food to the travellers.
problems, web base solutions can be used. Here we
came with a food delivery application, where the

CSEIT195152 | Received : 25 Jan 2019 | Accepted : 05 Feb 2019 | January-February -2019 [ 5 (1) : 203-206 ]
203
Shivali Singh et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 203-206

III. LITERATURE REVIEW by the restaurant owners to adopt information and


communication technologies like PDA, wireless
In [1] a food ordering system is proposed with LANs, costly multi-touch screens, etc. to enhance the
Delivery Routing Optimization using GPS dining experience. This paper highlights limitations
Technology and Google Maps. In this paper they based on PDA base food ordering system using an
have explained about the use of GPS and Google Map android smartphone or tablet as a solution. The
for ordering purpose. The major problem in this customer tablet, kitchen display connects with each
paper is routing optimization. The system also uses other through Wi-Fi. Orders did by the customer
GPS in Android mobiles and take advantage of then reach the kitchen module instantly. This
Google Map for map solution. wireless application is user friendly which improves
accuracy and efficiency for hotels by saving their
In [2] a Digital Ordering System for Restaurant using time, reduces human errors and provides customer
Android was proposed. In this application they feedback.
combine the hotel management system by web IV. PROPOSED SYSTEM
technologies. Digital Hotel Management combines
lots of system of restaurant and hotels such as To overcome the limitations of the above system, an
Ordering System Kitchen Order Ticket, Billing Online Food Ordering System based on web based
System Customer Relationship Management System technologies is proposed. Here .NET technology is
together. This combine solution can add or expand mainly used. Our main objective in the project is in
the software system in any size of hotel chain designing completely automated menu in hotels with
environment. This system increases the speed of the help of Android phones using Wi-Fi module and
service and quality. This system also increases to provide a user-friendly environment. The menu
attraction of many customers. will be displayed automatically on the customer
mobile application using wireless connectivity and
In [3] this system explained Implementing we can directly order the menu with the help of press
Customizable Online Food Ordering System Using on the menu. The greatest advantage of this system is
Web Based Application. Previously the process its flexibility.
requires that the customer has to be seated in the
hotels before starting. An alternative method of this ASP.NET is an open-source server side web
for the customer is that ‘Food Pre-Ordering System application framework which is designed for web
using Web Based Application’ in which customer can development to produce dynamic web-sites, web-
be able to do the order before they approach to the applications and web services.
hotel, the saved order can be confirmed by touching
the smartphones. The list of theie selected order There are following steps in our process which first
items shall be shown on the kitchen screen, and includes that the customer has to log in to our system.
when confirmed, order slip shall be printed for If he is new to the system, he need to first register
further processing. This solution provides the easy himself and then log in to the system. The next page
and convenient way to select the order transaction that will appear will be the menu card where the
from customers. customer needs to choose the items from the list and
place his order. This order is collected and
In [4] an Implementation of Smart Restaurant with e- acknowledge by our system. After this he has to fill
menu card is proposed. Lots of work have been done all regarding details like his credit card details, etc.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


204
Shivali Singh et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 203-206

Now the verification of the customer details will be shortest path. In this project through .NET
done to see that whether it is valid or invalid. So for technology we have created login page, checking our
this we have to contact the credit card company. password and through MySQL we created database.
After this if we find all the details correct in all After login our site will show variety of nearby hotels
means then the order is placed successfully. But if through which a customer can order the meals.
given details mismatch then the order will be
cancelled. In the future research, we suggest adding more
variable for routing optimization process such as type
Now, the last step is to make the product ready and vehicle, the size of food packages, holiday session the
get the delivery. driver license type owned by delivery service and the
maximum capacity of the vehicle type. These
ADVANTAGES additional variables will add the complexity to our
application. This system can also enhance to produce
• Online Ordering and easy to use for all. an online system by adding the email notification
• Freedom for demanding food. feature for order confirmation with official receipt.
• Food available on decided places.
• Remove time consumption. VI. ACKNOWLEDGMENT
• Meal comes with good packing.
• Hygiene has first priority. Special Thanks to our Head of Department CSE/IT
• Hassle-free food preparation. Prof Prasanna Lohe , for his valuable suggestions and
encouragement.
FLOW DIAGRAM
VII.REFERENCES

[1]. Roy Deddy Hasiolan Tobing, “A Food Ordering


System with Delivery Routing Optimization
Using Global Positioning System Technology and
Google Maps”, Internetworking Indonesia
Journal, vol. 8, 2016.
[2]. Apurava Joshi, Prachi Oke, Niranjan Jadhav,
Ashutosh Bhargave, Prof. Mr. S.R. Lahane,
“Digital Ordering System for Restaurant Using
Android”. In International Journal of Scientific
and Research Publications, Volume 3, Issue
4April 2013.
[3]. Varsha Chavan, Priya Jadhav, Sneha Korade,
V. CONCLUSION Priyanka Teli,Prof. Mr. R.B.Anpat
“Implementing Digital Restaurant and Inter-
In this project, we have worked on the A* algorithm Restaurant Navigation Using Smart Phone.” in
for GPS tracking, All Pair Shortest Path Algorithm International Journal of Computer Science and
for finding the nearest route of the customer, Mobile Computing, Volume 4. Issue 2, February
Heuristic Search algorithm for again to find out the 2015.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


205
Shivali Singh et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 203-206

[4]. Mayur D. Jakhete, Piyush C. Mankar,


“Implementation of Smart Restaurant with e-
menu card.” in International Journal of Computer
Application, Volume 119, June 2015.
[5]. N. M. Z. Hashim “Smart Ordering System via
Bluetooth” in International Journal of Computer
Trends and Technology (IJCTT) – volume 4 Issue
7–Month 2013
[6]. S. R. Patil “e-converse An affordable touch screen
solution to intrigue Dining experience”

Cite this article as :

Shivali Singh, Aishwarya Narnaware, Kirtikumar Jain,


Kash Patel, Prof. S Dhawale, "Smart Meals",
International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 203-206, January-February 2019.
Journal URL : http://ijsrcseit.com/CSEIT195152

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


206
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195159

A Review on Different Face Recognition Techniques


Aishwarya Admane1, Afrin Sheikh1, Sneha Paunikar1, Shruti Jawade1, Shubhangi Wadbude1,
Prof. M. J. Sawarkar2
1 BE Students, Department of Computer Science & Engineering, Priyadarshini J. L. College of Engineering,
Nagpur, Maharashtra, India
2Assistant Professor, Department of Computer Science & Engineering, Priyadarshini J. L. College of
Engineering, Nagpur, Maharashtra, India

ABSTRACT

Face recognition has been a quickly developing, testing and intriguing region progressively applications. An
extensive number of face recognition calculations have been produced in a decade ago. In this paper an
endeavor is made to survey an extensive variety of techniques utilized for face recognition exhaustively. This
incorporate PCA, LDA, ICA, SVM, Gabor wavelet delicate registering instrument like ANN for recognition and
different cross breed blend of this systems. This audit examines every one of these techniques with parameters
that difficulties face recognition like illumination, pose variation, facial expressions.
Keywords: Principal Component Analysis (PCA), Linear Discriminant Analysis (LDA), Face Recognition,
Independent, Component Analysis (ICA), Artificial Neural Networks (ANN)

I. INTRODUCTION the 1970s [4] and after the fundamental work of


Kanade [5].
Face recognition is an essential piece of the ability of
human discernment framework and is a standard In 1995, an audit paper [6] gave an exhaustive review
errand for people, while building a comparable of face recognition innovation around then [7].
computational model of face recognition. The Around then, video-based face recognition was still in
computational model add to hypothetical bits of a beginning stage. Amid the previous decades, face
knowledge as well as to numerous down to earth recognition has gotten expanded consideration and
applications like mechanized group observation, get to has progressed actually. Numerous business
control, outline of human PC interface (HCI), content frameworks for still face recognition are presently
based picture database administration, criminal accessible. As of late, noteworthy research endeavors
distinguishing proof et cetera. The soonest chip away have been centered around video-based face
at face recognition can be followed back in any event displaying/following, recognition and framework
to the 1950s in brain science [1] and to the 1960s in incorporation. New databases have been made and
the designing writing [2]. A portion of the soonest assessments of recognition strategies utilizing these
examines incorporate work on facial demeanor databases have been done. Presently, the face
feelings by Darwin [3]. In any case, inquire about on recognition has turned out to be a standout amongst
programmed machine recognition of faces began in the most dynamic uses of example recognition, picture
examination and comprehension.

CSEIT195159 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 207-213 ]
207
Aishwarya Admane et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 207-213

II. FACE RECOGNITION ALGORITHMS LDA is more touchy than PCA and ICA on fractional
impediments, however PCA is less delicate to halfway
A. Principal Component Analysis (PCA) impediments contrasted with LDA and ICA. PCA is
PCA otherwise called Karhunen-Loeve technique is utilized as a measurement lessening method in [12]
one of the famous strategies for include determination and for demonstrating articulation disfigurements in
and measurement lessening. Recognition of human [13].
faces utilizing PCA was first done by Turk and
Pentland [8] and recreation of human faces was A recursive calculation for figuring the discriminant
finished by Kirby and Sirovich [9]. The recognition highlights of PCA-LDA method is presented in [14].
technique, known as eigenface strategy characterizes This technique focuses on testing issue of figuring
an element space which lessens the dimensionality of separating vectors from an incrementally arriving
the first information space. This diminished high dimensional information stream without
information space is utilized for recognition. In any processing the comparing covariance grid and without
case, poor separating power inside the class and huge knowing the information ahead of time. The proposed
calculation are the outstanding basic issues in PCA incremental PCA-LDA calculation is exceptionally
technique. This restriction is overwhelmed by Linear proficient in memory utilization and it is extremely
Discriminant Analysis (LDA). LDA is the most productive in the estimation of first premise vectors.
predominant calculations for highlight choice in This calculation gives an adequate face recognition
appearance based techniques [9]. achievement rate in correlation with extremely
acclaimed face recognition calculations, for example,
Be that as it may, numerous LDA based face PCA and LDA. Two appearance– based systems, for
recognition framework initially utilized PCA to lessen example, Modified PCA (MPCA) and Locality
measurements and after that LDA is utilized to boost Preserving Projections (LPP) are consolidated in [15]
the separating energy of highlight determination. The to give a high face recognition rate. PCA is utilized as
reason is that LDA has the little specimen measure a component extraction strategy in [16]. These
issue in which dataset chose ought to have bigger component vectors are thought about utilizing
examples per class for good segregating highlights Mahalanobis separations for basic leadership. Tensor
extraction. Accordingly executing LDA based Multilinear PCA approach is proposed in [17]
straightforwardly brought about poor extraction of which removes highlight specifically from the tensor
separating highlights. In the proposed strategy [10] portrayal as opposed to the vector portrayal. This
Gabor channel is utilized to channel frontal face technique demonstrates a superior execution in
pictures and PCA is utilized to decrease the correlation with the outstanding strategies in separate
measurement of separated component vectors and changing situations.
after that LDA is utilized for highlight extraction. The
exhibitions of appearance based factual techniques, for PCA can outflank over numerous different systems
example, PCA, LDA and ICA are tried and looked at when the span of database is little. In proposed
for the recognition of hued faces pictures in [11]. PCA calculation [18] the database was sub grouped utilizing
is superior to LDA and ICA under various a few highlights of enthusiasm for faces. Just a single
illumination variations however LDA is superior to of the acquired subgroups was given by PCA to
ICA. recognition. Notwithstanding the great aftereffects of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


208
Aishwarya Admane et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 207-213

PCA, this system has the inconvenience of being Another face recognition technique in light of PCA,
computationally costly and complex with the LDA and neural system were proposed in [21]. This
expansion in database estimate, since every one of the technique comprises of four stages:
pixels in the picture are important to get the portrayal
used to coordinate the information picture with all i. Preprocessing
others in the database. ii. Dimension lessening utilizing PCA
iii. include extraction utilizing LDA
Diverse dimensionality diminishment methods, for iv. grouping utilizing neural system.
example, PCA, Kernel PCA, LDA, Locality
safeguarding Projections and Neighborhood Blend of PCA and LDA were utilized for enhancing
Preserving inserting were chosen and connected so as the capacity of LDA when a couple of tests of pictures
to decrease the loss of grouping execution due to were accessible and neural classifier was utilized to
changes in facial appearance. The execution of diminish number misclassification caused by not-
recognition while utilizing PCA and in addition LDA directly distinguishable classes. The proposed strategy
for dimensionality diminishment is by all accounts was tried on Yale face database. Exploratory outcomes
parallel as far as precision. However, it was watched on this database showed the viability of the proposed
that LDA requires long time for preparing more technique for face recognition with less
number of numerous face pictures notwithstanding misclassification in examination with past strategies.
for little databases. In the event of Locality Preserving
Projections (LPP) and NPE techniques, the An alternate approach for face recognition was
recognition rate was less if expanding number of face proposed in [22] which limit calculation time while
pictures were utilized when contrasted with that of accomplishing higher discovery precision. PCA was
PCA and KPCA strategies. The proposed technique utilized to decrease the measurement separating a
[19] gave significant enhancements on account of component vector. GRNN utilized as a capacity
illumination variations, PCA and bit PCA are the best estimate system to identify whether the information
entertainers. picture contains a face or not and if existed then
reports about its introduction. The proposed
Altered PCA calculation for face recognition were framework had demonstrated that GRNN can perform
proposed in [20], this strategy depended on superior to back propagation calculation and give
diminishing the impact of eigenvectors related with some answer for better regularization.
the expansive eigen esteems by normalizing the
component vector component by its comparing B. Support Vector Machine (SVM)
standard deviation. The reenactment comes about Support Vector Machines (SVM) are a standout
demonstrate that the proposed strategy brings about a amongst the most valuable systems in grouping issues.
superior execution than customary PCA and LDA One clear case is face recognition. Be that as it may,
approaches and the computational cost continues as SVM can't be connected when the component vectors
before as that of PCA and significantly less than that characterizing tests have missing sections. A grouping
of LDA. calculation that has effectively been utilized as a part
of this structure is the all-known Support Vector
Machines (SVM) [23], which can be connected to the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


209
Aishwarya Admane et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 207-213

first appearance space or a subspace of it acquired in This approach depended on the exhibitions of a firmly
the wake of applying an element extraction technique segregating optical connection strategy alongside the
[24] [25] [26]. The benefit of SVM classifier over heartiness of the ICA show. Independent component
customary neural system is that SVMs can accomplish analysis (ICA) demonstrate had started enthusiasm for
better speculation execution. looking for a direct change to express an arrangement
of arbitrary factors as straight mixes of factually
C. Independent Component Analysis (ICA) independent source factors [36]. ICA gave a more
Independent component analysis (ICA) is a strategy intense information portrayal than PCA as its
for finding basic elements or components from objective was that of giving an independent instead of
multivariate (multidimensional) measurable uncorrelated picture deterioration and portrayal. A
information. There is have to execute face recognition quick incremental primary non Guassian headings
framework utilizing ICA for facial pictures having analysis calculation called IPCA_ICA was proposed in
face introductions and distinctive illumination [37].This calculation figures the key components of an
conditions, which will give better outcomes as arrangement of picture vectors incrementally without
contrasted and existing frameworks [27] [28] [29] . evaluating the covariance grid and in the meantime
What recognizes ICA from different strategies is that, change these important components to the
it searches for component that is both measurably independent bearings that amplify the non-
independent and non-gaussian [27]. The ICA is like Guassianity of the source. IPCA_ICA is extremely
visually impaired source detachment issue [30] that effective in the estimation of the main premise vectors.
comes down to finding a straight portrayal in which PCA_ICA makes higher normal progress rate than
the components are measurably independent. The Eigenface, the Fisherface and FastICA strategies.
correlation of face recognition utilizing PCA and ICA
on FERET database with various classifiers [31] [32] III. CONCLUSIONS
were talked about and discovered that the ICA would
be advised to recognition rate as contrasted and PCA This paper has endeavored to survey significant
with factually independent premise pictures and number of papers to cover the current advancement
furthermore with measurably independent in the field of face recognition. Introduce think about
coefficients. Face recognition utilizing ICA with uncovers that for upgraded face recognition new
extensive revolution edges with poses and variations calculation needs to advance utilizing half and half
in illumination conditions was proposed in [33]. A strategies for delicate figuring apparatuses, for
novel subspace technique called successive line example, ANN, SVM, SOM may yields better
segment independent component analysis for face execution. The rundown of references to give more
recognition is proposed in [34]. In ICA each face itemized comprehension of the methodologies
picture is changed into a vector before figuring the portrayed is enrolled. We apologize to specialists
independent components. RC_ICA decreases face whose critical commitments may have been neglected.
recognition blunder and dimensionality of recognition
subspace ends up noticeably littler. A novel system for
face recognition consolidated the independent
component analysis (ICA) demonstrate with the
optical connection procedure was proposed in [35].

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


210
Aishwarya Admane et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 207-213

IV. REFERENCES [11]. Önsen TOYGAR Adnan ACAN ,‖Face


recognition using PCA,LDA and ICA approaches
[1]. Bruner, I. S. and Tagiuri, R. The perception of on colored images‖, Journal Of Electrical and
people. In Handbook of Social Psychology, Vol. Electronics Engineering, vol- 13,2003
2, G. Lindzey, Ed., Addison-Wesley, Reading, [12]. Y. Cheng, C.L. Wang, Z.Y. Li, Y.K. Hou and
MA, 634–654.1954 C.X. Zhao,‖ Multiscale principal contour
[2]. Bledsoe, W. W. The model method in facial direction for varying lighting face
recognition. Tech. rep. PRI:15, Panoramic recognition‖,Proceedings of IEEE 2010
research Inc., Palo Alto, CA.1964 [13]. F. Al-Osaimi·M. Bennamoun · A. Mian,‖ An
[3]. Ekman, P. Ed., Charles Darwin’s The Expression Expression Deformation Approach to Non-rigid
of the Emotions in Man and Animals, Third 3D Face Recognition‖, Springer
Edition, with Introduction, Afterwords and Science+Business Media, LLC 2008
Commentaries by Paul Ekman. Harper- [14]. Issam Dagher,‖Incremental PCA-LDA
Collins/Oxford University Press, New York, algorithm‖, International Journal of Biometrics
NY/London, U.K.1998 and Bioinformatics (IJBB), Volume (4): Issue (2)
[4]. Kelly, M. D. Visual identification of people by [15]. J. Shermina,V. Vasudevan,‖An Efficient Face
computer. Tech. rep. AI-130, Stanford AI recognition System Based on Fusion of MPCA
Project, Stanford, CA. 1970 and LPP‖, American Journal of Scientific
[5]. Kanade, T. Computer recognition of human Research ISSN 1450-223X Issue 11(2010), pp.6-
faces. Birkhauser, Basel, Switzerland, and 19
Stuttgart, Germany 1973 [16]. Ishwar S. Jadhav, V. T. Gaikwad, Gajanan U.
[6]. Chellapa, R., Wilson, C. L., and Sirohey, S. Patil,‖Human Identification Using Face and
Human and machine recognition of faces: A Voice Recognition‖, International Journal of
survey. Proc. IEEE, 83, 705–740.1995 Computer Science and Information
[7]. Samal, A. and Iyengar, P. Automatic recognition Technologies, Vol. 2 (3), 2011
and analysis of human faces and facial [17]. Yun-Hee Han,Keun-Chang Kwak,‖ Face
expressions: A survey. Patt. Recog. 25, 65– Recognition and Representation by Tensor-
77.1992 based MPCA Approach‖, 2010 The 3rd
[8]. M. Turk and A. Pentland, "Eigenfaces for International Conference on Machine Vision
recognition," J. Cognitive Neuroscience,vol. 3, (ICMV 2010)
71-86., 1991. [18]. Neerja,Ekta Walia,‖Face Recognition Using
[9]. D. L. Swets and J. J. Weng, "Using discriminant Improved Fast PCA Algorithm‖,Proceedings of
eigenfeatures for image retrieval", IEEE Trans. IEEE 2008
PAMI., vol. 18, No. 8, 831-836, 1996. [19]. S.Sakthivel,Dr.R.Lakshmipathi,‖Enhancing Face
[10]. C.Magesh Kumar, R.Thiyagarajan , Recognition using Improved Dimensionality
S.P.Natarajan, S.Arulselvi,G.Sainarayanan,‖ Reduction and feature extraction
Gabor features and LDA based Face Recognition Algorithms_An Evaluation with ORL database‖,
with ANN classifier‖,Procedings Of ICETECT International Journal of Engineering Science
2011 and Technology Vol. 2(6), 2010

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


211
Aishwarya Admane et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 207-213

[20]. Lin Luo, M.N.S. Swamy, Eugene I. Plotkin, ―A Applications‖ Neural Networks Research Centre
Modified PCA algorithm for face recognition‖, Helsinki University of Technology P.O. Box
Proceedings of IEEE 2003 5400, FIN-02015 HUT, Finland, Neural
[21]. A. Hossein Sahoolizadeh, B. Zargham Heidari, Networks, 13(4-5):411-430, 2000
and C. Hamid Dehghani,‖ A New Face [31]. Bruce A. Draper, Kyungim Baek, Marian
Recognition Method using PCA, LDA and Stewart Bartlett, ―Recognizing faces with PCA
Neural Network‖, International Journal of and ICA‖, Computer Vision and Image
Electrical and Electronics Engineering 2:8 2008 Understanding 91 (2003) 115-137.
[22]. Feroz Ahmed Siddiky, Mohammed Shamsul [32]. Jian Yang, David Zhang, Jing-yu Yang, ―Is ICA
Alam,Tanveer Ahsan and Mohammed Saifur Significantly Better than PCA for Face
Rahim,‖An Efficient Approach to Rotation Recognition?‖ Proceedings of the Tenth IEEE
Invariant Face detection using PCA,Generalized International Conference on Computer Vision
Regression Neural network and Mahalanobis (ICCV’05) 1550- 5499/05.
Distance by reducing Search space‖,Proceedings [33]. Kailash J. Karande, Sanjay N.Talbar,‖ Face
Of IEEE 2007 Recognition under Variation of Pose and
[23]. Vapnik. Statistical Learning Theory. JohnWiley Illumination using Independent Component
and Sons, New York, 1998. Analysis‖, ICGST-GVIP, ISSN 1687-398X,
[24]. E. Osuna, R. Freund, and F. Girosit. Training Volume (8), Issue (IV), December 2008
support vector machines: an application to face [34]. Quanxue Gao , LeiZhang, DavidZhang,‖
detection. Proc. of CVPR, pages 130–136, 1997. Sequential row–column independent
[25]. B. Heisele, T. Serre, and T. Poggio. A component analysis for face recognition‖,
componentbased framework for face detection Elsevier 2008
and identification. IJCV, 74(2):167–181, 2007. [35]. A. Alfalou and C. Brosseau,‖ A New Robust and
[26]. Q. Tao, D. Chu, and J. Wang. Recursive support Discriminating Method for Face Recognition
vector machines for dimensionality reduction. Based on Correlation Technique and
IEEE Trans. NN, 19(1):189–193, 2008. Independent Component Analysis Model‖,
[27]. Marian Stewart Bartlett, Javier R. Movellan, Optics Letters 36 (2011) 645-647
Terrence J. Sejonowski, ―Face Recognition by [36]. P. Comon, ―Independent component analysis:
Independent Component Analysis‖, IEEE a new concept?‖, Signal Process. 3, 287-314
Transactions on Neural Networks, vol-13, No- (1994).
6,November 2002, PP 1450- 1464. [37]. Issam Dagher and Rabih Nachar,‖Face
[28]. Pong C.Yuen, J.H.Lai, ―Face representation Recognition Using IPCA-ICA algorithm‖, IEEE
using independent component analysis‖, Pattern Transactions On Pattern Analysis and Machine
Recognition 35 (2002) 1247-1257. Intelligence, VOL. 28, NO. 6, JUNE 2006
[29]. Tae-Kyun Kim, Hyunwoo Kim, Wonjum [38]. Arindam Kar, Debotosh Bhattacharjee, Dipak
Hwang, Josef Kittler, ―Independent component Kumar Basu, Mita Nasipuri, Mahantapas
analysis in a local facial residue space for face Kundu,‖ High Performance Human Face
recognition‖, Pattern Recognition 37 (2004) Recognition using Independent High Intensity
[30]. Aapo Hyvärinen and Erkki Oja ―Independent Gabor Wavelet Responses: A Statistical
Component Analysis: Algorithms and Approach‖, International Journal of Computer

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


212
Aishwarya Admane et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 207-213

Science & Emerging Technologies 178 Volume Cite this article as :


2, Issue 1, February 2011
[39]. Lades M, Vorbruggen J, Buhmann J, Lange J, Aishwarya Admane, Afrin Sheikh, Sneha Paunikar,
Cvd Malsburg, Wurtz R, Konen W Distortion Shruti Jawade, Shubhangi Wadbude, Prof. M. J.
invariant object recognition in the dynamic link Sawarkar , "A Review on Different Face Recognition
architecture. IEEE Trans Comput 42(3):300– Techniques", International Journal of Scientific
311,1993 Research in Computer Science, Engineering and
[40]. Wiskott L, Fellous JN, Kruger N, Cvd Malsburg Information Technology (IJSRCSEIT), ISSN : 2456-
Face recogniton by elastic bunch graph 3307, Volume 5 Issue 1, pp. 207-213, January-
matching. IEEE Trans PAMI 197:775–779,1997 February 2019.
Journal URL : http://ijsrcseit.com/CSEIT195159

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


213
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195160

A Survey on Various Bio-Metric Techniques and its Application


for Child Care
Shanon Paul1, Payal Ramteke1, Sugam Sangai1, Minakshee Shende1, Aafreen Qureshi1, Prof. Umesh Samarth2
1 Department of Information Technology, Rashtrasant Tukdoji Maharaj Nagpur University, Nagpur,
Maharashtra, India
2 Professor, Department of Information Technology, JD College of Engineering and Management, Nagpur,
Maharashtra, India

ABSTRACT

With various developing applications requiring biometric acknowledgment of children (e.g., following tyke
immunization plans, recognizing missing children and counteracting infant swaps in emergency clinics),
exploring the worldly solidness of biometric acknowledgment precision for children is critical. The diligence of
acknowledgment precision of three of the most ordinarily utilized biometric qualities (fingerprints, face and iris)
has been examined for grown-ups. Notwithstanding, the industriousness of biometric acknowledgment
precision has not been concentrated methodicallly for children in the age gathering of 0-4 years. Given that
youthful children are regularly uncooperative and don't fathom or adhere to directions, as we would see it,
among all biometric modalities, fingerprints are the most suitable for perceiving children. This is principally in
light of the fact that it is simpler to catch fingerprints of youthful children contrasted with other biometric
attributes, e.g., iris, where a tyke needs to gaze straightforwardly towards the camera to start iris catch. In this
report, we detail our drive to explore the perseverance of unique finger impression acknowledgment for
children.
Keywords : Bio-Metric, Fingerprint, Child Identification

I. INTRODUCTION pay nations commonly don't have any type of ID


reports which can be utilized for this reason, we
One of the significant objectives of most national, address the accompanying inquiry: can fingerprints
worldwide and non-administrative wellbeing be adequately used to perceive children from birth to
associations is to kill the event of immunization- 4 years? We have gathered 1,600 unique mark
preventable youth sicknesses (e.g., polio). Without a pictures (500 ppi) of 20 babies and little children
high inoculation inclusion in a nation or a geological caught over a 30-day time frame in East Lansing,
locale, these lethal illnesses take an overwhelming Michigan and 420 fingerprints of 70 newborn
toll on children. In this manner, it is essential for a children and babies at two diverse wellbeing centers
viable vaccination program to monitor children who in Benin, West Africa. We conceived the
have been inoculated and the individuals who have accompanying systems to enhance the unique finger
gotten the required supporter shots amid the initial 4 impression acknowledgment exactness when looking
years of life to enhance the immunization inclusion. at the gained fingerprints against an all-
Given that children, just as the grown-ups, in low encompassing exhibition database of 32,768 newborn

CSEIT195160 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 214-220 ]
214
Shanon Paul et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 214-220

child fingerprints gathered by VaxTrac in Benin: (I) immunizations obtained, $50 will never go into the
upsample the obtained unique finger impression arm of a tyke in need"3. Accordingly, the kid death
picture to encourage particulars extraction, (ii) rates keep on being high in the low-salary and
coordinate the question print against formats made creating nations.
from every enlistment impression and wire the
match scores, (iii) intertwine the match scores of II. RELATED WORK
the thumb and forefinger, and (iv) refresh the
display with fingerprints procured over various A powerful inoculation program needs to monitor
sessions. which babies and little children have been
vaccinated and how frequently they have gotten
The United Nations Children's Fund (UNICEF's) the required sponsor shots from birth to 4 years old
"2013 Progress Report on Committing to Child (Fig. 1 demonstrates the UNICEF prescribed
Survival: A Promise Renewed" [2] makes reference vaccination plan for newborn children; for CDC
to that while more children currently make due suggested timetable for children up to 6 years see
past their fifth birthday celebration than at any [1]). In creating nations, regularly, there are no
other time, the poorest countries still lose an national distinguishing proof projects which can be
extensive number of children to immunization- utilized to recognize children all through the
preventable illnesses. The 2011 Grand Challenges in vaccination schedule4. This brings up the
Global Health Explorations Round 7 issued by the accompanying issue: can fingerprints, or, so far as
Bill and Melinda Gates Foundation [4] states that that is concerned some other biometric
"every year roughly 25 million babies don't get the methodology, be utilized to distinguish children
vital immunizations1, and in any event, 2.4 million from birth to 4 years old.
children bite the dust from antibody preventable
sicknesses."

With the point of destroying antibody preventable


infections, standard and obligatory immunization
programs are a standard in high salary nations. For
example, as per the Centers for Disease Control and
Prevention (CDC), "in the United States,
inoculation programs have wiped out or
fundamentally decreased numerous antibody
preventable infections" [3]. Thus, the youngster
death rates have diminished extensively in high Fig. 1 : Universally prescribed vaccination plan for
salary nations. Then again, routine inoculation newborn children by UNICEF
programs have not been as powerful in decreasing
the event of immunization-preventable sicknesses A few endeavours have been made to examine the
in low-pay nations. VaxTrac2, a non-legislative suitability of utilizing distinctive biometric qualities
association working in West-African nations, for distinguishing babies and little children. In 1899,
expresses that the "antibody wastage rates are Sir Francis Galton [4] first concentrated the varieties
higher than half is probably the most difficult experienced in the inked unique finger impression
geologies", and "for each $100 in new impressions of a baby caught after some time (from

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


215
Shanon Paul et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 214-220

around 9 days to 4.5 years old). He inferred that it


was not attainable to distinguish newborn children
in the age scope of 0-2.5 years utilizing inked unique
mark impressions. All the more as of late, the Joint
Research Center of the European Commission
distributed a specialized report [5] dedicated to the
(a) (b)
topic of regardless of whether robotized unique mark
Fig. 2 : Use of fingerprints for following the
acknowledgment for children is achievable.
inoculation calendar of newborn children and little
children.
The examination inferred that (I) children can be
recognized utilizing fingerprints when the time
Weingaertner et al. [9] researched the utilization of
contrast between the two caught impressions is
palmprints and impressions for distinguishing infants
under 4.5 years, and (ii) picture quality is a
(0-2 days old). Manual coordinating precision was
conclusive factor in unique mark acknowledgment.
accounted for to be around 83% and in Benin, Africa.
(a) Mothers holding up in a wellbeing center to get
Gottschlich et al. [5] contemplated the impact of
their children inoculated, and (b) a human services
youthful development on the precision of unique
specialist fingerprinting a kid before overseeing
mark coordinating frameworks, and demonstrated
immunization. These pictures were caught by the
that (I) finger impression development can be
creators amid their visit to the immunization bases in
displayed utilizing an isotropic development model,
and on Cotonou, Benin in June 2014. Roughly 67%
and (ii) coordinating exactness of finger impression
for palm prints and impressions of infants, separately.
frameworks can be enhanced by upscaling the
In spite of the fact that various diverse biometric
unique mark pictures utilizing this model while
modalities for distinguishing children have been
coordinating unique finger impression pictures of
investigated, there is no unmistakable agreement on
young people gathered after some time.
(I) regardless of whether it is attainable to perceive
newborn children and little children utilizing
Corby et al. [6] considered the reasonability of
biometrics, and (ii) if biometric acknowledgment is
utilizing business sensors to catch iris pictures of 1.5-
surely plausible, which methodology is most
multi year old children. They announced a high
appropriate for this assignment. In view of various
inability to enlist (FTE) rate of around 57%, in spite
contemplations, for example, simplicity of catch
of the fact that the acknowledgment exactness for
the selected subjectswas extremely high (about 99%). (palmprints are hard to catch since babies and
newborn children keep their clench hands shut),
Tiwari et al. [7] and Bharadwaj et al. [8] caught face
parental concerns (e.g., infrared enlightenment for
pictures of infants (0-3 days old) and inferred that it
iris catch), tirelessness of biometric attribute (facial
was hard to catch great quality face pictures because
qualities change after some time), as we would like to
of (I) net head reflexes, and (ii) posture and
think, fingerprints seem, by all accounts, to be the
demeanor varieties.
most feasible biometric for baby and little child
acknowledgment (see Tab. 1). Undoubtedly, VaxTrac
has built up a versatile immunization library
framework which utilizes fingerprints to recognize
children in Benin1 (Fig. 2(b)). In the VaxTrac

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


216
Shanon Paul et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 214-220

framework, the left and right thumb prints of both The connection-based techniques utilized in spatial
the kid and his mom are gathered. In the event that or in the recurrence area associate two unique mark
the kid's fingerprints can't be coordinated effectively, pictures to process the similitude between them.
mother's fingerprints are utilized for building
up/confirming the youngster's personality. While A tale way to deal with unique mark arrangement
VaxTrac does not report the coordinating exactness and coordinating was proposed by Arun Ross et.al
of children's fingerprints, they notice that they [10]. The list of capabilities utilized in this strategy
constantly end up utilizing the mother's fingerprints was an edge include map. A lot of Gabor channels
for this reason in light of the fact that coordinating pre-tuned to a particular recurrence were utilized to
children's fingerprints flops frequently 5. Proceeded catch the neighborhood edge qualities at different
with endeavors are, thusly, expected to propel the introductions and concentrate the nearby edge
unique finger impression innovation, both detecting attributes. The convolution was performed in the
innovation just as coordinating calculation behind recurrence space. A standard deviation picture that
the versatile immunization library framework. In caught the variety in the edge quality at different
this paper, we present the underlying consequences introductions was developed utilizing the separated
of our progressing study on utilizing fingerprints to pictures. The edge highlight map was acquired
perceive newborn children and little children. utilizing standard deviation picture. A 2D connection
of the edge highlight maps of the question picture
III. LITERATURE SURVEY and format was resolved to produce a coordinating
score.
Countless for programmed unique finger impression
coordinating have been proposed in the writing. Two Karthik Nandakumar and Anil K. Jain [11] proposed a
given fingerprints are thought about by a connection based unique mark matcher that used the
coordinating calculation and return either a level of nearby relationship of districts around the details to
closeness (additionally called coordinating score) or assess the level of comparability between two unique
an acknowledgment/dismissal choice. This finger impression pictures. Particulars focus and the
workgroup the unique finger impression coordinating related edge focuses were separated from the layout
methodologies into connection based, particulars and the inquiry unique finger impression pictures.
based methodologies, and non-details include based, This technique utilized Procrustes examination to get
for example, edge shape, surface data and so forth. a decent gauge of comparing edge bends to adjust the
The motivation behind the study is to break down inquiry to the layout. The two pictures were
those calculations and talk about the execution improved utilizing a bank of Gabor channels of
assessment. various introductions. The standardized cross
correlation was processed to decide the nature of the
The current unique finger impression particulars coordinate. A database comprising of
acknowledgment frameworks utilize methods unique finger impression impressions of 160 clients
dependent on the neighborhood and worldwide were utilized to assess this technique.
element portrayals of the unique mark pictures, for
example, details, edge shape, surface data and so on. A Robust Fingerprint Matching Algorithm for
Verification dependent on connection was
introduced by Abdullah Cavusoglu et. al [12]. They
utilized a fluctuation based division technique to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


217
Shanon Paul et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 214-220

fragment the unique mark picture. Neighborhood the connection between's edges of neighboring
edge introduction was resolved to utilize Sobel particulars that shows the similarity of regions in the
administrators. The strategy used to discover the middle of the two comparing particulars sets were
Reference point depended on the differential entirety processed. The item rule was utilized to join the
of sine estimations of the headings of the pixels quantity of coordinated particulars sets,
situated on a specific sweep. The proposed calculation neighborhood connection score and edge relationship
determined 12 distinctive cross connections score that spoke to an all-out aggregate of
dependent on certain span esteems from the coordinating zones in two fingerprints. The proposed
reference point for both the info and format pictures coordinating calculation recognized real coordinated
and found an all-out total to decide if the pictures details and impostor coordinated particulars in this
relate to a similar unique finger impression. The manner enhancing the coordinating precision.
proposed calculation was assessed on an open space
database. Haiyun Xu et.al [15] exhibited a novel technique for
unearthly details portrayal for unique finger
Koichi Ito et.al [13] proposed a proficient unique impression check. The ghostly particulars spoke to
mark acknowledgment calculation utilizing the stage detail set as a settled length include vector and
parts in 2D Discrete Fourier Transforms of the depended on the move, scale, and revolution
pictures. They utilized Phase-Only Correlation (POC) properties of the two-dimensional (2-D) ceaseless
capacity and Band-Limited Phase-Only Correlation
(BLPOC) capacity to decide the stature and area of Fourier change. The two details portrayal techniques
the crest from stage attributes in Fourier space, which utilized for particulars coordinating were Location-
gave the comparability measure and the translational Based Spectral Minutiae and Orientation-Based
uprooting between the pictures separately. The Spectral Minutiae. To lessen the higher frequencies, a
revolution and the removal between the enlisted Gaussian low-pass channel was connected on the
unique mark and the information finger impression range. Two coordinating calculations were exhibited.
pictures were standardized utilizing BLPOC work. The relationship of two ghostly pictures (direct
The covered locale of the two pictures was removed coordinating) was picked as a comparability score.
lastly the coordinating score is assessed as the The second calculation is the Fourier– Mellin
entirety of the most elevated two pinnacles. The coordinating, in
execution of their calculation was assessed on a
database comprising of 330 unique finger impression which the extent of the Fourier change of the
pictures. particulars range was taken to compute the likeness
score. The proposed calculations were assessed on
Jiang Li et.al [14] depicted a unique mark three unique finger impression databases. A basic
coordinating calculation that consolidates minutia relationship based unique mark confirmation
based coordinating strategy with relationship-based framework was proposed by Asker M. Bazen et.al
coordinating technique. Minutia based coordinating [16]. The proposed framework specifically utilized
calculation was utilized to separate the rundown of the more extravagant dark scale data of the
coordinated minutia sets from the two unique mark fingerprints.
pictures. Relationship of the nearby neighborhood
districts around each coordinating details pair that Suitable trademark layouts were chosen in the
speaks to the neighborhood likeness and furthermore essential unique mark and their relating positions in

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


218
Shanon Paul et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 214-220

the info unique finger impression were resolved. The [2]. Committing to child survival: A promise
format was moved over the whole unique finger renewed.
impression picture and the area where the separation http://www.unicef.org/lac/Committing to Child
is insignificant was picked as the relating position of Survival APR 9 Sept 2013.pdf.
the layout in the info finger impression. The data of [3]. Vaccines and immunizations.
all layout sets was converged to get the ultimate http://www.cdc.gov/vaccines
choice. The proposed framework execution was /vacgen/whatifstop.htm.
assessed on pictures from four diverse unique mark [4]. F. Galton. Finger prints of young children.
databases comprising of 880 fingerprints altogether. British Association for the Advancement of
Science, 1899.
IV. CONCLUSIONS [5]. C. Gottschlich, T. Hotz, R. Lorenz, S.
Bernhardt, M. Hantschel, and A. Munk.
Current biometric information catch and Modeling the growth of fingerprints improves
acknowledgment arrangements cook essentially to matching for adolescents. IEEE Transactions on
grown-ups (more than 16 years old). National ID Information Forensics and Security, 6(3):1165–
programs, for example, India's Aadhaar program, 1169, Sept 2011.
order catching fingerprints and iris pictures of people [6]. P. M. Corby, T. Schleyer, H. Spallek, T. C. Hart,
who are 5 years old or more established. There is R. J. Weyant, A. L. Corby, and W. A. Bretz.
currently a developing requirement for creating Using biometrics for participant identification
capacities to perceive exceptionally youthful children in a research study: a case report. J. Am.
(from infant to 4 years of age) in light of biometrics. Medical Informatics Assoc., 13(2):233–235,
For biometric acknowledgment to be fruitful, 2006.
uniqueness and constancy properties of a biometric [7]. S. Tiwari, A. Singh, and S. K. Singh. Intelligent
characteristic should be fulfilled for the number of method for face recognition of infant.
inhabitants in intrigue. While these two central International Journal of Computer
fundamentals of biometrics have been researched for Applications, 52(4):36–50, 2012.
the three essential biometric characteristics [8]. S. Bharadwaj, H. Bhatt, R. Singh, M. Vatsa, and
(fingerprints, face, and iris) of the grown-up populace, S. Singh. Face recognition for newborns: A
there is no thorough examination led to explore the preliminary study. In 4th IEEE BTAS, pages 1–
perseverance of biometric acknowledgment 6, Sept 2010.
(longitudinal investigation) for children. In this [9]. D.Weingaertner, O. R. P. Bellon, L. Silva, and
examination, we will likely research the tirelessness M. N. Cat. Newborn’s biometric identification:
of unique mark acknowledgment for children in the Can it be done? In VISAPP (1), pages 200– 205,
age gathering of 0-4 years. 2008.
[10]. Arun Ross, James Reisman and Anil Jain,
V. REFERENCES (2002), “Fingerprint Matching Using Feature
Space Correlation”, Biometric Authentication.
[1]. 2014 recommended immunizations for children Springer Berlin Heidelberg, pp. 48-57.
from birth through 6 years old. [11]. Karthik Nandakumar and Anil K. Jain ,(2004),
http://www.cdc.gov/vaccines/ “Local Correlation- based Fingerprint
parents/downloads/parentver- sch-0-6yrs.pdf. Matching”, Conference proceeding of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


219
Shanon Paul et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 214-220

Computer Vision, Graphics and Image Cite this article as :


Processing, Kolkata, India, pp. 503-508.
[12]. Abdullah Cavusoglu and Salih Gorgunoglu, Shanon Paul, Payal Ramteke, Sugam Sangai,
(2007), "A Robust Correlation Based Minakshee Shende, Aafreen Qureshi, Prof. Umesh
Fingerprint Matching Algorithm for Samarth, "A Survey on Various Bio-Metric
Verification", Journal of Applied Sciences 7, pp. Techniques and its Application for Child Care",
3286-3291. International Journal of Scientific Research in
[13]. Koichi Ito, Ayumi Morita, Takafumi Aoki, Computer Science, Engineering and Information
Tatsuo Higuchi, Hiroshi Nakajima, and Koji Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
Kobayashi, (2005), “A Fingerprint Recognition 5 Issue 1, pp. 214-220, January-February 2019.
Algorithm using Phase-Based Image Matching Journal URL : http://ijsrcseit.com/CSEIT195160
for low quality fingerprints”, IEEE
International Conference on Image Processing,
Vol. 2, pp. 33-36.
[14]. Jiang Li, Sergey Tulyakov and Venu
Govindaraju, (2007), “Verifying Fingerprint
Match by Local Correlation Methods”, First
IEEE International Conference on Biometrics:
Theory, Applications, and Systems, pp.1-5.
[15]. Haiyun Xu, Raymond N. J. Veldhuis, Asker M.
Bazen, Tom A. M. Kevenaar, Ton A. H. M.
Akkermans and Berk Gokberk ,(2009),
“Fingerprint Verification Using Spectral
Minutiae Representations”,IEEE Transactions
On Information Forensics And Security, Vol. 4,
No. 3, pp. 397-409.
[16]. Asker M. Bazen, Gerben T.B. Verwaaijen, Sabih
H. Gerez, Leo P.J. Veelenturf and Berend Jan
van der Zwaag, (2000), "A correlation- based
fingerprint verification system", ProRISC 2000
Workshop on Circuits, Systems and Signal
Processing, pp. 205-213

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


220
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT183818
Liver Cancer Detection
L. S. Rohith Anand*, B. Shannmuka, R. Uday Chowdary, K. Satya Sai Krishna
CSE, Vasireddy Venkatadri Institute of Technology, Guntur, Andhra Pradesh, India

ABSTRACT

Machine learning techniques play an important role in building predictive models by learning from Electronic
Health Records (EHR). Predictive models building from Electronic Health Records still remains as a challenge
as the clinical healthcare data is complex in nature and analysing such data is a difficult task. This paper
proposes prediction models built using random forest ensemble by using three different classifiers viz. J48, C4.5
and Naïve Bayes classifiers. The proposed random forest ensemble was used for classifying four stages of liver
cancer. Using a feature selection method the reliable features are identified and this subset serves as input for
the ensemble of classifiers. Further a majority voting mechanism is used to predict the class labels of the liver
cancer data. Experiments were conducted by varying the number of decision trees generated using the J48, C4.5
and Naïve Bayes classifiers and compared with the classification made using decision stump and Adaboost
algorithms.
Keywords : Ensemble, Feature Selection, C4.5, J48 and Random Forest

I. INTRODUCTION format for documenting patient information than


narrative text which is unstructured in nature. EHR
In health care industry, patient’s medical data size helps to streamline the clinical workflow information.
grows day to day. The process of applying computer Ensemble learning is a well-known approach used in
based information system (CBIS), including new machine learning for prediction by combining
techniques, for discovering knowledge from data is various ensemble models [1]. Ensemble of classifiers
called data mining. The process of machine learning is aggregations of multiple classifiers are J48, C4.5 and
is similar to that of data mining. Machine learning Naive Bayes etc. [2]. Ensembles aim for better
algorithms may be distinguished by either supervised performance than any of the base classifiers. The
or unsupervised learning methods. Supervised proposed work aims to improve the accuracy of
learning methods are widely used for predictive healthcare data for prediction and classification, by
modelling. Predictive modelling is a branch of building a hybrid predictive classifier model using
clinical and business intelligence branch which is ensemble of classifiers [3][4].
used for health risk classification and also to predict
the future health status of the individuals. Electronic The remaining part of the paper described in the
health records (EHR) are used to store large scale following section.
information of patient conditions, treatments etc.
The EHR information may be structured or Section 2 describes related works on classification,
unstructured. Using controlled vocabulary, electronic feature selection, subset generation, pre- processing
health records are maintained in structured data and boosting algorithms such as adaptive boosting for

CSEIT183818 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 221-228 ]
221
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

electronic health records. Section 3 explains the constraint projection for mutli-label ensemble. They
overall architecture of proposed system. Section 4 have adopted boosting methods to construct a multi-
reports the experimental results. Section 5 concludes label ensemble to increase the generalization ability
the paper. [6].

II. RELATED WORK Jia Zhua et. al. (2015) employed multiple classifier
systems (MCS) to improve the accuracy of disease
This section discusses the existing methods for pre- detection for Type-2 Diabetes Mellitus. Multi
processing, feature extraction, boosting methods such classifier system performs worse when design is not
as adaptive boosting. Aydin et. al. (2009) investigated proper [7]. They have proposed a dynamic weighted
the various factors involved on ensemble voting scheme for multiple classifier decision
construction using a wide variety of learning combination. Yan Li et. al. (2015) stated data mining
algorithms, data sets and evaluation criteria [5]. They framework for distributed healthcare information
have provided the idea of subset selection to the level based on privacy preserving constraints [8]. Neesha
of discriminating whether the discrimination is Jothi et. al. (2015) surveyed the data mining
applicable or not at the level of classifier.Ping Li et. al. techniques and has classified the articles have
(2013) surveyed about supervisedmulti-label suggested data mining plays important role in
classification and proposed variable pairwise medical diagnosing for predicting diseases [9].

Table 1. Comparative analysis of different ensembles of classifiers

Author Methods Used Data Sets Used Number of Performance Metrics


Iterations

Nikunj C. K nearest neighbor, Learning Datasets from 110 Average, Geometric


Oza vector quantization, Multi-layer UCI repository mean
et.al.method perceptron’s, Radial basis
(2008)[10] functions, Support
vector machines
Yong Seog Naive Bayes, Support vector German credit 120 AUC, Accuracy,
Kim machines data and COII False positive rate,
et.al.method ,Artificial neural networks, 2000 competition Hit rate gain
(2009)[11] Pruned tree classifier data
Hesam Sagha Quadrant discriminant analysis 2 real datasets 45 Entropy, Mutual
et.al.method containing data Information
(2013)[12] from body
mounted inertial
sensors
Ping Li Bagging, Boosting, Random 12 datasets 20 Hamming loss,
et.al.method Forest, Random subspace, including test Ranking loss,
(2013) [6] Rotation forest categorization, One error,
image Coverage, Average

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


222
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

classification and Precision, F1-


bioinformatics metrics, Recall
Ritaban Binary tree, Linear 24 Holstein- 20 AUC, Accuracy
Dutta discriminant analysis classifier, Friesian cows
et.al.method Naïve Bayes Classifier, K- from Tasmanian
(2015) [13] nearest Institute of
neighbor, Adaptive Neuro Agriculture
Fuzzy Inference classifier Dairy Research
Facility
Yang Zhang SVM classifier, BPNN classifier Benchmark 11 Average regression
et.al.method datasets from
(2015)[14] UCI repository
Bing Gong Artificial neural network, Datasets from 30 F measure, G mean
et.al.method Support vector machines, UCI repository
(2016)[15] CART
Yan Li Adaboost algorithm 9948 real world 20 F-measure,
et.al.method EHRs of diabetes Sensitivity,
(2016)[8] patients Precision
Cátia M. Apriori decision, Aposteriori Benchmark 12 AUC,
Salgado decision datasets from Accuracy,
et.al.method UCI collection, Sensitivity,
(2016)[16] MIMIC II Specificity
datasets

Based on the literature survey carried out a proposed system and how limitations in existing
comparative analysis of the ensemble of classification system is resolved.
methods, the data sets used for experiments by
different researchers, the number of iterations for III. PROPOSED WORK
which the experiments were conducted and the
metrics used for measuring the classification accuracy The proposed architecture is shown in Figure 1. The
are tabulated in the table given below. Electronic health records contain features like patient
id, status, age, sex, hepato, ascites, edema, billi,
From the above table the conclusion drawn is an cholestrol, albumin etc. The data considered have to
ensemble of C4.5, J48 and Naïve Bayes classifier with be clinically transformed i.e. to make it suitable for
majority voting scheme was not studied and hence further processing. The clinical transformation step is
this work focusses on building a predictive model also identified as preprocessing step.
based on building a random forest using these three
classifiers. The proposed system has been compared The unprocessed has null values, irrelevant values
with the existing decision stump and Adaboost and noisy values. These data errors would lead to
algorithms. The next section discusses about the misclassification and hence need to be clinically
transformed. The missing data in the considered

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


223
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

dataset is imputed with values computed using mode The proposed system with its role and advantages is
function. discussed. The experimental result analysis of the
proposed work has been discussed in the next section.
After pre-processing of data, for classifying instances
under Random forest, three subsets from the datasets IV. RESULTS AND DISCUSSION
are generated. The subset will be generated
considering three features like platelet count, alkaline Experiment Results
phosphate and cholesterol values.
The proposed system is implemented using Java and
The random forests are built using three classification Weka tool. The liver cancer dataset having 500
algorithms namely C4.5, J48 and Naïve Bayes. There instances and breast cancer dataset are used for the
are many voting mechanisms followed for ensemble experiments. The ensemble of classifiers is used for
of classifiers, here we are using majority vote method classifying these datasets on which voting is
to perform voting with different classifiers. Here the performed.
output will be the final outcome of the majority of
classifiers Pre-processing

Figure 1 shows the architecture of proposed system. In this module, pre-processing of data is done. The
dataset which we have contains null values,
irrelevant values and noisy values. As missing values
in the dataset lead to misprediction of the final result,
the dataset is pre-processed by filling missed values
on basis of mode function. The dataset without pre-
processing which contains irrelevant values is shown
below in figure 2.

Figure 2. Dataset before preprocessing

Figure 1. Architecture of Proposed System The dataset is preprocessed to fill missing and
irrelevant values as shown in figure 3

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


224
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

Figure 5. Results achieved after implementing subset


Figure 3. Dataset after preprocessing
generation

Feature selection
Performance Evaluation

Feature selection method used for model


This section evaluates the performance of J48
construction by choosing a subset of relevant
Random forest classifier, C4.5 Random Forest
predictors. It also called as variable selection or
classifier and Naïve Bayes classifier using Accuracy,
attribute selection [17].
True positive, False positive, Precision, Recall and F-
measure. Figure 6 shows the performance of J48
Random Forest classifier for prediction different
stages of liver cancer. Figure 7 show the performance
of C4.5 Random forest classifier for prediction
different stages of liver cancer and Figure 8 shows the
performance of Naïve Bayes classifier for prediction
different stages of liver cancer.

Figure 4. Dataset after feature selection

Subset Generation

For classifying instances under Random forest, Figure 6. Comparison of J48 Random forest classifier
generate three subsets from the datasets. Subset will at different stages
be generated considering some features as first,
middle and last stages as shown in figure 5.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


225
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

Comparison of Existing and Proposed System

The comparative analysis of classification accuracy


for Liver and Breast cancer dataset shown in figure 9
and also existing and proposed system based on
different threshold values shown in Table 2 and
figure 10.

Figure 7. Comparison of C4.5 Random forest classifier


at different stages

Table 2. Accuracy of Existing and Proposed system


based on different threshold values

Syste Weight Weigh Weight Weight


m ed ted ed ed
Thresh Thresh Thresh Thresh
old old old old
value value value value Figure 9. Comparison of existing and proposed system
100 in 200 in 500 in 1000 in for two different datasets
% % % %
Existi
ng 46 48 46 44
Propo
sed 51 53 50 48

Figure 10. Comparison of existing and proposed


system based on weighted threshold

In this section, we had a brief discussion about


implementation and experimental results of Clinical
feature transformation, Feature selection, Subset
generation, J48 classifier, C4.5 classifier, Naive bayes
classifier and Majority voting.
Figure 8. Comparison of Naïve Bayes classifier at
different stages

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


226
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

V. CONCLUSION construction of classifier and discriminant


ensembles, Information Sciences 179 (2009)
The prediction model is built by series of steps such 1298–1318.
as clinical feature transformation, feature selection, [6] Ping Li , Hong Li , Min Wu , Multi-label
ensemble of classifiers and Majority voting which ensemble based on variable pairwise
aimed to improve rate of correct predictions. The constraint projection, Information Sciences
prediction accuracy is improved by an ensemble of 222 (2013) 269–281.
classifiers and when majority voting mechanism was [7] Jia Zhua,, Qing Xie, Kai Zheng,An improved
applied on them. The proposed system here achieve early detection method of type-2diabetes
an accuracy of 40% for C4.5 Random forest classifier, mellitus using multiple classifier system,
43% for J48 Random forest classifier, 38% for Naïve Information Sciences 292 (2015) 1–14.
bays classifier when tested for Liver cancer dataset [8] Yan Li, Changxin Bai, Chandan K. Reddy, A
which is more than existing system. The accuracy of distributed ensemble approach for mining
45% for C4.5 Random forest classifier, 52% for J48 healthcare data under privacy constraints,
Random forest classifier and 47% for Naïve Bayes Information Sciences (2015).
classifier was achieved when tested for Breast cancer [9] Neesha Jothi, Nur’Aini Abdul Rashid,
dataset which is more than existing system which has Wahidah Husain, Data Mining in
an accuracy of 46%. The number of trees generated Healthcare – A Review, Procedia Computer
was varied and the prediction accuracy of the Science 72 ( 2015 ) 306 – 313.
proposed work was studied. [10] Nikunj C. Oza , Kagan Tumer, Classifier
ensembles: Select real-world applications,
VI. REFERENCES Information Fusion 9 (2008) 4–20.
[11] YongSeog Kim, Boosting and measuring the
[1] Dietterich TG.(2000), Ensemble methods in performance of ensembles for a successful
machine learning. In: Proceedings of database marketing, Expert Systems with
Multiple Classifier System, vol. 1857, Applications 36 (2009) 2161–2176.
Springer (2000), pp. 1–15. [12] Hesam Sagha, Hamidreza Bayati, José del R.
[2] Zhi-Hua Zhou, Ensemble Methods: Millán, Ricardo Chavarriaga, On-line
Foundations and Algorithms, Machine anomaly detection and resilience in classifier
Learning & Pattern Recognition Series, ensembles, Pattern Recognition Letters
2012. (2013).
[3] Yongjun Piao, Minghao Piao, Keun Ho Ryu, [13] Ritaban Duttaa, Daniel Smitha, Richard
Multiclass cancer classification using a Rawnsley, Greg Bishop-Hurley, James Hills,
feature subset- based ensemble from Greg Timms, Dave Henry, Dynamic cattle
microRNA expression profiles, Computers in behavioural classification using supervised
Biology and Medicine 80 (2017) 39–44. ensemble classifiers, Computers and
[4] ManjeevanSeera, Chee Peng Lim, A hybrid Electronics in Agriculture 111 (2015) 18–28.
intelligent system for medical data [14] Yang Zhang, Li Zhang, M.A. Hossain,
classification, Expert Systems with Adaptive 3D facial action intensity
Applications 41 (2014) 2239– 2249. estimation.
[5] Aydın Ulas, Murat Semerci, Olcay Taner [15] Bing Gong, Joaquín Ordieres-Mere,
Yıldız, Ethem Alpaydın, Incremental Prediction of daily maximum ozone

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


227
L. S. Rohith Anand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 221-228

threshold exceedances by preprocessing and


ensemble artificial intelligence techniques:
Case study of Hong Kong, Environmental
Modelling & Software 84 (2016) 290-303.
[16] Cátia M. Salgado, Susana M. Vieira, Luís F.
Mendonça, Stan Finkelstein, João M.C.
Sousa, Ensemble fuzzy models in
personalized medicine: Application to
vasopressors administration, Engineering
Applications of Artificial Intelligence
(2015).
[17] B. Seijo-Pardo, I. Porto-D’iaz, V. Bol’on-
Canedo, A. Alonso-Betanzos, Ensemble
Feature Selection: Homogeneous and
Heterogeneous Approaches, Knowledge-
Based Systems (2016).
[18] J.Prathyusha, G.Sandhya, V.Krishna Reddy,”
An Improvised Partition-Based Workflow
Scheduling Algorithm”, International
Innovative Research Journal of Engineering
and Technology, vol 02, no 04,pp.120-
123,2017.

Cite this article as :

L. S. Rohith Anand, B. Shannmuka, R. Uday


Chowdary, K. Satya Sai Krishna, "Liver Cancer
Detection", International Journal of Scientific
Research in Computer Science, Engineering and
Information Technology (IJSRCSEIT), ISSN : 2456-
3307, Volume 5 Issue 1, pp. , January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT183818
Journal URL : http://ijsrcseit.com/CSEIT183818

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


228
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195150

Emotion Detection using Facial expressions with Transfer


Learning
Sairama Geetanand, Shaik Nazeer Basha, Sajja Nageswara Rao, Rayudu Srinivas
CSE, Vasireddy Venkatadri Institute of Technology, Guntur, Andhra Pradesh, India

ABSTRACT

The use of machines to perform different tasks is constantly increasing in society. Providing machines with
perception can lead them to perform a great variety of tasks. Machine perception requires that machines understand
about their environment. Recognizing facial emotions will thus help in this regard. We use the TensorFlow library
and the Inception model and apply transfer-learning for our dataset to retrain the model. We then identify the facial
emotions: happiness, sadness, anger, surprise, disgust and neutral.
Keywords: Deep learning, TensorFlow, Inception, Transfer Learning, Convolutional neural networks.

I. INTRODUCTION recognition is particularly difficult for two reasons: (1)


There does not exist a large database of training images
Facial expressions play a very important role in human and (2) classifying emotions can be difficult depending
communication. The human face is the richest source of on whether the input image is static or in a transition
emotions. As society continues to make more use of frame into a facial expression. The latter issue is
human-machine interactions, it is important for particularly difficult for real-time detection where facial
machines to be able to interpret facial expressions in expressions vary dynamically.
order to improve their authenticity or to make them less
machine and more human. Our brains make vision In the last few years the field of machine learning has
seem easy. It doesn't take any effort for humans to tell made tremendous progress on addressing these difficult
apart a lion and a tiger, read an article, or recognizing a problems. In particular, the discovery of the model
human's face. But these are actually hard problems to called a deep convolutional neural network can achieve
solve with a computer: they only seem easy because our reasonable performance on hard visual recognition
brains are incredibly good at understanding images. tasks matching or exceeding human performance in
Emotion can be recognized through a variety of means some domains. It is now being used by popular tech-
such as voice intonation, body language, and more giants, Facebook being one of the leading users.
complex methods such electroencephalography (EEG). Researchers have demonstrated steady progress in
However, the easier, more practical method is to computer vision by validating their work against
examine facial expressions. There are seven types of ImageNet an academic benchmark for computer vision.
human emotions shown to be universally recognizable
across different cultures: anger, disgust, fear, happiness, II. METHODS AND MATERIAL
sadness, surprise, contempt. Interestingly, even for
complex expressions where a mixture of emotions A. Dataset
could be used as descriptors, cross-cultural agreement is
still observed. Therefore, a utility that detects emotion We scraped image urls from google images, using java
from facial expressions would be widely applicable. script, downloaded the images from the generated urls
Such advancement could bring applications in medicine, using python. The images are then filtered manually
marketing and entertainment. The task of emotion

CSEIT195150 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 229-234 ]
229
Sairama Geetanand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 229-234

and the ones that don‟t suit the emotion are removed. part of ImageNet Large Sale Visual Recognition
The current dataset of images we used contain 2192 challenge 2014. It can classify almost every day-to-day
images for the six emotions : Happy, Sad, Angry, objects. The architecture of the Inception is briefly
Surprise, Disgust, Neutral. described in . Inception consists of 22 layers. The
penultimate layer is called as “Bottlenecks”. The final
B. TensorFlow layer is called as softmax layer. This is the layer that
can be retrained to classify the required image group.
Tensor Flow (TF) [2] is an open source software library We retrained it to classify 6 emotions: neutral, happy,
for machine learning written in Python and C++. The sad, surprise, disgust and anger. There are 3 metrics.
main reason behind it is that TF was developed by Training accuracy describes the accuracy of the
Google Brain Team. Google has already been using TF inception model in classifying the images on which it
to improve some tasks on several products. These tasks was trained. Validation accuracy describes the accuracy
include speech recognition in Google Now, search of the model in classifying the images that are not part
features in Google Photos, and the smart reply feature of the training dataset. Cross entropy is the loss
in Inbox by Gmail. Some design decision in TF have function. It describes the error rate.
lead to this framework to be early adopted by a big
community. One of them is the ease of going from
prototype to production. There is no need to compile or
to modify the code to use it on a product. Then, the Fig.1: The 3 metrics: Train accuracy, Class entropy,
framework is not only thought as a research tool, but as Validation accuracy
a production one. Another main design aspect is that
there is no need to use different API when working on D. Implementation
CPU or GPU. Moreover, the computations can be
deployed over desktops, servers and mobile devices. A Once we are sure that inception is correctly installed
key component of the library is the data flow graph. and is working correctly we retrain the model for our
The sense of expressing mathematical computations dataset like so. Modern object recognition models have
with nodes and edges is a TF trademark. Nodes are millions of parameters and can take weeks to fully train.
usually the mathematical operations, while edges define Transfer learning is a technique that shortcuts a lot of
the input / output association between nodes. The this work by taking a fully-trained model for a set of
information travels around the graph as a tensor, a categories like ImageNet, and retrains from the existing
multidimensional array. Finally, the nodes are allocated weights for new classes. We retrained the final layer
on devices where they are executed asynchronously or from scratch, while leaving all the others untouched.
in parallel when all the resources are ready. How is it done?

C. Inception We evaluate whether features extracted from the


activation of a deep convolutional network trained in a
Inception is a pre-trained deep neural network, fully supervised fashion on a large, fixed set of object
developed to classify images. It took around 2 weeks to recognition tasks can be re-purposed to novel generic
train the inception model using 8 Tesla K40 GPUs and tasks. Our generic tasks may differ significantly from
it costed around $30,000. It considers around 25 million the originally trained tasks and there may be
parameters and does 5 billion add-multiply operations insufficient labelled or unlabeled data to conventionally
to classify a single image. Given today‟s computation train or adapt a deep architecture to the new tasks. We
power, it can be done in fraction of a second. Inception investigate and visualize the semantic clustering of
takes images as its input. It can process only JPEG deep convolutional features with respect to a variety of
format images. The recommended resolution is such tasks, including scene recognition, domain
299*299. If the image is of higher resolution, it will be adaptation, and fine-grained recognition challenges.
compressed automatically. It produces a private class of Before we start any training, we need a set of images to
array of the image as its output. It was developed as a teach the network about the new classes you want to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


231
Sairama Geetanand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 229-234

recognize. Here we use a dataset of emotions gathered


from a variety of sources. Once you have the images,
from the root of your TensorFlow source directory you
can begin the training process.

We load the pre-trained Inception v3 model, remove the


old top layer, and train a new one on the emotion
Fig.3: Bottleneck values generated for an image
photos we've downloaded. None of the emotions were
in the original ImageNet classes the full network was
Once the bottlenecks are complete, the actual training
trained on. The magic of transfer learning is that lower
of the top layer of the network begins. We'll see a series
layers that have been trained to distinguish between
of step outputs, each one showing training accuracy,
some objects can be reused for many recognition tasks
validation accuracy, and the cross entropy. The training
without any alteration. The script can take thirty
accuracy shows what percent of the images used in the
minutes or more to complete, depending on the speed of
current training batch were labelled with the correct
the machine. The first phase analyzes all the images on
class. The validation accuracy is the precision on a
disk and calculates the bottleneck values for each of
randomly-selected group of images from a different set.
them. 'Bottleneck' is an informal term we often use for
The key difference is that the training accuracy is based
the layer just before the final output layer that actually
on images that the network has been able to learn from
does the classification. This penultimate layer has been
so the network can overfit to the noise in the training
trained to output a set of values that's good enough for
data. A true measure of the performance of the network
the classifier to use to distinguish between all the
is to measure its performance on a data set not
classes it's been asked to recognize. That means it has
contained in the training data -- this is measured by the
to be a meaningful and compact summary of the images,
validation accuracy. If the train accuracy is high but the
since it has to contain enough information for the
validation accuracy remains low, that means the
classifier to make a good choice in a very small set of
network is overfitting and memorizing particular
values. The reason our final layer retraining can work
features in the training images that aren't helpful more
on new classes is that it turns out the kind of
generally. Cross entropy is a loss function which gives
information needed to distinguish between all the 1,000
a glimpse into how well the learning process is
classes in ImageNet is often also useful to distinguish
progressing. The training's objective is to make the loss
between new kinds of objects. Because every image is
as small as possible, so you can tell if the learning is
reused multiple times during training and calculating
working by keeping an eye on whether the loss keeps
each bottleneck takes a significant amount of time, it
trending downwards, ignoring the short-term noise.
speeds things up to cache these Bottleneck values on
disk so they don't have to be repeatedly recalculated
By default, it will run 4,000 training steps. Each step
and if you rerun the script they'll be reused so you don't
chooses ten images at random from the training set,
have to wait for this part again.
finds their bottlenecks from the cache, and feeds them
into the final layer to get predictions. Those predictions
are then compared against the actual labels to update
the final layer's weights through the back-propagation
process. As the process continues you should see the
reported accuracy improve, and after all the steps are
done, a final test accuracy evaluation is run on a set of
images kept separate from the training and validation
pictures. This test evaluation is the best estimate of how
the trained model will perform on the classification task.
Fig.2 : TensorFlow creating bottleneck values for each You should see an accuracy value of between 90% and
image 95%, though the exact value will vary from run to run

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


232
Sairama Geetanand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 229-234

since there is randomness in the training process. This


number is based on the percent of the images in the test
set that are given the correct label after the model is
fully trained.

Tensor Board summaries make it easier to understand,


debug, and optimize the retraining. We can visualize
the graph and statistics, such as how the weights or
accuracy varied during training. The Inception v3
network with a final layer retrained to our categories to
get the output graph and a text file containing the labels
–output labels.

III. RESULTS AND DISCUSSION


Fig.5: A line graph depicting the average percentage
values, obtained by testing various images of the six
After the training process, we provide the retrained
emotions.
model with the image we wish to classify. Note that
the retrained model can classify images only with the
The graph values are taken by testing the output label
classes we provided initially in the retraining
with various images, some taken from the dataset and
inception part, so if the image is for example, an
some outside of it. Each emotion is tested using 15
image of a tall building the result would be the best
images and the average of all these values is then
possible match among the classes. The system can
converted to percentage, which are given as inputs to
identify only the images it is trained for just like
the graph.
humans, seeing something we have never seen before
we shall not be able to identify it. It will classify the
IV.CONCLUSION
images in the weights of percentage, for example if an
image is happy it will identify say 0.9025 to say it
In this work, we detect and recognize emotions in
static human images using TensorFlow and Inception
model. The future enhancement can be an action that
Fig.4: Test image classification values. is done upon recognition of the emotions. If we get a
sad emotion, we can have the system play a song or
The above image shows the classification value of an tell a joke or send his/her best friend a message. This
image, which depicts the emotion „sad‟, and its value can be the next step of AI where the system can
generated is 0.931057 which approximates to 93%. understand, comprehend the user’s feelings and
emotions and react accordingly. This bridges the gap
between machines and humans.

We can also have an interactive keyboard where the


users can just use the app and the app will then
identify the emotion and convert that emotion to the
emoji of choice.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


233
Sairama Geetanand et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 229-234

V. REFERENCES

[1]. Raghav Puri, Archit Gupta, Manas Sikri (2018).


Emotion Detection using Image Processing in
Python. 5th International Conference on
“Computing for Sustainable Global
Development (IndiaCom)” IEEE conference ID
: 42835.
[2]. https://codelabs.developers.google.com/codelab
s/tensorflow-for-poets
[3]. Hong-Wei Ng, Viet Dung Nguyen, Vassilios
Vonikakis, Stefan Winkler(2015). Deep
Learning for Emotion Recognition on Small
Datasets Using Transfer Learning.
[4]. Alexandr Rassadin, Alexey Gruzdev, Andrey
Savchenko(2017). Group-level Emotion
Recognition Using Transfer Learning from Face
Identification.
[5]. https://www.tensorflow.org/hub/tutorials/imag
e_retraining
[6]. https://www.tensorflow.org/install/source
[7]. https://docs.docker.com/toolbox

Cite this article as :

Sairama Geetanand, Shaik Nazeer Basha, Sajja


Nageswara Rao, Rayudu Srinivas , "Emotion
Detection using Facial expressions with Transfer
Learning ", International Journal of Scientific
Research in Computer Science, Engineering and
Information Technology (IJSRCSEIT), ISSN : 2456-
3307, Volume 5 Issue 1, pp. 229-234, January-
February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195150
Journal URL : http://ijsrcseit.com/CSEIT195150

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


234
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195167

A Review on Various Algorithms for Student Performance


Prediction
Mayuri S. Dongre1, Neha S.Vidya1, Vanita D. Telrandhe1, Shweta R. Chaudhari1, Anup A. Umrikar1,
Prof. Prachiti V. Adghulkar2
1 BE Students, Department of computer Science & Engineering, Suryodaya collage of Engineering &
Technology, Nagpur, Maharashtra, India
2Assistant Professor, Department of computer Science & Engineering, Suryodaya collage of Engineering &
Technology, Nagpur, Maharashtra, India
ABSTRACT

Data in educational institutions are growing progressively along these lines there is a need of progress this
tremendous data into helpful data and information utilizing data mining. Educational data mining is the zone of
science where diverse techniques are being produced for looking and investigating data and this will be
valuable for better comprehension of understudies and the settings they learned. Classification of data objects in
view of a predefined learning of the articles is a data mining and information administration procedure utilized
as a part of collection comparable data questions together. Decision Tree is a valuable and well known
classification method that inductively takes in a model from a given arrangement of data. One explanation
behind its prominence comes from the accessibility of existing calculations that can be utilized to assemble
decision trees. In this paper we will survey the different ordinarily utilized decision tree calculations which are
utilized for classification. We will likewise contemplating how these decision tree calculations are appropriate
and valuable for educational data mining and which one is ideal.
Keywords : Educational data mining (EDM), Classification, Dropout Prediction, Selection Failure

I. INTRODUCTION The data in any given educational association is


growing rapidly. There is a need to change this data
As of late the scientist concentrates on the new into helpful data and learning; henceforth we make
region of research, which is EDM (Educational Data utilization of data mining.
mining). EDM is the information disclosure in
database strategy or the data mining in instruction. Distinguish and find valuable data covered up in
The specialist concentrates on the improvement of substantial databases is a troublesome errand. An
technique to better comprehend understudies and the exceptionally making an assurance to answer for
settings in which they learn. There are great cases of accomplish this goal is the utilization of learning
how to apply EDM systems to make models that revelation in databases techniques or data mining in
foresee dropping out and understudy disappointment. training, called Educational Data Mining, EDM [15].
These works have indicated promising outcomes Educational data mining techniques drawn from an
regarding those sociological, monetary, or assortment of writings, including data mining and
educational qualities that might be more pertinent in machine learning, psychometrics and different ranges
the forecast of low scholarly execution.

CSEIT195167 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 235-239 ]
235
Mayuri S. Dongre et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 235-239

of measurements, data perception, and computational amid this stage the tree is recursively parcelled till
displaying. every one of the data things have a place with a
similar class name. Tree pruning is done is a base up.
As we examined over the new territory of center for It is utilized to enhance the forecast and classification
the specialists is EDM. In introduce the vast majority exactness of the calculation by limiting over-fitting.
of the procedures was produced for the educational Over-fitting in decision tree calculation brings about
data mining, which foresee the disappointment and misclassification mistake. Tree pruning is less
dropout understudies. Be that as it may, a large entrusting contrasted with the tree development
portion of the examination on the use of EDM to stage as the preparation data set is filtered just once.
determine the issues of understudy disappointment In this examination we will audit Decision tree
and drop-outs has been connected principally to the calculations executed, recognize the ordinarily
particular instance of advanced education and all the utilized calculations.
more particularly to on the web or separation
instruction. Less almost no data about particular II. RELATED WORK
research on basic and optional instruction has been Decision tree is a vital strategy for both acceptance
found, and what has been discovered uses just research and data mining, which is basically utilized
measurable strategies, not DM methods. Along these for display classification and expectation. ID3
lines, there is a need to proposed Educational data calculation is the most broadly utilized calculation in
mining system that is attainable for rudimentary and the decision tree up until this point. Through
optional instruction. outlining on the fundamental thoughts of decision
tree in data mining. ID3 calculation is a forerunner of
Classification can be depicted as a regulated learning C4.5 calculation and it was produced by a data
calculation in the machine learning process. It mining software engineering scientist Ross Quinlan
relegates class marks to data objects in light of earlier in 1983.It is utilized to build a decision tree by testing
information of class which the data records have a every hub's trait of tree in top-down way. ID
place. In classification a given arrangement of data calculation performs property choice component
records is partitioned into preparing and tests data utilizing Entropy and Information Gain idea. A
sets. The preparation data set is utilized as a part of decision tree is an essential method for data mining
building the classification display, while the test data and inductive realizing, which is generally used to
record is utilized as a part of approving the model. frame classifiers and expectation models [2].
The model is then used to order and foresee new
arrangement of data records that is not the same as C4.5 is a standout amongst the most exemplary
both the preparation and test data sets. Decision tree classification calculations on data mining. Decision
calculation is a data mining enlistment methods that Trees a Decision Tree is a helpful and prevalent
recursively segments a data set of records utilizing classification procedure that inductively takes in a
profundity first voracious approach or expansiveness model from a given arrangement of data. One
initially approach until the point when every one of explanation behind its fame originates from the
the data things have a place with a specific class. The accessibility of existing calculations that can be
tree structure is utilized as a part of arranging obscure utilized to manufacture decision trees [3].
data records. Decision tree classification system is
performed in two stages: tree building and tree CART is Classification and relapse trees. It was
pruning. Tree building is done in top-down way. It is presented by Breiman in 1984. It fabricates the two

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


236
Mayuri S. Dongre et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 235-239

classifications and regressions trees. The classification classification and for recognition of connection
tree development via CART. It depends on paired between factors [16].
part of the characteristics. It is additionally in view of
Hunt's model of decision tree development. It The present decision tree calculations likewise
additionally can be executed serially by Breiman et al. contrast in their capacity to deal with various sorts of
It utilizes gini record part measure for choosing the data, and by a wide margin the most exceptional
part quality. Pruning is finished by utilizing a calculation of the gathering is C4.5. C4.5 depends on
segment of the preparation data set in truck. This Quinlan's prior work with ID3 and is equipped for
approach utilizes both numeric and clear cut taking care of ceaseless data, data with missing
properties for building the decision tree and has in- esteems, and even has worked in ventures for
constructed highlights that arrangement with missing rearranging the resultant decision tree. Another
qualities [11]. purpose for the ubiquity of decision trees is that they
are regularly interpretable by human analyzers. The
The fundamental distinction between CART, ID3 and structure of a decision tree gives thinking to each
C4.5 is the manner by which the parceling of data is statement of class esteem, and it is trusted that this
performed. Truck utilizes the Gini list to choose the thinking is anything but difficult to understand.
part quality, though ID3 and C4.5 utilize an
estimation of data pick up and the data pick up The classifier is tried first to group inconspicuous
proportion. data and for this reason coming about decision tree is
utilized. C4.5 calculation takes after the standards of
SLIQ is the Supervised Learning in Ques approach. It ID3 calculation. Also C5.0 calculation takes after the
was presented by Mehta et al in 1996. It is a quick, tenets of calculation that is the vast decision tree can
adaptable decision tree calculation that can be see as an arrangement of guidelines which is
executed in serial and parallel example. It does not straightforward. C5.0 calculation gives the recognize
depend on Hunt's calculation for decision tree on clamor and missing data. Issue of over fitting and
classification. It segments a preparation data set mistake pruning is tackled by the C5 calculation. In
recursively. It utilizes expansiveness first ravenous classification procedure the C5 classifier can envision
procedure that is coordinated with pre-sorting which properties are important and which are not
method amid the tree building stage [17]. applicable in classification [5]

Dash is Scalable Parallelizable Induction of decision


Tree calculation. It was presented by Shafer et al. It is
a quick and adaptable decision tree classifier. Like
SLIQ it utilizes one time kind of the data things. It
has no limitation on the info data estimate [18].

CHAID is a kind of decision tree method. It depends


on balanced hugeness testing. This procedure was
produced in South Africa. It is a comparable form to
relapse examination. This adaptation of CHAID being
initially known as XAID. It is valuable for

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


237
Mayuri S. Dongre et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 235-239

[2]. IU Qin, “Data Mining Method Based on


Computer Forensics-based ID3 Algorithm”,
978-1-4244-5265-1/10/$26.00©2010IEEE
[3]. Rong Cao, Lizhen Xu, “Improved C4.5
Algorithm for the Analysis of Sales”, Sixth Web
Information Systems and Applications
Conference, 978-0-7695-3874-7/09 $25.00 DOI
10.1109/WISA36, © 2009 IEEE
[4]. Gaganjot Kaur , Amit Chhabra , “Improved J48
Classification Algorithm for the Prediction of
Diabetes” , International Journal of Computer
Applications (0975 – 8887) Volume 98 – No.22,
July 2014.
[5]. Tomasz Bujlow, Tahir Riaz, Jens Myrup
Pedersen, “A method for classification of
Table 1. Suervy Table network traffic based on C5.0 Machine
Learning Algorithm”, Workshop on Computing,
III. CONCLUSION Networking and Communications, 2012 IEEE.
[6]. Raisul Islam Rashu, Naheena Haq, Rashedur M
This paper is a review of the best in class regarding Rahman, “Data Mining Approaches to Predict
EDM and reviews the most applicable work around Final Grade by Overcoming Class Imbalance
there to date. It would be exceptionally hard to Problem”, 17th International Conference on
physically experience the tremendous arrangement of Computer and Information Technology (ICCIT)
scholarly records to distinguish the understudy 2014.
patterns and conduct and the example in which they [7]. Hina Gulati, “Predictive Analytics Using Data
learn. Rather, if client makes utilization of data Mining Technique”, 978-9-3805-4416-
mining strategies on the expansive measure of 8/15/$31.00c 2015 IEEE.
scholastic record, he/she can without much of a [8]. Rutvija Pandya , Jayati Pandya , “C5.0
stretch gathering the understudies, distinguish Algorithm to Improved Decision Tree with
shrouded designs about their learning styles, discover Feature Selection and Reduced Error Pruning ”,
unfortunate understudy conduct and perform International Journal of Computer Applications
understudy profiling. In this way, data mining can (0975 – 8887) Volume 117 – No. 16, May 2015.
unquestionably be a vital apparatus and part of [9]. Alana M. de, Morais and Joseana M. F. R.
mechanically progressed educational systems. Araújo, Evandro B. Costa, “Monitoring Student
IV. REFERENCES Performance Using Data Clustering and
[1]. Carlos Márquez-Vera, Cristóbal Romero Predictive Modelling”, 978-1-4799-3922-
Morales, and Sebastián Ventura Soto, 0/14/$31.00 ©2014 IEEE.
“Predicting School Failure and Dropout by [10]. Kin Fun Li, David Rusk and Fred Song,
Using Data Mining Techniques”, Ieee Journal “Predicting Student Academic Performance”,
Of Latin-American Learning Technologies, Vol. 2013 Seventh International Conference on
8, No. 1, February 2013 IEEE. Complex, Intelligent, and Software Intensive
Systems.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


238
Mayuri S. Dongre et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 235-239

[11]. Mishra, T., Kumar, D. Gupta, S.,” Mining


Students' Data for Prediction Performance”,
Advanced Computing & Communication
Technologies (ACCT), 2014 Fourth
International Conference on 8-9 Feb. 2014.
[12]. B.K.Bhardwaj and S.Paul, “Mining Educational
Data to Analyze Students Performance”,
International Journal Advanced Computer
Science and application Vol. 2 No. 6, 2011.
[13]. Yohannes Kurniawan, Erwin Halim, “Use Data
Warehouse and Data Mining to Predict Student
Academic performance in Schools: A Case
Study (Perspective Application and Benefits)”.
[14]. B.K.Bhardwaj and S.Paul, “Mining Educational
Data to Analyze Students Performance”,
International Journal Advanced Computer
Science and application Vol. 2 No. 6, 2011.
[15]. Lewis, R.J. (200). An Introduction to
Classification and Regression Tree (CART)
Analysis. 2000 Annual Meeting of the Society
for Academic Emergency Medicine, Francisco,
California.
[16]. Mehta, M., Agrawal, R., and Rissanen, J. (1996).
SLIQ: A fast scalable classifier for data mining.
In EDBT 96, Avignon, France
[17]. Shafer, J., Agrawal, R., and Mehta, M. (1996).
Sprint: A scalable parallel classifier for data
mining. Proceedings of the 22nd international
conference on very large data base. Mumbai
(Bombay), India.

Cite this article as :

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


239
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195166

E-Toll System Using RFID


Ankita Bambole , Shrishti Thakur1, Suvidha Ramteke1, Sneha Borkar1, Hitesh Kasturi1, Prof. M. M. Baig1
1

1Department of Computer Science and Engineering, JD College of Engineering and Management, Nagpur,
Maharashtra, India
2Assistant Professor, Department of Computer Science and Engineering, JD College of Engineering and

Management, Nagpur, Maharashtra, India

ABSTRACT

ETS (E-Toll System) is an electronic Toll Tax collection system by which we can collect the toll charges using
RFID card detection and scanning.These RFID Cards are related to the respective vehicle holders.This project
hence, reducing common problems like traffic congestion, Pollution caused due to vehicles, Time dilemma,
requirement of labours etc. Makes it effective to collect the required amount of charge from the driver. There
will be an RFID tag implemented on every vehicle. Whenever vehicle passes through the toll plaza RFID
reader with the help of sensors will detect tag and the required amount will be deducted. Driver don’t need to
stop to pay the money ,and this RFID card is used to track & bill the vehicle owner through a payment gateway.
Keywords : Electronic Toll Collection, RFID, XAMPP, Vehicle, Tag.

I. INTRODUCTION Sanral:-E-toll (in South Africa) consists of the


electronic toll collection (ETC) processes employed
The world is changing furiously, the role of by South Africa's roads agency Sanral on selected toll
automation in our day to day life is increasing at a roads or toll lanes, subject to the Sanral Act of 1998.
very fast rate, hence the consumption of fuel and use
of road is increasing proportionally. People charge of FASTag:-FASTag is an automatic electronic toll
using roads as nothing in this universe is free of Cost. collection system operated by National Highway
This leads to the motive of our project i.e. “The Authority of India[1] (NHAI). It is also based on
Automation”.We have chosen this topic because Radio Frequency Identification (RFID) technology
digitizing current toll plaza will overcome for making toll payments from savings account linked
disadvantages like time delay, traffic congestion etc. to it
The charge form of manual and semi-manual will not
meet demand of the charging management system E-ZPass:-The E-ZPass Group is comprised of toll
and may lead to many vehicle blocked at entrances entities stretching across 17 states that operate the
and exits causing huge economic losses when it extremely successful E-ZPass electronic toll
reached a certain threshold enhancing the collection program.
convenience and safety of the travelers.
The current status of RFID-based ETC system
worldwide exampled and described as follows:

CSEIT195166 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 240-243 ]
240
Ankita Bambole et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 240-243

II. THE COMPOSITION processing will be done. Transaction will begin


depending upon the balance in the account of the
Our Toll Collection system is madeup of following drivers card the amount will be deducted as the toll
modules:- tax if the vehicle has to encounter another toll plaza
through the journey she can request for deduction of
• Hardware part: This is a hardware-cum-sensor amount wrt the upcoming toll also.All these tolls are
system which manages vehicle detection and is connected using the centralized server using LAN.If
installed at the toll booth. The hardware is based there is no sufficient amount of money present the
on Arduino boards development platform. user can pay manually or other transactions means.
• Server: This is the backend server hosted in The software updates the data of the transaction after
XAMPP which is responsible for authenticating, the process successfully terminated.
identifying and billing a vehicle based on the
unique RFID tag. Thus it manages the complete Following is the algorithm associated with the
toll collection process. project:-
• Database: Consists of user and administrator data
and information regarding the total transaction Step 1: Start
and input output of the vehicles in and through Step 2: Sensors 0.5 km before the toll plaza detects
the toll plaza. transport.
• User website: It is a web site frontend we have Step 3: Scanner and Camera will start processing
created in PHP for users and toll people to according to step 2.
register new drivers,Tolls,Vehicles and updating Step 4: Vehicle slow down for some seconds at the
information regarding the vehicles. same time High quality scanner scans QR-Code
Step 5: if Vehicle= Truck or Other transport then
III. DESIGN AND IMPLEMENTATION return the amount of tax to be deducted.
Vehicle = special Transport then
When a person buys a new vehicle she’ll need to Return(null);
register herself and her vehicle at the toll office or Step 6: At the same time the high quality Camera
RTO office. The RTO officer will allot a unique ID to clicks picture of the vehicle.
the RFID card the user along with No. Plate which
will contain a unique ID related to the Card only. At If{
this time the officer will also create an account for Traffic rules are followed then
the user although user can herself make her account return (null);
using the website by signing up. After it has else
registered the vehicle a password and user id will be return(penalty);
generated and once it is generated the user can Step 7: if the Transaction is completed
recharge his/hers card any time. successfully then the Tax will be delivered on their
address.
Each time the drivers vehicle approaches the toll Step 8:Message will be sent on owners phone
plaza the IR sensor implemented on the Step 9:END
plaza(Hardware)will sense the vehicle approaching
and sends a ready signal to the RFID reader and
Scanner the scanner will scan the Card and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


241
Ankita Bambole et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 240-243

IV. HARDWARE SETUP (a)RS232 compatible output and (b)TTL compatible


output. A TTL compatible output pin is connected
At the toll booth, Arduino board hardware are directly to Arduino.
situated which manages the toll total collection
processes .We have used Arduino UNO which is a
microcontroller board ATmega328P consisting of 14
digital input/output pins ,6 analog input ,a quartz
crystal, USB cable connection , a power supply jack,
ICSP header and a reset button.
We have simply connected it to the computer using
USB cable.

Fig 3. RFID with Arduino Uno Rev3

V. DATABASE

Database is the Heart of our project. We have used


MY-SQL for database query generation and updating
the entries. We obtained a database for user, admin
Fig1. Arduino Uno Rev3,Code: A000066 and transaction using XAMPP control panel .There
are three Data information’s stored in the database as
To track the vehicles, Arduino UNO is interfaced follows:-
with the RFID reader and scanner for getting the 1.Admin’s Database:-Administrator database consists
information regarding the vehicle. of all the details of central database and all tolls under
Construction.I.e. RTO’s Database.
Both the RFID reader and the tags comes with the 2.Central Database:- Central database consist records
coil in them power supply is given and the data is of all toll plaza in work. This is managed by the RTO
fetched We collect the read data. administrator.The Customer registers themselves into
this account system. This account information about
the users is stored into The RTO database. When the
registered customer passes through the particular toll
plazas then, automatically data will be updated.
3.Integrated database:-Integrated database is
disconnected to the central database This database
will update automatically. This database includes all
Registered vehicles and the details of vehicle such as
Fig 2. Interfacing Arduino with RFID vehicle owner, vehicle number, license number,
account ID, account balance,current charges, etc.
.Power supply requirement of RFID Readers varies
for different products. Reader we have used is of 12
Volts. There are 2 outputs from RFID Reader.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


242
Ankita Bambole et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 240-243

VI.CONCLUSION Cite this article as :

In this report, the conclusion that came out is that, to Ankita Bambole, Shrishti Thakur, Suvidha Ramteke,
implement modern system of toll collection new Sneha Borkar, Hitesh Kasturi, Prof. M. M. Baig, "E-
technique RFID came into light. The RFID safety is Toll System Using RFID", International Journal of
key important o this project. Scientific Research in Computer Science, Engineering
This project is reliable and easy way to pay toll as and Information Technology (IJSRCSEIT), ISSN :
compared to manual one. 2456-3307, Volume 5 Issue 1, pp. 240-243, January-
February 2019.
Project is all about automated toll system which will Journal URL : http://ijsrcseit.com/CSEIT195166
overcome disadvantages of current toll system. Our
project helps driver to pass through toll without
stopping his vehicle to pay the toll tax. RFID tag on
each vehicle will be detected by RFID reader and
amount will be deducted. Open Road Tolling
System[2] and Automated Toll system[3] these are
two previously implemented projects similar to our
project. We have studied them and we are trying to
overcome disadvantages of these two. The project if
implemented by government then toll system will be
fully automated and one who is passing through toll
don’t need to wait for a long time this will also
reduce pollution, traffic and accidents.

VII. REFERENCES

[1]. Electronic Toll Collection System in India.


Operated by the National Highway Authority
of India
https://en.wikipedia.org/wiki/National_High
way_Authority_of_India.
[2]. OPEN ROAD TOLLING - IMPROVED
SAFETY, LESS CONGESTION RSS FIRST
PUBLISHED IN ITS INTERNATIONAL
JANUARY FEBRUARY 2010 AS ORT: THE
SILVER BULLET FOR SAFETY
[3]. AUTOMATIC TOLL COLLECTION
ITERNATIONAL JOURNAL OF
COMPUTER SCIENCE AND MOBILE
COMPUTING, VOL.5 ISSUE.8, AUGUST-
2016

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


243
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/ CSEIT195165

A Review on Indoor Navigation Systems and Technologies for


Guiding the Users
Meenakshi D. Dhobale1, Pooja S. Verma1, Pushpa R. Joshi1, Jyotsna K. Tekade1, Swati Raut2
1BE Students, Department of Computer Science and Engineering, JD College of Engineering and Management,
Nagpur, Maharashtra, India
2Assistant Professor, Department of Computer Science and Engineering, JD College of Engineering and
Management, Nagpur, Maharashtra, India

ABSTRACT

GPS is in actuality the main answer for outdoor positioning, no unmistakable arrangement has so far risen for
indoor positioning notwithstanding the business hugeness. Along these lines, the principal point is to grow
ongoing indoor following framework and utilizing cell phone sensors to give dependable and precise indoor
localization. Indoor navigation is a vital empowering innovation for application, for example, finding a
gathering room in a place of business, security exit amid a crisis or focused on the retail commercial in a
shopping center. Utilizing this thought, an application can be made and by methods for that application, we are
handling indoor navigation utilizing continuous navigation. The android application gives mindfulness,
straightforwardness to discover your direction, commercial and effortlessness. To give solid indoor navigation
to business structures like shopping centers, enterprises, universities, hospitals, inns. In this paper, we give an
order and arrangement of the current indoor positioning systems and distinguish some conceivable zones of
improvements.
Keywords : Indoor Navigation, Pedestrian Navigation, GPS, Object Tracking, Location Based Services

I. INTRODUCTION and Non-Line-of-Sight (NLoS) conditions, nearness


of moving individuals that adjust the indoor
Recently, the utilization of indoor localization proliferation channel, more noteworthy thickness of
methods has turned out to be progressively critical in obstructions that reason a high weakening and flag
an extensive number of uses and settings, for example, dissipating, request of a higher exactness and
human services, homecare, checking, following, and precision.
so forth. In outside localization settings, the most
notable and broadly spread innovation is the Global In the decision of the best innovation to plan an
Positioning System (GPS). It can ensure fantastic indoor localization framework, an extensive number
execution in outside situations yet not work of parameters ought to be considered (e.g., cost,
legitimately in indoor conditions because of poor precision, vigor, versatility, inclusion, and so on.).
inclusion of satellite flag. Additionally, acquiring Clearly, a solitary arrangement that works fine for
position data in indoor situations is especially testing any situation does not exist. At that point, it is vital to
a direct result of a few reasons: mistakes by multipath circumspect the execution parameters all things

CSEIT195165 | Received : 01 Feb 2019 | Accepted : 08 Feb 2019 | January-February -2019 [ 5 (1) : 244-250 ]
244
Meenakshi D. Dhobale et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 244-250

considered and coordinate them with the client II. LITERATURE REVIEW
prerequisites, which must be broke down and
portrayed decisively for every application. In Scientists and organizations have made considerable
addition, the estimations of the execution parameters commitments to the theme of Indoor Positioning
are not univocally definable since they thus rely systems and conditions. Indoor Positioning zone is
upon different factors and conditions. Consequently, brimming with valuable commitments. In this
it is important to locate the correct exchange off segment, we talk about the diverse indoor conditions
among execution parameters, client necessities, and and the best in the class of indoor positioning systems.
natural conditions so as to plan a modified
arrangement. A. Fixed Indoor Positioning

In an outside situation, the Global Positioning System Normal design of Fixed IPS systems is having settled
(GPS) works productively in positioning and focusing number of Base Stations (BS), which is the
on various sorts of substances. It has been utilized in fundamental PC that gets every one of the signs from
numerous outdoor applications for limiting the diverse hubs and ascertains the directions of the
individuals, autos, just as different items. Be that as it objective article utilizing the parameters sent from
may, GPS comes up short on a similar dimension of the sensor hubs. All BS are required to be introduced
effectiveness when utilized inside indoor situations. in settled areas inside the building. In addition, a
This issue is because of the presence of hindrances Wireless Sensor arranges including the sensor hubs
that can debilitate the flag of the GPS (e.g., building are required. In this structure, the sensor hubs are in
engineering, dividers) where the presence of various charge of proliferating the flag got from a portable
gear can cause a commotion in the GPS transmitted station to the base station just as doing the basic
flag. count of the goal or time of transmissions. This
conduct relies upon the structure of the framework.
Pedersen [1] proposed a miniaturized scale The third principle segment of the settled indoor
positioning procedure that ought to be actualized positioning engineering is the cell phones or labels
inside the indoor condition so as to position and track conveyed by individuals or joined to target objects.
objects. He expressed that this procedure would fill in These labels give a novel recognizable proof to each
as a trade for the GPS positioning framework. item or individual. Along these lines, blunders will be
Furthermore, Fhelelboom [2] found that a remote restricted. Next, we list every remote innovation
neighbourhood (WLAN) can be utilized inside any with the systems, which are produced dependent on
indoor condition to position objects. this innovation. Elements of IPS are talked about in
[3].
In this paper, we audit the distinctive positioning
conditions, the diverse systems connected for every a. Infrared Positioning Systems
condition and the calculations utilized inside every
framework. We indicate two situations for These systems utilize infrared flags so as to transmit
positioning individuals and articles inside an indoor signals from sensor hubs to the BS. A standout
domain. Every one of these situations has its very amongst the most well-known infrared positioning
own difficulties, which specialists attempted in the systems is the dynamic identifications created by
past to moderate by proposing a few arrangements AT&T Cambridge [4]. In this framework, clients
throughout the most recent ten years. convey an ID card furnished with an infrared LED.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


245
Meenakshi D. Dhobale et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 244-250

The infrared LED sends a special code at regular and its directions. Cricket utilizes the Time
intervals. Moreover, there are infrared sensors Difference of Arrivals (TDOA) credit so as to figure
introduced on a roof and if the IR identification is the directions of the objective item. That is by
inside six meters, the sensor can peruse the code. The ascertaining the contrast between the Radio
BS gets the information from the IR sensors Frequency flag Time Arrival and the ultrasonic flag
occasionally. At last, the BS can assemble a guide of Time Arrival. The Difference of the season of the
each identification area utilizing the data recovered landing of the two signs is determined to utilize
from the sensors. Dynamic Badges have for the most conditions since the Radio Frequency is a lot quicker
part four directions, WITH, LOOK, NOTIFY, and than sound. At that point, the TDOA is equivalent to
HISTORY, every one of which gives an alternate the separation the flag went over the speed of the flag.
capacity. For instance, WITH demonstrates the The Cricket framework is professed to give a
identifications that exists in the sensor region, LOOK precision of up to 2 cm. Be that as it may, when
is utilized to search for an identification by a sensor, confronting NLOS blunder, the Cricket framework
NOTIFY is utilized to tell the BS when the does not give an exact computation of the facilitate.
identification is found and HISTORY demonstrates
the identifications positions over a specific timeframe. Popa [8] proposes joining Cricket with Inertial
Navigation sensors. This blend gives a superior
b. Ultrasonic Positioning Systems precision and relief of NLOS blunder as guaranteed
by Popa. That is on the grounds that Inertial
Ultrasonic reference points are utilized more Navigation sensors have a more extensive inclusion
frequently than infrared innovation. Ultrasonic and their flag isn't influenced by the NLOS mistake.
systems give increasingly exact positions to objects.
Ultrasonic based systems are more exact than Radio c. RF Positioning Systems
Frequency based systems, as we will find in the talk
area. Nevertheless, Ultrasonic systems need a settled The for the most part utilized remote innovation is
structure of the framework [5]. Radio Frequency (RF). This is because of the minimal
effort and the high scope of inclusion of the systems
Instances of the systems created dependent on created dependent on RF innovation. Some RF based
Ultrasonic innovation are the Active Bats [6] created systems that we will examine in this paper are
by AT&T Cambridge as well. This framework has a RADAR, Spot-On, LANDMARC, and UWB systems.
comparative structure to the dynamic identifications.
LANDMARC framework [9] depends on Radio
The Crickets framework [7] created by the Frequency signs and RFID labels. LANDMARC
Massachusetts Institute of Technology (MIT) is decreases the expense of utilizing RFID per users by
another outstanding ultrasonic based indoor lessening the quantity of per users and utilizing
positioning framework. The Cricket framework has reference labels. These reference labels have a well-
two kinds of hubs, signals, and audience members. known area and they transmit to the per users the
Reference points' areas are settled and they are area of the objective items. LANDMARC is a decent
connected to the roof while the audience members framework however its exactness is 1-2 meters.
are appended to the objective items and individuals.
Guides send intermittent data to the BS containing its Guang [10] gave a component to enhance the
ID, scope of inclusion or physical space related to it execution and exactness of LANDMARC. This

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


246
Meenakshi D. Dhobale et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 244-250

component works by lessening the quantity of


applicants of reference labels while ascertaining the B. Indoor Pedestrian Positioning
situation of an article. This diminishes the count
exertion and results in a quicker computation and a The person on foot positioning, as referenced prior in
superior precision. When testing LANDMARC the paper, happens when finding individuals who are
utilizing this component, Guang claims the precision conveying localization sensors while the building is
of LANDMARC was preferred utilizing this system not furnished with an indoor positioning framework.
over utilizing the conventional LANDMARC. In these kinds of systems, Inertial Navigation Systems
Guang's component gives a blunder rate. or dead retribution are for the most part utilized.
Dead Retribution is characterized as a navigation
Jiang [11] proposed a framework that gives open air procedure that begins with an outstanding area [13].
positioning utilizing GPS and UWB to give indoor At that point, includes the position changes in the
positioning. Jiang's framework contains PDA, UWB directions of the beginning stage. It likewise adds the
sensor organize and a Base Station. GPS programming progressions to the heading (course), speed or
is introduced on the PDA just as an interface for the separation. Besides, Pedestrian Dead Reckoning (PDR)
UWB sensor organize. At the point when the client is characterized as assessing the speed of development
moves outside the building the GPS application is and the heading or bearing of development. We will
actuated and when the client moves inside the demonstrate a portion of the intriguing commitment
building the system perceives the gadget and the to this sort of positioning.
UWB application is initiated to empower the client to
explore through the building and enables the Base A few methodologies in Indoor Pedestrian
Station to position this client. The exactness was Positioning use Particle Filters, for example, Bayesian
observed to be inside 10 meters which needs a ton of Filter and Kalman Filter. A Bayesian channel [14] is
upgrades later on. utilized to appraise the progression of the passer by at
a specific time when knowing the past strides of a
d. Optical Indoor Positioning similar walker at a number of times before it. Kalman
Filter [15] is a scientific model which is utilized to
Optical Indoor Positioning is another sort of settled precisely appraise the situation with the presence of
indoor positioning where we have a framework commotion.
introduced in the building and a camera conveyed by
the client. Beauregard [16] utilizes the Pedestrian Dead
Reckoning methodology so as to position. He utilizes
Tilch [12] proposed CLIPS (Camera and Laser based head protector mounted sensors as a novel
Indoor Positioning System). This framework methodology for person on foot positioning. That is
consolidates the two advances to position objects on the grounds that the protective cap is the most
indoor. The camera goes about as the cell phone for astounding position the sensor can reach as expressed
positioning articles. The laser gadget is situated by Beauregard. The calculation utilized in this
towards the roof and laser pillars are on the roof. The framework has two stages, step discovery and
camera tracks the laser shafts and modifies its estimation and heading location and estimation. In
introduction concerning the laser bars area. Optical recognizing and evaluating the progression, they
indoor positioning is utilized more for robot self- distinguish the speed of the development and the
localization inside an indoor domain. length of the progression while in heading location

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


247
Meenakshi D. Dhobale et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 244-250

and estimation the bearing of development. The Project onto way, (iii) Find the objective position,
constraint in this framework is that the sensor on the and (iv) Compute the direction point.
protective cap must be coordinated to the bearing of
the person on foot's development. Woodman [19] proposed a Bat framework. Bat
framework utilizes a foot-mounted inertial unit for
Another methodology given by Robertson [17], he Pedestrian Dead Reckoning and a Wifi innovation to
proposes utilizing foot-mounted inertial sensors position people on foot inside a multi-story building.
which give the person on foot dead retribution. He Bat framework utilizes molecule channel calculation
gave a framework called FootSlam. This framework which is Bayesian channels. Woodman proposed an
utilizes a foot mounted Inertial Measurement Unit adjustment calculation to address the float in inertial
(IMU). In addition, it assembles a 2D guide of the estimations with the person on foot development.
working with no earlier information of the structure Woodman's calculation helps in assessing the
of the building. The learning of all client expectation situation of the person on foot at time k when
assists with shaping the guide of the building and knowing his situation at time k-1.
managing the client through it. Notwithstanding, the
more occasions places are visited inside the building Shao [20] talks about a relative region; he executed
the better data or guide worked in respects of that self-ruling people on foot in a 3D domain. He
put. Recreation results demonstrate a precision of 1-2 demonstrated how his self-sufficient people on foot
meters. associate inside nature and how they fabricate maps
of the earth. He characterized the maps as
Fischer [18] recommends utilizing Ultrasound Topological guide, recognition guide and way outline.
reference points to give better precision and less The topological guide speaks to the ecological
heading mistakes in Pedestrian Dead Reckoning. His districts and the bolts speak to the openness between
methodology joins Ultrasonic and PDR. The PDR them. The discernment maps include stationary
calculation contains two sorts of stages inside the articles, which are the nearby items. Where the
means, a positioning stage, and a swing stage. Fischer versatile articles are the worldwide items. At long last,
saw there are likewise two kinds of mistakes that can the way maps can be either framework, which is for
happen inside PDR, these blunders are the heading short way mapping (e.g., seat) or Quadtree which are
blunder and the separation blunder. As examined for long way arranging. Shao's maps technique for his
before, the heading is the bearing of the development; independent people on foot may be extremely
in this way, the heading blunder is the mistake valuable whenever utilized for walker direction
happens in the appeared of the walker. The through indoor positioning.
separation blunder as expressed by Fischer has fewer
events. Fischer's framework is connected primarily to III.CONCLUSION
the safeguard group specialist on call. The framework
functions as pursues; the specialists on call put the In this paper, we have surveyed indoor positioning
ultrasonic sensors on their way as they continue systems, the standards of positioning and calculations
inside the building. PDR framework is utilized to get utilized in different conditions. We talked about the
the area of the people on foot. Be that as it may, difficulties that can be confronted when planning an
ultrasonic sensors influence the modifications so as to indoor positioning framework, expounded on various
lessen the heading mistake. His calculation has four arrangements proposed to beat these difficulties and
stages (I) Estimate the individual's position, (ii) given a conceivable zone of utilization and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


248
Meenakshi D. Dhobale et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 244-250

improvement inside indoor positioning systems. We Workshop on Software Technologies for Future
saw Indoor Positioning is a tremendous zone with Embedded Systems.
numerous applications and numerous enhancements [9] Lionel M.NI, Yunhao Liu, Iu Cho Lau,
to be conveyed. As referenced before in this paper, Abhıshek P. Patil; LANDMARC: Indoor
explores inside indoor positioning could be Location Sensing Using Active RFID
increasingly valuable whenever conveyed for people [10] Guang-yao Jin, Xiao-yi Lu, Myong-Soon Park.
on foot positioning since the exactness came to isn't 2006. An Indoor Location Mechanism Using
as precise as the precision came to in settled indoor Active RFID. Proceedings of the IEEE
positioning. International Conference on Sensor Networks,
Ubiquitous and Trustworthy Computing.
IV.REFERENCES [11] Lijan Jiang, Lim Nam Hoe, Lay Leong Loon.
2010. Integrated UWB and GPS Location
[1] Solrun Furuholt Pedersen, "Micro Positioning". Sensing System In Hospital Environment. IEEE
Master Thesis. ITEM NTNU, Jun 15, 2014. International Conference of Sensor Networks.
[2] Fhelelboom, Zuher. "Equipment tracking and [12] Tilch, S. Mautz, R. 2010. Current investigations
security systems for hospitals". Master Thesis at the ETH Zurich in optical indoor
Universitiy Teknology Malaysia. May, 2017. positioning. IEEE Positioning Navigation and
[3] I. Guvenc, C.-C. Chong, “A Survey on TOA Communication (WPNC) conference.
Based WirelessLocalization and NLOS [13] S. Beauregard and H. Haas. Pedestrian dead
Mitigation Techniques”, IEEE Communications reckoning: A basis for personal positioning. In
Surveys and Tutorials, vol. 11, no. 3, 2009, Proceedings of the 3rd Workshop on
Pages: 107-124. Positioning, Navigation and Communication,
[4] R.Want, A.Hopper, V.Falcao and J.Gibbons; 2006.
The active Badge location system, ACM [14] V. Fox, J. Hightower, L. Lin, D. Schulz, and G.
Transactions on Information systems Vol. 40, Borriello, “Bayesian filtering for location
No. 1, pp. 91-102, January 2012. estimation,” IEEE Pervasive Computing, vol. 2,
[5] Hazas, M., Hopper, A; A Novel Broadband no. 3, pp. 24–33, 2003.
Ultrasonic Location System for Improved [15] Matthies, L. H., Szeliski, R., and Kanade, T.
Indoor Positioning, IEEE Transactions on (1989). Kalman filter-based algorithms for
mobile Computing, Vol. 5, No. 5, May 2006. estimating depth from image sequences.
[6] Michael Popa, Junaid Ansari, Janne Riihij¨arvi, International Journal of Computer Vision, 3,
and Petri M¨ah¨onen. 2008. Combining 209–236.
Cricket System and Inertial Navigation for [16] Beauregard, S., (2006). A Helmet-Mounted
Indoor Human Tracking. WCNC proceedings. Pedestrian Dead Reckoning System.
[7] Priyantha, N. B; The cricket indoor location Proceedings of IFAWC2006, TZI University
system: PhD Thesis, Massachusetts Institute of Bremen, Germany, pp. 79-89
Technology. 199 p, June 2005. [17] Robertson, P., Angermann, M, Krach, B.,
[8] Fukuju, Y.; Minami, M.; Morikawa, H.; Simultaneous Localization and Mapping for
Aoyama, T.; Dolphin. 2003. An autonomous Pedestrians using only Foot-Mounted Inertial
indoor positioning system in ubiquitous Sensors. In Proc. Ubi Comp 2009, ACM (2009)
computing environment, in Proc of the IEEE 93-96.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


249
Meenakshi D. Dhobale et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 244-250

[18] C. Fischer et al., “Ultrasound-Aided Pedestrian


Dead Reckoning for Indoor Navigation,” Proc.
Int’l Workshop Mobile Entity Localization and
Tracking in GPS-Less Environments, ACM
Press, 2008, pp. 31–36.
[19] O. Woodman and R. Harle. Pedestrian
localisation for indoor environments. In Proc.
of the UbiComp 2008, Seoul, South Korea, Sept.
2008.
[20] SHAO W., TERZOPOULOS D.: Autonomous
pedestrians. In SCA ’05: Proceedings of the
2005 ACM SIGGRAPH/Eurographics
symposium on Computer animation (New
York, NY, USA, 2005), ACM Press, pp. 19–28.

Cite this article as :

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


250
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/ CSEIT195175

A Review Study on Online Job Portal


Aafreen Khan1, Anirudha Wankhade1, Priyanka Pakhide1, Sagar Meshram1, Prof. Sonali Zunke2
1BE Students, Department of Information Technology, JD College of Engineering and Management, Nagpur,
Maharashtra, India
2Assistant Professor, Department of Information Technology, JD College of Engineering and Management,

Nagpur, Maharashtra, India

ABSTRACT

Gaining information and explicit job skills have turned into the fundamental destinations for understudies in
the colleges. Learning is important to settle on educated choices, particularly, in a basic circumstance. Learning
and information the executives in any organization are pivotal to give it an aggressive edge in the present
testing and globalized condition. In this paper we present the plan of different on-line recruitment framework,
that enables businesses to post their job advertisements, which job searcher can allude to, when searching for
jobs. This job portal can catch job prerequisites dependent on industry needs.
Keywords: Knowledge Sharing, Web Portal, Job Portal, Online Recruitment.

I. INTRODUCTION to get another line of work, he/she can present a


resume utilizing word handling programming like
Joblessness is one of the genuine social issues Microsoft Office Word, open an internet browser to
looked by both creating and created nations. For send the resume and get an email. Online
instance, in Europe the rate of joblessness has been recruitment has turned into the standard strategy for
expanding quickly since the 1970's. Dorn and Naz [1] businesses and jobseekers to meet their particular
referenced that one reason for this issue is the targets. The businesses transfer the job contributions
unjustifiable appropriation or absence of data on job in to the job portals. The majority of substantial
openings so individuals cannot presently the new job organizations as well as the little ones has
opportunities. It implies that there are a few jobs acknowledged online recruitment. The organizations
accessible, yet jobseekers do not approach that data. send data or jobs opening for posting on the portals
An effective pursuit of the web may help to and speak with the candidates through the Internet
jobseekers in their job chase. Some web-based and Email.
interfaces give a productive method to look the web
for online data on job opportunities for jobseekers [2]. Gangle [4] characterized the idea of online
recruitment or e-recruitment as the utilization of the
Today, the web has changed numerous parts of our Internet to scan for jobs, which have been publicized
life, for example, the manner in which we search for electronically. Along these lines, the businesses
jobs [3]. On the off chance that one individual needs publicize the job openings, spare the resume and

CSEIT195175 | Received : 01 Feb 2019 | Accepted : 09 Feb 2019 | January-February - 2019 [ 5 (1) : 251-255 ]
251
Aafreen Khan et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 251-255

educational modules vitae (CV) of candidates, and different arrangements they need to make. Securing
contact the jobseekers who are qualified, on the web. all accessible position opening is a principle venture
Today, the Internet has turned out to be one of the at in the job-chasing process.
key strategies for landing data identifying with
position opening. Substantial foundations, similar to The Internet is presently an incredible asset that
colleges incorporate data on vocation prospects in jobseekers can utilize. Today, numerous locales
their sites, which are likewise connected to publicize job positions to be filled by individuals with
recruitment locales. specific skills in different fields. The Internet assumes
an essential job in the region of human asset
II. LITERATURE REVIEW arranging and advancement. Most arranging and
advancement organizations are currently utilizing PC
A. Job Procurement: Old and New Ways innovation and the Internet for staff recruitment. It
Job looking for more often than not includes ought to be noticed that despite the fact that the
distinctive approaches to search for jobs, for example, Internet has encouraged the procedure of job chasing,
through close to home contacts, direct phone calls to it has not supplanted the conventional techniques,
businesses, job organization office, filtering on the totally.
web job postings, and so on [3]. Prior to the Internet,
turned out to be generally utilizes as a strategy for B. Significance of Job Portals
looking for jobs; jobseekers invested a heap of energy
utilizing different strategies to search for job In the period of innovation, the Internet has
openings. Today, jobseekers utilize online techniques, turned into the principle wellspring of data for
which are extremely advantageous and spare a great jobseekers. Extensive partnerships, foundations, and
deal of time. Galanaki [5] records the accompanying colleges incorporate data on vocation prospects on
strategies to be the conventional (old) routes for their sites. As per a review, 70% of the workforce
recruitment: utilizes sites or portals on the Internet to look for jobs
in France. These sites or portals give an internet
• Employment recruitment offices searcher to get to data on job openings [7].
• Job fairs
• Advertising in the broad communications, for Sulaiman and Burke [8] found that most businesses
example, papers are quick to utilize online recruitment strategies for
• Advertisement in TV and radio getting staff. He referenced that online recruitment
• Management Consultants strategies can recognize the best candidates. That is
• Existing worker contacts the motivation behind why progressively created
• Schools universities or colleges understudies nations, for example, Malaysia have begun to utilize
administrations office online job portal as one of the vital method to enlist
• Workers or expert referrals individuals to fill job opening. An examination done
in 2006, found that 21% of web clients in the EU
These old jobs looking for techniques are utilized the web to look for jobs or to send job
excessively moderate, distressing, testing and applications. In 2007, this had expanded to 67% for
furthermore need quality [6]. What's more, the jobless individuals [5].
candidates need to consider the expense and the
measure of time to get the data they need, and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


252
Aafreen Khan et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 251-255

Most organizations distribute their job opening on individuals approach the job data from anywhere on
their site, or utilize online jobsites. These techniques the planet, while with the paper, data is spread at
result in incredible sparing in expenses. Mochol and nearby dimension. One key part of job portal is the
Nixon [9] expressed that the utilization of semantic expense. Organizations spend less to distribute or
web innovation gives advertise straightforwardness, publicize job opportunities on the portals or sites,
higher speed of acquisition yet decreased exchange when contrasted with the utilization of other media,
cost. Today, the Internet is utilized for an expansive for example, paper or job fairs. Besides, online
number of business exchanges. Individuals observe recruitment is quick, and spares time. Once the
the Internet to be a compelling specialized apparatus. managers transfer the job opening on the portal, the
In a report in 2005, it was discovered that 90% of jobseekers can see it and send in their resume.
jobseekers in Germany utilize the web to search for
jobs. A purpose behind this high rate of Internet is Along these lines, cost and efficient are two
that candidates are youthful and very qualified and noteworthy points of interest of job portals. Ganalaki
utilize the web a great deal, and numerous [5], refered to in Rosita and Nadianatra (2007),
organizations distributed their job openings on the expressed that the Internet is an instrument without
web and through their portal. fringes, and is an amazing strategy to come to an
around the world gathering of people.
Job portals are the beginning stage of jobseekers
while scanning for jobs. In this way, some job portals C. Highlights of Job Portals
charge businesses high expenses to distribute data on
job opportunities. Regardless of this, numerous One of the approaches to enhance work portability
businesses still keep on publicizing or distribute data is to give online job offer administrations. Online job
on job openings hands on portal, however portals can help jobseekers as they contain all
constrained to minimize expenses. Numerous required data about accessible opening in a solitary
businesses still trust that a jobseeker will visit job point. Such portals upgrade effectiveness in job
portals while hunting down job opening [10]. recruitment as candidates can coordinate their
capabilities and skills to the prerequisites of bosses.
A great job portal can likewise bolster learning
sharing among the individuals. The quantity of For the most part, looking for jobs on the web
online job portals keeps on expanding. It is trusted includes a procedure of data gathering in light of the
that seventy-five percent of individuals who are fact that the jobseeker accumulates data contained in
looking for jobs, utilize the web and on the web the job portals, amid the inquiry [11]. A great job
portals. Gangle [4] expressed that online recruitment portal imparts data and encounters to its
has the accompanying points of interest: bosses can individuals/clients. This spare time and endeavours
distinguish an extensive number of qualified job and better choices can be made [12]. Job openings
searchers and get their data effectively. It implies that necessities can be coordinated to a candidate's
organizations or organizations can broaden the capability and skills. Along these lines, job portals
inquiry area, subsequently, they have better prospect return the exact matches as well as return the most
of choosing the most qualified applicants. Web gives comparable match. The individuals from the
managers an approach to draw in a higher number of European Commission (EC) expressed that online job
applicants, particularly, the individuals who satisfy portals ought to have very comparable attributes. The
the job necessities. With online recruitment, attributes included are:

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


253
Aafreen Khan et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 251-255

• An online accessible database of positions for job portals to have the necessary information for the
searcher; offices to send cvs to the site; students. Fig. 1 shows the schema of the proposed job
• Email alarms of jobs which coordinate the clients web portal, in illustrating the relationship between
profile; the students and the faculty, and the unemployed
• Additional guidance, for instance, about working graduates and the potential employers.
in remote nations or profession direction;
• The capacity to oversee job applications; Nowadays, most modern universities web portals
• Businesses must be able to distribute and oversee provide information and services to achieve various
job positions, look through the CV database; goals and to improve the quality of education with
• Have online contact with potential jobseekers. these services. Based on information from the
literature review, as well as the students’
III.Discussion and Analysis requirements identified from the survey, a job portal
should be develop. The portal should be a knowledge
Analysis of the data collected from the survey sharing system for the students to create a better
indicates that most of students are not satisfied with academic environment for their studies.
the current faculty portal. The respondents said that
they have problem to get the needed information IV.CONCLUSION
when they are looking for a job. In addition, there is
not enough information and resources related to help In perspective of the prior discoveries, the invalid
them. The students looking for a proper Job want a speculation expressing that there is no huge contrast
knowledge sharing system to get such information in the respondents' assessment on the viability of the
and help them in their decisions. created Online Job Recruitment System regarding
execution, dependability, security, and cost-adequacy
Base the development of the new web portal should was acknowledged. It is inferred that the created
on the feedback from the survey and the programming was successful in choosing qualified
requirements of the students. In recent years, the candidates inside a shorter timeframe. Subsequently,
structure of university portals has changed from static it would turn into a noteworthy supporter of the
information to dynamic and transaction mode. It is value contract. It is additionally inferred that
very important for the faculty to provide the students Performance, Reliability, Security, and Cost-
with online access to different services and adequacy could be used as criteria in assessing web
transactions such as online enrolment, course support recruitment programming.
and job information. Suresh Ram stated that 70% of
students, who finished their studies at tertiary level In light of the ends drawn, it is prescribed that the
education, could not find a job. This shows how product is actualized in the organization. After
important it is to have access to the right information execution, it is recommended that the effect of the
to find a job. It is important for the universities that framework on the recruitment procedures is resolved
provide information to jobless graduates on the jobs and that the adjustments in the recruitment forms
or career opportunities. with the presentation of the online recruitment
programming be examined. It is proposed that
Sulaimun and Burke [8] emphasized that it is reviews on the client experience, reception or non-
necessary for the universities to pay more attention selection of the product, and mistakes submitted in
to their knowledge management system or web

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


254
Aafreen Khan et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 251-255

utilizing the product be directed and that the [8] N. Sulaiman and M. Burke, “A case analysis of
Usability model of FURPS be surveyed. knowledge sharing implementation and job
searching in Malaysia,” International Journal of
As far as an upgrade of the product, it is Information Management, 2009.
unequivocally prescribed that an online test is joined [9] M. Mochol, H. Wache, and L. Nixon,
in the recruitment and that additional security “Improving the accuracy of job search with
highlights, for example, the dimension of access semantic techniques,” in Proc. Business
grouped by the situation in the organization likewise Information Systems, 10th International Conf.,
be fused in the product. BIS 2007, Poznan, Poland, April 25-27, 2007,
Springer, pp. 301-313.
V. REFERENCES [10] M. Mansourvar, Development of a Job Web
Portal to Capture Industry’s Needs, 2011.
[1] J. Dorn and T. Naz, “Integration of Job portals [11] A. Weber and H. Mahringer, “Choice and
by Meta-search,” in Proc. 3rd International success of job search methods,” Empirical
Conf. on Interoperability for Enterprise Economics, vol. 35, no. 1, pp. 153-178, 2008.
Software and Applications, Funchal, Portugal, [12] M. Mansourvar and N. B. M. Yasin,
2007, pp. 401-412. “Knowledge portal: a tool to capture university
[2] S. Bsiri, M. Geierhos, and C. Ringlstetter, requirements,” in Proc. 2011 International
“Structuring job search via local grammars,” Conf. on Graphic and Image Processing,
Advances in Natural Language Processing and International Society for Optics and Photonics,
Applications, pp. 201, 2008. October 2011, pp. 82850F-82850F.
[3] M. Mansourvar and N. Y. Mohd, “Web portal
as a knowledge management system in the Cite this article as :
universities,” World Academy of Science,
Engineering and Technology, vol. 70, pp. 968- Aafreen Khan, Anirudha Wankhade, Priyanka
974, 2010. Pakhide, Sagar Meshram, Prof. Sonali Zunke, "A
[4] M. Gangle, “The only way is up? Employment Review Study on Online Job Portal", International
protection and job mobility among recent Journal of Scientific Research in Computer Science,
entrants to European labour markets,” Engineering and Information Technology
European Sociological Review, vol. 19, pp. 429, (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
2007. 251-255, January-February 2019.
[5] E. Galanki, “The decision to recruit online: a
descriptive study,” Career Development
International, vol. 7, pp. 243-251, 2002.
[6] S. Mauno, U. Kinnunen, and M. Ruokolainen,
“Job demands and resources as antecedents of
work engagement: A longitudinal study,”
Journal of Vocational Behavior, vol. 70, 2007,
pp. 149-171.
[7] A. Doyle, Internet Your Way to a New Job:
How to Really Find a Job Online, Happy about,
2008.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


255
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195158

Voice Based Navigation System for the Blind People


N. Vijayalakshmi 1, K. Kiruthika 2
1 Assistant Professor, Department of Computer Science, Bon Secours College for Women, Bharathidasan
University, Thanjavur, Tamil Nadu, India
2M. Sc Computer Science, Bon Secours College for Women, Bharathidasan University, Thanjavur, Tamil Nadu,
India
ABSTRACT

This main aim of this project to provide voice-based navigation system for blind people using voice recognition
module and it is intended to provide overall measures object detection and real time assistance via Global
Positioning System (GPS) and ultrasonic sensors. This project aims at the development of an Electronic
Travelling Aid (ETA) kit to help the blind people to find obstacle free path. This ETA is fixed to the stick of the
blind people. When the object is detected near to the blinds stick it alerts them with the help of vibratory
circuit. The blind person will give the destination’s name as the input to voice recognition module. GPS module
continuously receives the latitude and longitude of the current location. GPS compares it with the destination’s
latitude and longitude. The blind person receives the pronounced directions which he needs to follow to reach
his destination.
Keywords : Global Positioning System, Electronic Travelling Aid, Navigation System, Latitude and Longitude.

I. INTRODUCTION By the introduction of a basic sensor, the provisions


to the aid of blind people have remarkably increased.
Visual impairment can limit people’s ability to Many universities and companies like IBM have
perform everyday tasks and can affect their quality of given a part of their focus to the development of aids
life and ability to interact with the surrounding for the blind people. Some of the popular ones are
world. Blindness, the most severe form of visual smart canes and obstacle sensors. Recent
impairment, can reduce people’s ability to perform development includes self-driving cars and smart
daily tasks, and move about unaided. Good quality glasses. In the last 30 years, various other strides that
rehabilitation allows people with different degrees of have been developed are the text or speech softwares
visual impairment to fully profit from life, achieve and smartphone apps. Moreover the systems that are
their goals and be active and productive in today’s being developed on robotics and artificial intelligence
society. Ample efforts have been made to aid the will be very advantageous to them as well.
blind by innovating and improving technologies.
According to [23] “The emerging ethics of human One of the major factors in developing these
centric GPS tracking and monitoring”, the main technical aids is the compatibility with the user. He
factors focused while tracking a person are privacy, should not have trouble getting acquainted with the
accuracy and accessibility. product. The features of the product should not be
too difficult to use. Even the notification systems
used to alert the blind should be comfortable and

CSEIT195158 | Received : 01 Feb 2019 | Accepted : 10 Feb 2019 | January-February -2019 [ 5 (1) : 256-259 ]
256
N. Vijayalakshmi, K. Kiruthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 256-259

reliable. Another major factor is the cost of such also in case of an emergency, the location of the blind
products. Since they are already paying for person can be sent to a relative or friend via SMS.
treatments or other nursing cares, the price of the
product should be in the range that is reasonable. II. METHODOLOGY
Other feature of these products should be the
durability. The users might not be able to charge the
system. So, appropriate measure should be taken for
it.

To aid visually impaired and to provide a compact


and complete solution, a smart stick can be developed
which will include a GPS system, obstacle detection
mechanism and an audio module, so that the stick
can instruct the directions to the user. The obstacle
detection mechanism is combined with navigation Figure 1 : Block Diagram
and location detection using GPS-GSM and an audio
module with Bluetooth transceiver to provide a The block diagram of main board is shown in Figure1.
better solution than the already existing solutions. In this diagram the 32-bit ARM processor (LPC2148)
is used, which is the heart of this project. The
The technologies we plan to use are sonar technology, LPC2148 microcontrollers are based on a 16-bit/32-
GPS navigation, GSM network, voice recognition and bit ARM7TDMI-S CPU with real-time emulation and
notification. Sonar technology is used for obstacle embedded trace support, that combine
detection. The principle of the system works in the microcontroller with embedded high speed flash
way that the detector sends and receives a signal. memory ranging from 32 kB to 512 kB. A 128-bit
Upon the reception of the signal, the distance is wide memory interface and unique accelerator
calculated using the time it took for the signal to architecture enable 32-bit code execution at the
reflect back. It is a simple procedure which works maximum clock rate. For critical code size
with the properties of electromagnetic waves. This is applications, the alternative 16-bit Thumb mode
mainly used to find distances in various fields. reduce code by more than 30 % with minimal
Obstacle detection is done by ultrasonic sensors performance penalty.GPS receiver is used to get the
attached to the stick. The sensors are selected to current location in the form of longitude and latitude.
satisfy their criteria of sensor angle, distance and The GPS used here is GR87.The output of GPS
other parameters. When the sensors come across receiver is given to the processor using serial
some obstacle, it sets off a voice announcement from communication. In this system output is in the form
the set of announcements already recorded, of voice hence speaker/headphones are used.
according to their semantics and alerts the person
about the obstacle in the path. In order to reduce the GPS GR87 is a highly integrated smart GPS module
noise and sound dissipation, an earpiece is attached to with GPS patch antenna that is ceramic. The antenna
give the audio announcements. The GPS system is is connected to the module through the LNA. The
used to detect the current location and helps the user module has 51 channel acquisition engine and 14
to reach the destination by announcing the current channel track engine, which is capable of receiving
location name whenever required by the user and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


257
N. Vijayalakshmi, K. Kiruthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 256-259

signals from up to 65 GPS satellites and transferring 3.2 GPS RECEIVER


them into the precise position and timing
information that can be read over either UART port Global Positioning System (GPS) satellites broadcast
or RS232 serial port. Low power consumption is signals from space which are used by GPS receivers,
needed for the small size and high end GPS to provide three-dimensional location (latitude,
functionality. Both the LVTTL-level and RS232 longitude, and altitude) and precise time. Reliable
signal interface are present on the interface positioning, navigation, and timing services are
connector and a supply voltage of 3.6V~6.0V is provided by the GPS receivers to users all around the
supported. The smart GPS antenna module is world continuously in all weather, day or night,
available as an off-the-shelf component which is 100% anywhere on or near the Earth. The GPS receiver
tested. The smart GPS antenna module can be offered used in this project is GR87.Its main features are
for OEM applications. Also, the antenna can be tuned On chip 1Mb SRAM
to the final systems conditions. Low power consumption
Multi path mitigation hardware
III. HARDWARE DESCRIPTION Reacquisition time 0.1 seconds

3.1 MICROCONTROLLER 3.3 VOICE RECOGNITION


A speech analysis is done after the user speaks in a
The Controller used will be ARM LPC2148 which is microphone and inputs are thus taken. The
based on 32/16 bitARM7TDMI-S CPU with real-time manipulation of the input audio signal is done at the
emulation and embedded trace support that combines system level. Different operations are performed at
the microcontroller with embedded high-speed flash different levels on the input signal such as Pre-
memory ranging from 32 kB to 512 kB. It has many emphasis, Framing, Windowing, Mel Cepstrum
important features like. 16/32-bit ARM7 analysis and Recognition (Matching) of the spoken
microcontroller in a tiny package. It has on-chip word. The speech recognition system consists of two
static RAM and on-chip flash program memory. It distinguished phases. The first one is training session,
offers real-time debugging and high-speed tracing of while, the second one is referred to as operation
instruction execution. The features of LPC2148 are session or testing phase. During training phase,
speaker has to provide samples of their speech to
• USB 2.0 Full Speed compliant Device Controller. train the system. During recognition phase, speaker
• 10 bit A/D converters. has to give samples of his speech to match with
• Multiple serial interfaces with two UARTs. existing database and provides exact match. The voice
• Low cost, low consumption, easy handling and recognition IC used here is IC HM2007.
flexibility.
IV. CONCLUSION
These features make the controller reliable for the
project. The world's largest number of blind people and
India’s current population is over 1.22 billion. Earlier
majority of visually impaired people prefer to not use
electronic aids, and use only canes or guide dogs. The
underlying reasons for this include the relatively

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


258
N. Vijayalakshmi, K. Kiruthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 256-259

high costs and relatively poor levels of user sensors. Information and Automation (ICIA),
satisfaction associated with existing electronic 2010 IEEE International Conference on June
systems. So we tried to develop a low cost and user 2010
friendly system for blind people with greatest [7]. Baranski, P.; Polanczyk, M.; Strumillo, P.; A
possible accuracy. This method offers innovative remote guidance system for the blind. e-Health
solutions in order to replace the conventional Networking Applications and Services
methods of guiding visually impaired person. Also, it (Healthcom), 2010 12th IEEE International
can be easily applied anywhere where it can handle Conference .
places like mall, airports etc. In this project we have
used ARM processor which contains more memory Cite this article as :
and its operating speed is high. We guide or navigate
the blind people using voice. N. Vijayalakshmi, K. Kiruthika, "Voice Based
Navigation System for the Blind People",
V. REFERENCES International Journal of Scientific Research in
Computer Science, Engineering and Information
[1]. Amit Kumar,Rusha Patra, M. Manjunatha, J. Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
Mukhopadhyay and A. K. Majumdar An 5 Issue 1, pp. 256-259, January-February 2019.
electronic travel aid for navigation of visually Available at doi :
impaired Communication Systems and https://doi.org/10.32628/CSEIT195158
Networks (COMSNETS), 2011 Third Journal URL : http://ijsrcseit.com/CSEIT195158
International conference on 4-8 jan 2011.
[2]. Shamsi, M.A.; Al-Qutayri, M.; Jeedella, J.; Blind
assistant navigation system Biomedical
Engineering (MECBME), 2011 1st Middle East
Conference on 21-24 Feb. 2011
[3]. Calder, David J.; Curtin .An obstacle signaling
system for the blind ,Digital Ecosystems and
Technologies Conference (DEST), 2011
Proceedings of the 5th IEEE International
Conference on 30 June 2011
[4]. Dunai, L.; Fajarnes, G.P.; Praderas, V.S.; Garcia,
B.D.; Lengua,I.L.; Real-time Assistance
prototype- a new Navigation Aid for Blind
people, IECON 2010 - 36th Annual Conference
on IEEE Industrial Electronics Society
[5]. Shang Wenqin; Jiang Wei; Chu Jian; A machine
vision based navigation system for the blind
China Computer Science and Automation
Engineering (CSAE), 2011 IEEE International
Conference on 10-12 June 2011
[6]. Hashino, S.; Ghurchian, R.; A blind guidance
system for street crossings based on ultrasonic

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


259
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195161

Data Encryption Key Sharing Using Image Pixel Color Value


Vijay Gokul Koli, Raj Kumar Paul
Department of Computer Science & Engineering, Vedica Institute of Technology, Bhopal, India

ABSTRACT

Security is required to transmit confidential information over the network. Security is also demanding in wide
range of applications. Cryptographic algorithms play a vital role in providing the data security against
malicious attacks. Massage communication over internet facing problems like data security, copyright control,
data size capacity, authentication etc. A new idea is to apply reversible data encoding algorithms images by
wishing to remove the embedded data after data receiving with the help of image pixel color value.

Keywords: Network Security, Cryptographic Algorithms, Image Pixel, Data Encryption, Public-Key
Cryptosystem, Encryption, Decryption

I. INTRODUCTION In this technique a public-key cryptosystem (RSA)


using two public key and some mathematical
In the field of networking, role of network security relation. This two public keys are sent
is immense. In the age of information need to keep separately.Security is most important to transmit
information about every aspect of our live. These confidential data over the network, in the today’s
information needs to be hidden from unauthorized world. In wide range of applications, Security is
access (confidentiality), protected from also demanding. For data security Cryptographic
unauthorized change (integrity), and available to an algorithms play a vital role against malicious
authorized entity when it is needed (availability). attacks. In the most popular implementations of
Hence the way of keeping the information securely Public Key Infrastructures, RSA algorithm is
is known as cryptography, which comes from a extensively used. In this paper [1] an algorithm
word with Greek origin, means “secret writing”. has proposed for RSA a method for implementing
Many cryptographic algorithms are developed to a public-key cryptosystem (RSA) using two public
achieve the above said goal. The algorithms should key and some mathematical relation. This two
be such that an opponent cannot defeat its purpose. public keys are sent separately, this makes the
attacker not to get much knowledge about the key
and unable to decrypt the message. Two different
II. METHODS AND MATERIAL keys are used in Public Key cryptography. One key
is used for decryption & only the other
A Modified RSA Encryption Technique Based on corresponding key must be used for encryption.
Multiple public keys Not any other key is possible to decrypt the
message, even the original (i.e. the first) key can’t
used for encryption. Every communicating party

CSEIT195161 | Received : 01 Feb 2019 | Accepted : 10 Feb 2019 | January-February -2019 [ 5 (1) : 260-265 ]
260
Vijay Gokul Koli, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019; 5(1) : 260-265

requires pair of key for communicating with any side, the customer representatives only see the
number of other parties. It is beauty of this names list.
scheme. Once someone obtains a key pair, he can
When operation is needed, software installed on
communicate with anyone else. They have done
the customer representatives’ computer or cell
implementation of RSA algorithm efficiently using
phone were decrypt the data and send them to the
two public key pairs and using some mathematical
call center directly without touching the
logic rather than sending the e value directly as a
representatives.
public key.

Fig 1. RSA algorithm

A new personal information protection approach


based on RSA cryptosystem

With the widespread and rapid development


application of the information technology, the
Fig. 2. The encryption approach of customers’
communication pattern has obviously changed
information
among individuals, corporations and even nations.
However, convenient network-based
Key Distribution by Using Public Key Algorithm
communication method brings not only the benefits
(RSA)
but also some disadvantages such as individual
information leak. This paper [2] introduced that,
personal information can be transformed from plain Authors suggested a new model for quantum key
text into cipher text. Customer representatives can distribution among three parties or more where
be able to contact their clients without seeing the there is a trusted center that providing the
privacy. On the server side, the system necessary secret information of clients to securely
administrator has the permission of authorization communicate to each other.
management. They devolve the authorization to Using the current computing systems classical
database administrators and then database cryptography is based on the computational
administrators input customers’ information into difficulty to compute the secret key. Depending
the system. At the same time, sensitive information only on the difficulty of computational complexity
such as phone number is encrypted. On the client does not provide enough security because finding a
fast method to calculate the secret key. It

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


261
Vijay Gokul Koli, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019; 5(1) : 260-265

compromises the security of the systems. Law of In RSA computation is lengthy and some less
physics is used in Quantum computing for secure. This paper[5] present a new algorithm to
communication. In cryptography and key presents the modified form of new RSA algorithm
distribution quantum theorems and principles are in order to boost up the speed of the
applied. In this paper[3], new model for quantum implementation of RSA algorithm during data
key distribution are introducing among three exchange across the network world. In this method
parties or more where there is a trusted center that keys are stored offline before the process start.
providing the necessary secret information of Thus, the speed of process increased as compared
clients to securely communicate to each other. to original RSA method.

To compare the bases, classical channel is used by Enhancing the Security of The RSA Cryptosystem
quantum key distribution protocols BB84, B92 and
ERP.
This paper [6] increases the security of the RSA
Loop-based RSA Key Generation Algorithm using algorithm, this enhancement use randomized
String Identity parameter to change every encrypted message
block such that even if the same message is sent
more than once the encrypted message block is look
This paper [4] propose i-RSA algorithm that is focus
different. This paper suggests that how to use
on key generation algorithm. user identity is
randomized parameters in the encrypt the data to
Enhancement of this algorithm. it can be used as a
make RSA. By this enhancement it makes the RSA
public key, such as email address. The key
semantically more secure. Means an attacker cannot
certificates are used to authenticate the user's key
distinguish two encryptions from each other, even
pair. So certificate does work as important role in
if the attacker knows (or has chosen) the
secure communication but to issue the certificate is
corresponding plaintexts (original message). In this
a big challenge and it also increases the overhead
Work comparison between the modified RSA and
due to the increasing cost. For public key Previous
the basic RSA version introduced. Enhancement
algorithm has successful used email identity, but all
can easily be implemented on this Work. Also other
type of email can’t be used as a public key. So the
attacks are presented by this paper, also how to
propose i-RSA algorithm that can produces 66.6%
speed up the RSA encryption and decryption
compared to previous algorithm (46.67%) email can
process is an important issue for the RSA
be a string public key. in key generation looping
implementation.
process is the main differences between i- RSA and
previous algorithm, to get new value of p and q Here, RSA is more secure and it may be more
parameter, when value of k is equal to 1, then stronger by applying some techniques. Here They
looping process can stop, and the email can be a have seen that all authors are talking about many
public key. Detail explanations of i-RSA algorithm methods but no one is talking about image pixel for
in propose algorithm section. security purpose. So theyF can add image pixel
technique to make more powerful RSA

algorithm.
Modified RSA Cryptosystem Based on Offline
Storage and Prime Number

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


262
Vijay Gokul Koli, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019; 5(1) : 260-265

Proposed Methodology Algorithm


In this proposed work first of all p and q two prime For both encryption and decryption
number is selected then find p*q after this calculate algorithms are following.
(p-1)*(q-1). Then select e and

after all public key and private key is generated. But


before sharing public key they have to encrypt this.
So that other person who doesn’t belongs to my Encryption
group cannot get public key. (1) Select p and q both prime number, p is not
equal to q.
According to the comparisons and the characteristics (2) Calculate n = p x q.
of RSA, it determined to use RSA cryptography as the (3) Calculate ø (n) = (p -1) x (q-1).
core algorithm for personal information protection in (4) Select integer e whose gcd (Ø (n), e) = 1; 1 < e
information system. It makes users don’t have to <Ø (n).
store a mass of calculated secret keys. The (5) Calculate private key d = ℮-1 (mod Ø (n)).
information owner can easily send messages to the (6) Public key PU = {e, n}.
receiver when he got reliable public key from the (7) Private Key PR = {d, n}.
receiver. This approach makes things easy, only one (8) Im=Load any image img.
pair of keys is necessary. (9) Select pixel position p_id.
(10) For(i=1; i<=P_id; i++)

Encryption {

If(i==P_id)
After generating public key, before sharing to other Px=im(p_id);
people’s sender can be use an image that already
}
have in receiver side. First of all select an image that
already have all receivers then select any pixel color (11) E=Px+e.
value of that image. Then add that color value with (12) Public key to transmit = {E,P_id,n}.
e and save in E. now it doesn’t need to share actual (13) Message (M) Cipher text- C = M^e mod n.
value of e. now it share E and position of that pixel Decryption
in image.

(1) Public key to transmit = {E,P_id,n}. (2)


Decryption For(i=1; i<=P_id; i++)
When receiver get encrypted message then if he {
have same image then they can be select same image
If(i==P_id)
and select pixel position that they have received and
Px=im(p_id);
after all, pick color value of that pixel and minus
that value from E that they have received from }
sender. Means anyone who wants to share public
(3) e=E-Px.
key they only share E and pixel position.
(4) Message M = C^e mod n.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


263
Vijay Gokul Koli, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019; 5(1) : 260-265

Where M is message (Plane text), p and q are prime


numbers, N is common modulus, e and d are public
and private keys, p_id is a pixel position of selected
image, im is a program variable which contain all
pixels color values of selected image, Px is a color
value of pixel position p_id for selected image.

III. RESULT AND ANALYSIS

Figs shows front view of implementation of proposed


work. It has designed using MATLAB. Here sender
side “select image” button used for selecting the
image. Fist text box is using for entering message that
you want to send, second text box is using for giving
pixel position value. And in receiver side first text Fig. 5. Histogram of receiver side image
box is using for giving same pixel position that has
Here, in figs, clearly seen that histogram of sender
given in sender side.
side image is same as histogram of receiver side image
Time Required for Brute force attack because this algorithm don’t modifying image. only
80
data encryption using encrypted image
data encoding using pixel value
pixel color value is using for reference.
70

60
IV. CONCLUSION
Time in Seconds

50

40

30 In this work a new method has proposed for RSA


20 public key sharing. In this method before sharing the
10 public key, e is encrypted with any specified pixel
0
0 50 100 150 200 250 300 350 400 450 500
color value of any particular image. So, it doesn’t
Traffic
need to transfer e. because e is encrypted with any
pixel color value then possibility of attack is very less
Fig. 3. Comparison of brute force attack in existing
as compared to without encrypted that’s why if any
and proposed method using image pixel.
attacker got shared key then they don’t know what is
actual value of e. it is possible if and only if he has
same image. But in this method no one is sharing
image. image is predefined. So after all this method is
complex for attacker to getting.

V. REFERENCES

[1]. Amare Anagaw Ayele, Dr. Vuda Sreenivasarao,


Fig. 4. Histogram of sender side image June 2013, "A Modified RSA Encryption
Technique Based on Multiple public keys",
International Journal of Innovative Research in

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


264
Vijay Gokul Koli, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019; 5(1) : 260-265

Computer and Communication Engineering Cite this article as :


Vol. 1, Issue 4.
[2]. Liang Wang, Yonggui Zhang, 2011, "A New Vijay Gokul Koli, Raj Kumar Paul, "Data Encryption
Personal Information Protection Approach Key Sharing Using Image Pixel Color Value",
Based on RSA Cryptography", IEEE. International Journal of Scientific Research in
[3]. Ammar Odeh, Khaled Elleithy, Muneer Computer Science, Engineering and Information
Alshowkan, Eman Abdelfattah, 2013, Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
"Quantum Key Distribution by Using Public 5 Issue 1, pp. 260-265, January-February 2019.
Key Algorithm(RSA)", IEEE. Available at doi :
[4]. Norhidayah Muhammadi, Jasni Mohamad https://doi.org/10.32628/CSEIT195161
Zaini, Md Yazid Mohd Saman, "Loop-based Journal URL : http://ijsrcseit.com/CSEIT195161
RSA Key Generation Algorithm using String
Identity", 13th International Conference on
Control, Automation and Systems (ICCAS
2013).
[5]. Ms. Ritu Patidar, Mrs. Rupali Bhartiya, 2013,
"Modified RSA Cryptosystem Based on Offline
Storage and Prime Number", IEEE.
[6]. Malek Jakob Kakish, "Enhancing The Security
Of The Rsa Cryptosystem", Ijrras August 2011.
[7]. M. Jason Hinek, Another Look at Small RSA
Exponents, 2006.
[8]. B. A. Forouzan, D. Mukhopadhyay,
Cryptography and Network Security, Tata
McGraw-Hill, 2012.
[9]. Abdullah Darwish Imad Khaled Salah and Saleh
Oqeili, Mathematical Attacks on RSA
Cryptosystem, Journal of Computer Science
(2006).
[10]. J. M. Pollard, A Monte Carlo Method for
Factorization, BIT Numerical Mathematics
(1975).
[11]. H. Riesel, Prime Numbers and Computer
Methods for Factorization, Birkhauser, 1994.
[12]. William Stein, Elementary number theory.
Primes, congruences, and secrets. A
computational approach., New York, NY:
Springer, 2009 (English).

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


265
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195179

A Review on Prediction of Diabetic Retinopathy Using Data


Mining Algorithms
Kajal Sanjay Kothare1, Prof. Kalpana Malpe2
1 M-Tech Student Department of Computer Science and Engineering, Guru Nanak Institute of Engineering &
Technology, Nagpur, Maharashtra, India
2 Assistant Professor Department of Computer Science and Technology, Guru Nanak Institute of Engineering &
Technology, Nagpur, Maharashtra, India

ABSTRACT

The risking components of diabetic retinopathy (DR) were examined broadly in the past investigations, yet it
stays obscure which chance variables were more connected with the DR than others. On the off chance that we
can distinguish the DR related hazard factors all the more precisely, we would then be able to practice early
avoidance systems for diabetic retinopathy in the most high-chance populace. The motivation behind this
examination to study and consider the different predicting mechanisms for the DR in diabetes mellitus utilizing
data mining techniques including the support vector machines, decision trees, artificial neural networks, and
logistic regressions.
Keywords : Data Mining, Artificial neural fuzzy interference system, K-Nearest-Neighbor (KNN), Machine
Learning (ML), Support Vector Machines, Decision Trees

I. INTRODUCTION Diabetic retinopathy is microvascular


inconvenience of diabetes mellitus. It wards on
Diabetic retinopathy (DR) is the most well-known history of diabetes-related difficulties [2]. Diabetic
reason for recently analyzed visual deficiency retinopathy is a profoundly explicit vascular
consistently, particularly in working-age populace. confusion of both kind 1 and type 2 diabetes, with
Review surveys of the United Kingdom Prospective predominance emphatically identified with the
Diabetes Study (UKPDS) and the Diabetes Control length of diabetes [2]. Diabetic retinopathy comprises
and Complications Trial (DCCT) on sort 2 and type 1 of non-proliferative diabetic retinopathy (NPDR) and
diabetes mellitus, separately, both recommended that proliferative diabetic retinopathy (PDR). NPDR is
serious blood glycemic control can adequately lessen otherwise called foundation diabetic retinopathy
the danger of microvascular inconveniences (BDR). NPDR is beginning period of DR, and PDR is
including diabetic retinopathy. Past examinations for late phase of DR. In NPDR status, micro aneurysm,
the most part centered around controlling one discharge, hard exudates, cotton fleece spot,
noteworthy hazard factor which is glucose level, intraregional microvascular variations from the norm,
though few investigations have concentrated on and venous beading are common characters. In PDR
researching diverse hazard variables of DR [1]. organize, there are plate neovascularization, vitreous
discharge, and sinewy scarring. Macular edema is
testimony of hard exudates close macula. Diabetic

CSEIT195179 | Received : 01 Feb 2019 | Accepted : 10 Feb 2019 | January-February -2019 [ 5 (1) : 266-272 ]
266
Kajal Sanjay Kothare, Prof. Kalpana Malpe Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 266-272

retinopathy is the most regular reason for new vision with NPDR. However, when they created
instances of visual deficiency among grown-ups PDR with vitreous drain, they lost the vision all of a
matured 20– 74 years [2]. Around 21% of the sudden. Prior to vision misfortune, photocoagulation
recently determined patients to have type 2 diabetes can evade drain of PDR and diminished vision
(T2D) were likewise found to have co-bleak state of misfortune. In spite of the fact that instruction of
DR, though 60% of the patients with a ceaseless intricacy of T2D in our consideration unit is normal
history of 20 years of T2D were determined to have routine training for patients with T2D, the absence of
diabetic retinopathy [3]. About 20– 40% of patients knowledge is by all accounts the significant reason.
in T2D had diabetic retinopathy and 8% of patients
in T2D had locate compromising diabetic retinopathy One of the significant hazard factors inspected in a
(STDR) in United States [4]. In 2009, Prevalence of pooled examination from populace based
the diabetic retinopathy and poor vision/visual investigations around the globe was the long span of
impairment in Taiwanese patients with T2D were diabetes [8]. Other hazard factors distinguished in
8.91 and 0.62%, separately [5]. this investigation were abnormal state haemoglobin
A1C (HbA1C) and hypertension [8]. As indicated by
There are a few medicines for DR., To begin with, the UKPDS, the rate of diabetic retinopathy is
it is critical to immediately allude patients with any intently connected with the expanding length of T2D,
dimension of macular edema, extreme NPDR, or any and lower dimension of HbA1c can diminish the
PDR to an ophthalmologist who is proficient and danger of experiencing DR in these patients [9]. In
experienced in the administration and treatment of any case, it was seen in clinical practice that a few
diabetic retinopathy [2]. What's more, laser patients with long haul controlled HbA1c levels still
photocoagulation ought to be considered for eyes have dangers experiencing diabetic retinopathy in
with clinically critical macular edema, especially T2D [10]. This proposes the HbA1C level isn't the
when the focal point of the macula is included or main significant hazard factor, and different factors,
unavoidably undermined [6]. Hostile to vascular for example, hypertension, high blood glucose, and
endothelial development factor (against VEGF) length of diabetes may have conceivably played
treatment is additionally demonstrated for diabetic incomplete jobs in the improvement of diabetic
macular edema [2]. retinopathy in T2D. There were realized hazard
components of diabetic retinopathy, for example,
To limit the chances of visual misfortune or the long span of diabetes, poor glycaemic control,
new beginning of visual deficiency of diabetic hypertension, and hyperlipidaemia.
retinopathy, current rules of Taiwan diabetic
association proposes that the screening of fundus In the end, the most well-known hazard factor found
examination in patients with T2D should be by each bit of observational the proof is the term of
performed every year, and performed all the more diabetes. Poor glycaemic control drives a high fasting
much of the time in patients with diabetic glucose level, high, postprandial glucose, and high
retinopathy. It is noticed that the screening rate was HbA1C. Every one of the three previously mentioned
low, for there was just 28.9% of patients with T2D parameters advise diverse parts of diabetes and should
had eye fundus examination in Taiwan in 2009 [7]. all be considered.
One conceivable clarification for the low screening
rate might be that patients with T2D couldn't care
less about retinopathy when they have an ordinary

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


267
Kajal Sanjay Kothare, Prof. Kalpana Malpe Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 266-272

II. RELATED WORK 0.683, which means a decent dimension of


concordance between the watched event of DR and
A few examinations [11– 16] have been created to that anticipated by the model. For the outer approval,
anticipate diabetic retinopathy. A cross-sectional the qualities for C-record and CPE were 0.767 and
investigation of patients with T2D utilized routinely 0.697, separately. The AUC for 1-year survival from
gathered information at outpatient centers of the retinopathy was 0.825. There was no factual
Isfahan Endocrinology and Metabolism Research distinction between the C-record of that determined
Focus (IEMRC), Iran [11]. This investigation in the train informational index versus that
connected collector working trademark (ROC) bends determined on the test informational index (p =
to recognize the ideal estimation of diabetic patients 0.137). At that point, the investigation utilizes the
for deciding DR; affectability and explicitness for classification and relapse tree (CART) examination or
anticipating DR were determined for various cuts of the arbitrary woodland investigation for the train
the score. This examination exhibited the informational collection to confirm how the
consequences of utilizing calculated relapse models outcomes were steady with these distinctive
with DR as the needy variable. The region under the methodologies.
ROC bend (AUC) was 0.704, and furthermore
indicated affectability (60%) and explicitness (69%) Dr. Karim Hashim Al-Saedi et.al.[14] have
of a hazard score ≥ 52.5 for DR. structured and executed a framework to quantify the
effect of diabetic retinopathy utilizing information
Another examination talked about individual mining procedures In this exploration, an exact
hazard evaluation and data innovation to screen the estimation arrangement of diabetic retinopathy was
recurrence of diabetic retinopathy [12]. This created and researched utilizing information mining
investigation utilized a scientific calculation made method in which an early investigation of diabetic
utilizing epidemiological information on hazard retinopathy utilizing a precise and quick system gives
factors for diabetic retinopathy, through a site, the patient enough security treatment time. The
http://risk.is/, in which the calculation gets clinical shading fundus picture was utilized to consequently
information, including type and length of diabetes, identify and understand the different sores of diabetic
HbA1c or mean blood glucose, circulatory strain and retinopathy and its ordinary highlights, separately.
the nearness and grade of retinopathy. The AUC was The determinations of the typical shading fundus
0.76, and this number shows the model predicts the pictures were dissected and arranged by the
likelihood of a patient who creates locate extraction technique into ordinary or strange. The
compromising retinopathy (STR) 76% more right anomalous picture would then be able to be ordered
than who does not create STR. into three dimensions: Mild, moderate, and Severe.
To foresee the obscure class, an association standard
An investigation by Semeraro et al. anticipated and SVM classifier was utilized. The creator indicated
danger of diabetic retinopathy utilizing the c- promising outcomes to help the patients and the
measurement, survival beneficiary working quickening procedure.
trademark, and the Gonen and Heller concordance
likelihood gauge (CPE) for the Cox corresponding Abhilash Bhaisare et. al. [15] have proposed a
peril demonstrate [13]. For the inward approval, the framework to recognize the info picture as ordinary
C-file achieved an estimation of 0.746; the Gonen– or anomalous. At the point when the information
Heller CPE for the Cox relative risk technique was picture is discovered strange then examination for

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


268
Kajal Sanjay Kothare, Prof. Kalpana Malpe Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 266-272

further DR stages is finished. To recognize irregular of multi-display medoid based demonstrating


pictures there are different procedures and strategy approach in a gathering is exhibited to enhance the
utilized in picture mining. Picture mining is an precision of classification. The genuine
augmentation of information mining system. Microaneurysm districts are picked and characterized
Distinguishing proof of the unusual pictures is utilizing a hybrid classifier which is a weighted mix
finished utilizing preprocessing, including extraction of multivariate m-Mediods, GMM and SVM.
and classification calculations. The acquired outcome
is utilized to show the picture as ordinary or strange Vimala Balakrishnan et.al [18] have dealt with
and up to what degree. Integrating association guidelines and case-based
thinking to foresee retinopathy. They have proposed
K. R. Ananthapadmanaban and G. Parthiban [16] a retinopathy expectation framework dependent on
have utilized Naive Bayes and Support Vector information mining, especially association rules
Machine calculations to anticipate the early discovery utilizing Apriori calculation, and case-based thinking.
of eye ailment and DR. Utilizing Rapid Miner device The association rules are utilized to break down
they have assessed that Naive Bayes gives 83.37% examples in the informational index and to ascertain
precision and SVM gives 64.91% exactness. Execution retinopathy likelihood while case-based thinking is
of these strategies was additionally estimated by utilized to recover comparable cases. The creator
particularity as 95% and affectability as 96.65%. They trusted that extraordinary upgrades can be given to
began with a preprocessing task to enhance picture restorative professionals and furthermore to diabetics
quality by killing imperfections brought about by with the usage of this framework.
lighting and procurement forms. In the second step,
the optic plate has upset the programmed recognition. M.Tamilarasi and Dr.K.Duraiswamy [19] have
In the third step, the division of chart slices is utilized played out an overview on Automatic recognition of
so as to identify exudates areas. At long last, the non proliferative DR. Since there are numerous
neural system gave better outcomes with a information mining strategies and calculations that
component extraction of pictures by descriptors and analyze DR in retinal fundus pictures, in this paper,
Hu snapshot of GIST. The last outcomes were creators have checked on, characterized and looked
contrasted quantitatively and a manual exudates at the calculations and methods recently proposed so
division created by a specialist in ophthalmology. as to grow better and increasingly powerful
calculations.
M. Usman Akram, Shehzad Khalid, Shoab A.Khan
[17] have introduced a paper indicating three-arrange Ramon Casanova et.al.[20] concentrated on the
framework for early discovery of Microaneurysm utilization of Random Forest (RF) techniques to DR
utilizing channel banks. The framework extricates all classification investigation dependent on fundus
conceivable applicant districts for Microaneurysm photography information. The creators proposed a
present in the retinal picture. Contingent on a few way to deal with DR hazard evaluation dependent on
properties, similar to shape, shading, power, and metric got from reviewed fundus photography and
measurements an element vector for every area is fundamental information. they proposed that RF
framed to arrange a hopeful district as could be a profitable instrument to analyze DR
Microaneurysm or non-Microaneurysm. A crossover finding and assess its movement.
classifier which consolidates the Gaussian blend show
(GMM), bolster vector machine (SVM) and expansion

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


269
Kajal Sanjay Kothare, Prof. Kalpana Malpe Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 266-272

S.Sagar Imambi and T.Sudha [21] have connected Calculated relapse estimates the connection
content mining classification systems on the between the absolute ward variable and at least one
accessible clinical information to anticipate the autonomous factors by assessing probabilities. The
hazard factor of DR. They demonstrated that first expect a calculated capacity and the second a
moderately brief length of the case the executives standard typical dissemination work. The chances of
initiated before the beginning of clinically the reliant variable measuring up to a case are
recognizable retinopathy, essentially diminish the proportional to the exponential capacity of the direct
danger of creating retinopathy in patients with sort 2 relapse articulation. This shows how the rationale
diabetes. The objective was to build up an adaptable fills in as a connection work between the likelihood
and hearty clinical report classification framework and the direct relapse articulation. We additionally
that could be connected in extensive doctor's facility joined stepwise choice to choose discriminative
settings to encourage the doctors, with the goal that highlights in strategic relapse.
they can direct the patients effectively and diminish
the vision misfortune. SVM classifier is a machine learning calculation
proposed by Vapnik dependent on basic hazard
Hayrettin Evirgen, Menduh Çerkezi [22], have minimization rule of insights learning hypothesis. It
utilized Naive Bayes classification calculation to very well may be utilized to take care of classification
examine the genuine dataset so as to assembled and relapse issues. As a forecast of diabetic
prescient framework for DR. A sum of 385 diabetes retinopathy is a paired classification issue, SVM
patients' information was utilized to prepare the would be helpful for our motivation. During the time
expectation framework. With cross-approval creators spent model improvement, we utilize outspread
demonstrated that gullible Bayes calculation can be premise work (RBF) as the bit work in SVM.
utilized for diabetic retinopathy forecast with an
enhanced precision of 89%. An artificial neural Network is a group of
measurable learning models propelled by organic
III. MACHINE LEARNING ALGORITHMS TO neural systems and are utilized to evaluate or inexact
PREDICT DIABETIC RETINOPATHY capacities that can rely on countless and are
commonly obscure. ANNs are for the most part
DR expectation can be viewed as a multi-class exhibited as frameworks of interconnected neurons
classification issue. We consolidated choice trees, which send messages to one another. Positioned
calculated relapse, fake neural systems, and bolster forecast execution of choice trees dependent on 2-
vector machines to anticipate DR. The SAS year, 10-year, and a 15-year term of diabetes be
Enterprise Miner adaptation 12.1 programming of tuned dependent on experience, making neural nets
was utilized to produce expectation models. versatile to sources of info and fit for learning.

To help basic leadership forms, we utilized choice IV. CONCLUSIONS


trees to create interpretable guidelines for clinical
practice. We build choice trees and produce rules for Forecast and conclusion of Diabetic Retinopathy
clinical basic leadership, arranging dependent on rely upon the nearness of hemorrhages and
information gathering and unmitigated examination, microaneurysms in fundus pictures. There are
and producing choice trees as foreseeing models to numerous calculations which have been proposed
help with clinical basic leadership. and produced for the programmed recognition of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


270
Kajal Sanjay Kothare, Prof. Kalpana Malpe Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 266-272

diabetic retinopathy from highlight extraction. In angiographic and other retinal characteristics at
this paper condensed perspective of different baseline: ETDRS report no. 19", Early treatment
information mining methods is displayed which diabetic retinopathy study research group. Arch
demonstrates that KNN and SVM have given the best Ophthalmol. 1995; 113(9):1144–55.
correctnesses. This survey paper can go about as an [7] Chang TJ, Jiang YD, Chang CH, Chung CH, Yu
asset for the future scientists for the forecast of NC, Chuang LM. Accountability, utilization and
diabetic retinopathy utilizing information mining providers for diabetes management in Taiwan,
procedures. This will likewise be helpful for scientists 2000–2009: an analysis of the National Health
to get a diagram of this zone so as to grow Insurance database. J Formos Med Assoc.
increasingly proficient calculations. 2012;111(11):605–16. Tsao et al. BMC
Bioinformatics 2018, 19(Suppl 9):283 Page 120 of
V. REFERENCES 121
[8] Yau JW, Rogers SL, Kawasaki R, Lamoureux EL,
[1] Chew EY, Klein ML, Ferris FL 3rd, Remaley NA, Kowalski JW, Bek T, Chen SJ, Dekker JM,
Murphy RP, Chantry K, Hoogwerf BJ, Miller D., Fletcher A, Grauslund J, et al. Global prevalence
"Association of elevated serum lipid levels with and major risk factors of diabetic retinopathy.
retinal hard exudate in diabetic retinopathy. Diabetes Care. 2012; 35(3):556–64.
Early treatment diabetic retinopathy study [9] UK Pospective Diabetes Study Group. Tight
(ETDRS) report", 22. Arch Ophthalmol. 2016; blood pressure control and risk of macrovascular
114(9):1079–84. and microvascular complications in type 2
[2] American Diabetes Association. Standards of diabetes: UKPDS 38. UK Prospective Diabetes
medical care in diabetes–2014. Diabetes Care. Study Group. BMJ. 1998; 317(7160):703–13.
2014;37(Suppl 1):S14–80. [10] Kowall B, Rathmann W. HbA for diagnosis of
[3] Fong DS, Aiello L, Gardner TW, King GL, type 2 diabetes. Is there an optimal cut point to
Blankenship G, Cavallerano JD, Ferris FL 3rd, assess high risk of diabetes complications, and
Klein R, "American Diabetes A. Retinopathy in how well does the 6.5% cut-off perform?
diabetes. Diabetes Care", 2004;27(Suppl 1):S84–7. Diabetes Metab Syndr Obes. 2013; 6:477–91.
[4] Kempen JH, O'Colmain BJ, Leske MC, Haffner [11] Hosseini SM, Maracy MR, Amini M, Baradaran
SM, Klein R, Moss SE, Taylor HR, Hamman RF. HR. A risk score development for diabetic
"The prevalence of diabetic retinopathy among retinopathy screening in Isfahan-Iran. J Res Med
adults in the United States", Arch Ophthalmol. Sci. 2009;14(2):105–10.
2004; 122(4):552–63. [12] Aspelund T, Thornorisdottir O, Olafsdottir E,
[5] Huang YY, Lin KD, Jiang YD, Chang CH, Chung Gudmundsdottir A, Einarsdottir AB, Mehlsen J,
CH, Chuang LM, Tai TY, Ho LT, Shin SJ. Einarsson S, Palsson O, Einarsson G, Bek T, et al.
"Diabetes-related kidney, eye, and foot disease in Individual risk assessment and information
Taiwan: an analysis of the nation wide data for technology to optimise screening frequency for
2000-2009", J Formos Med Assoc. 2012; diabetic retinopathy. Diabetologia.
111(11):637–44. 2011;54(10):2525–32.
[6] Early Treatment Diabetic Retinopathy Study [13] Semeraro F, Parrinello G, Cancarini A, Pasquini
Research Group. "Focal Photocoagulation L, Zarra E, Cimino A, Cancarini G, Valentini U,
treatment of diabetic macular edema. Costagliola C. Predicting the risk of diabetic
Relationship of treatment effect to fluorescein

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


271
Kajal Sanjay Kothare, Prof. Kalpana Malpe Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 266-272

retinopathy in type 2 diabetic patients. J [21] S. Sagar Imambi and T. Sudha - Building
Diabetes Complicat. 2011;25(5):292–7. Classification System to Predict Risk factors of
[14] Dr. Karim Hashim Al-Saedi, Dr. Razi Jabur Al- Diabetic Retinopathy Using Text mining, (IJCSE)
Azawi, Rasha Asaad Kamil, - Design and International Journal on Computer Science and
Implementation System to Measure the Impact Engineering Vol. 02, No. 07, 2010, 2309-2312
of Diabetic Retinopathy Using Data Mining [22] Hayrettin Evirgen, Menduh Çerkezi- Prediction
Techniques, International Journal of Innovative and Diagnosis of Diabetic Retinopathy using
Research in Electronics and Communications Data Mining
(IJIREC) Volume 4, Issue 1, 2017, PP 1-6 Technique,ijaedu.ocerintjournals.org/tojsat/issue
[15] Abhilash Bhaisare, Sagar Lachure, Amol Bhagat, /22642/241874, Sep, 2014
Jaykumar Lachure - Diabetic Retinopathy
Diagnosis Using Image Mining, International Cite this article as :
Research Journal of Engineering and Technology
(IRJET), Volume: 03, Issue: 10, Oct -2016 Kajal Sanjay Kothare, Prof. Kalpana Malpe, "A
[16] K. R. Ananthapadmanaban and G. Parthiban. - Review on Prediction of Diabetic Retinopathy Using
Prediction of Chances - Diabetic Retinopathy Data Mining Algorithms", International Journal of
using Data Mining Classification Techniques. Scientific Research in Computer Science, Engineering
Indian Journal of Science and Technology, Vol and Information Technology (IJSRCSEIT), ISSN :
7(10), 1498–1503, October 2014 2456-3307, Volume 5 Issue 1, pp. 266-272, January-
[17] M. Usman Akram, Shehzad Khalid, Shoab A. February 2019.
Khan - Identification and classification of Journal URL : http://ijsrcseit.com/CSEIT195179
microaneurysms for early detection of diabetic
retinopathy, Pattern Recognition, Vol. 46, No. 1,
2013, pp. 107–116.
[18] Vimala Balakrishnan , Mohammad R. Shakouri
and Hooman Hoodeh - Integrating association
rules and case-based reasoning to predict
retinopathy, Maejo Int. J. Sci. Technol. 2012,
6(03), 334-343, ISSN 1905-7873
[19] M. Tamilarasi and Dr. K. Duraiswamy -A Survey
for Automatic Detection of Non- Proliferative
Diabetic Retinopathy, International Journal of
Innovative Research in Computer and
Communication Engineering, Vol. 2, Issue 1,
January 2014
[20] Ramon Casanova , Santiago Saldana, Emily Y.
Chew, Ronald P. Danis, Craig M. Greven,
Walter T. Ambrosius - Application of Random
Forests Methods to Diabetic Retinopathy
Classification Analyses, Published: June 18,
2014https://doi.org/10.1371/journal.pone.009858
7

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


272
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT195180

A Review on Predicting User Behaviour through Session Using


Web Mining
Veena Moharle1, Sonali Waghade1, Divyani Dhoke1, Sheetal Kolhe1, Snehal Narule1, Prof. Merajul Haque2
1BE Students, Department of Information Technology, JD College of Engineering and Management, Nagpur,
Maharashtra, India
2Assistant Professor, Department of Information Technology, JD College of Engineering and Management,
Nagpur, Maharashtra, India

ABSTRACT

Web usage mining is driving examination zone in Web Mining worried about the web client's behavior.
Weblog mining is one of the ongoing regions of research in Data mining. Web Usage Mining turns into an
imperative perspective in the present time because the amount of information is persistently expanding. We
manage the web server logs which keep up the historical backdrop of page demands. Web log records are the
documents, which contain total data about the clients, peruse exercises on the web server Web mining is the
utilization of information mining strategies to find designs from the World Wide Web. This paper gives a
consideration on Web usage mining to anticipate the behavior of web clients dependent on web server log
records. Clients utilizing web pages, a successive access way's and continuous access pages, joins are put away in
web server log records. Contingent on the recurrence of clients visiting each page mining is performed. By
finding the session of the client we can break down the client behavior when spending on a specific page. Web
log alongside the distinction of the client catches their perusing behavior on a website and talking about with
respect to the behavior from the investigation of various algorithms and diverse strategies.
Keywords : Knowledge Sharing, Web Portal, Job Portal, Online Recruitment.

I. INTRODUCTION between web pages through the web structure. Web


structure mining is again classified into hyper
Web Usage mining applies the information mining interface mining and inside structure mining. Web
method to extricate learning from weblog records usage mining examinations weblog records for
naturally. Web mining can be arranged into web discovering program examples of the client. Web
content mining, web structure mining and web usage usage mining is additionally called weblog mining.
mining. Web content mining is the way toward
separating data from an expansive measure of Understanding client get to examples can help a
gathered information. This strategy extricates the webmaster in tweaking the website content so as to
data from the substance of web pages. Web content expand the client's perusing knowledge. Web route
mining is again arranged into interactive media example might be portrayed by breaking down web
mining and web printed mining. Web structure server logs that contain nitty-gritty data of client
mining is the way toward dissecting the connections perusing exercises. An occasion in a weblog is made

CSEIT195180 | Received : 01 Feb 2019 | Accepted : 10 Feb 2019 | January-February -2019 [ 5 (1) : 273-277 ]
273
Veena Moharle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 273-277

out of IP address of the asking for customer, date and algorithms for playing out every one of these
time of the demand, asking for technique utilized, procedures [1].
status code, the asked for the record, and so forth. It
is conceivable to sort the arrangement of snap Anshul Bhargav and Munish Bhargav likewise the
exercises into sessions, and after that utilization them individuals from IEEE proposed edge work for web
for further investigation to get the predicted client usage mining. The edge work is to perform client’s
behaviour. Breaking down clients' Web watching characterization dependent on found examples and to
behaviours is one of the essential and testing research discover the qualities of clients. The casing work
subjects of Web usage mining. On the off chance that comprises of three principal steps: pre-processing,
clients' advantages can be naturally recognized from design disclosure and client arrangement. The pre-
clients' Web log information, they can be utilized for processing stage incorporates log document cleaning,
data suggestion and showcasing which are valuable client ID and session recognizable proof. The
for the two clients and Web webpage engineers. example disclosure process finds a new example from
Research for dissecting Web log information has the informational collection. In the client
been finished by numerous specialists in the field of arrangement process, every client is grouped by his
Web usage mining; finding incessant examples of log or her attributes. The groupings depend on nation-
information. The principle motivation behind the based arrangement, site section based order and
investigation of the client's visit record is to break access time-based characterization [2].
down the client's most worried about the outcomes
from the mining results. By breaking down the client S. S. Patil and H.P Khandagale proposed subjective
access to assets of the time, recurrence, etc., alter the client demonstrate which determine the foreseen
structure and plan of the site to anticipate that more usage behaviour dependent on the examples found in
clients should remain and better serve clients. Client past pre-processing, stage. The edge work
behaviour investigation has turned into another incorporates information cleaning, client
examination hotspot. recognizable proof, session distinguishing proof, way
finish, exchange ID and example revelation and
The work of this research mainly contemplates the extraction forms. Way culmination alludes to
weblog Mining technology in client behaviour creating rules for missing references dependent
investigation and assembles the client premium available structure, referrer, and other heuristic data.
model dependent on the client premium data, lastly By breaking down the found example, the behavior
draws the client's advantage. of the client is recognized and route refreshes are
given in the web page [3].
II. LITERATURE REVIEW
Virendra R. Rathod and Govind V. Patel proposed an
G. Neelima and Dr. Sireesha Rodda individuals from edge work dependent on FCM grouping and Markov
IEEE have proposed a thought for anticipating client show for foreseeing client behaviour utilizing weblog.
behavior through sessions utilizing weblog mining. Subsequent to pre-processing fluffy c implies (FCM)
They manage web server logs of NCSA normal log calculation is connected for example disclosure and
record design for mining. The proposed methodology examination. Markov demonstrate is utilized to next
comprises of information handling, client page expectation and better web page forecast
distinguishing proof and session recognizable proof precision. Fluffy c implies bunching calculation gives
procedures. They are utilizing three unique the preferred outcome over k-implies grouping.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


274
Veena Moharle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 273-277

Fluffy c implies bunching is a standout amongst the client's advantage dependent on the M5 display tree
most generally utilized fluffy grouping algorithms [4]. to break down the explanatory occasions.

In this exploration [5] propose the structure and In this examination [9], we present the "errand trail"
usage of a half and half framework by consolidating a to comprehend client seek behaviours. We
few information mining methods to catch the client's characterize an errand to be a nuclear client data
web perusing behavior. Client route sessions that need, while an assignment trail speaks to all client
speak to the association with a given website are exercises inside that specific undertaking, for
utilized to develop Hypertext Probability Grammar example, question reformulations, URL clicks. he
(HPG). Break down the HPG with the idea chain of considers investigations and correlations with assess
command show and apply the network bunching the viability of undertaking trails in a few inquiry
procedure to separate the usual portrayal of client applications: determining client fulfillment,
route behavior. anticipating client seek premiums, and
recommending related questions. Analyses on
This paper [6] portrays a technique for clearing up expansive scale datasets of a business internet
clients' interests dependent on an examination of the searcher demonstrate that:
site-catchphrase diagram. The technique is for (1) Task trail performs superior to anything session
removing sub-diagrams speaking to clients' primary and question trails in determining client
advantages from a website Keyword chart. It is fulfillment;
produced from increased Web crowd estimation (2) Task trail builds web page utilities of end clients
information (Weblog information). Our contrasting with session and inquiry trails;
commitments are 1) Web structure mining strategy (3) Task trails are equivalent to question trails yet
(PageRank) is appropriate for assessing the more touchy than session trails in estimating
characteristics of chose sub-diagrams and 2) two sorts distinctive positioning capacities;
of information (destinations and catchphrases) can be (4) Query terms from a similar assignment are more
spoken to and handled in a similar way utilizing a topically predictable to one another than
webpage catchphrase diagram. Our technique is basic question terms from various undertakings;
and it has capacities of separating sub-diagrams (5) Query proposal dependent on errand trail is a
identified with clients' principle advantages. Our decent supplement of question recommendations
work is required to give bits of knowledge for dependent on session trail and navigate
preparing boisterous heterogeneous chart structures. bipartitely.

This investigation of work [8] for the most part The discoveries in this paper check the need for
presents the technique for weblog mining, which can removing errand trails from web look logs and
find the method of web pages by burrowing web log improve applications in hunt and suggestion
records. By dissecting and investigating the standards frameworks.
of weblog records, we can recognize the potential
clients of the website and enhance the nature of data III.Discussion and Analysis
administrations to clients. In the phase of client
behaviour investigation, this examination investigates Research for breaking down Web log information has
the distinctions in client perusing behaviour in been finished by numerous analysts in the field of
various sorts of access occasions and figures the Web usage mining; finding regular examples of log

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


275
Veena Moharle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 273-277

information, demonstrating clients' route designs, At present, there is no standard meaning of session in
grouping clients of explicit Web webpage, and the International scholarly network. The meaning of
finding client networks. the session made by W3C Union is that it's a point of
confinement set of pages produced by clients' one or
A. Web Log Mining a few demand click. In the Weblog, the demand
records have a place with the clients who have
Web Usage Mining is the use of information mining distinctive IP addresses positively not in a similar
strategies to find intriguing usage designs from Web session. A client may peruse a similar site a few times
information to comprehend and better serve the and the time interim between each peruse
requirements of Web-based applications. Usage moderately expansive, at that point that is not in a
information catches the character or beginning of similar session. Session distinguishing proof's
Web clients alongside their perusing behaviour at a undertaking is to partition clients' peruse records into
Web webpage. a suitable session. We in every case initially do
distinctive presumptions on the client's perusing
behaviour on the customer, and after that distinguish
the session. These sessions can be utilized as a piece
of info information in grouping, bunching,
expectation and different errands. In light of a
uniform settled timeout, a conventional session
recognizable proof calculation is utilized.

Fig 1. Phases of Web Usage Mining IV.CONCLUSION

A server log is a log document (or a few records) Web usage mining is without a doubt one of the
consequently made and kept up by a server rising regions of research and imperative sub-space of
comprising of a rundown of exercises it performed. information mining and its systems. So as to take the
Web server logs are utilized to bunch web clients full favourable position of web usage mining and its
having comparative interests. It is likewise everything methods, it is essential to do pre-
characterized as adjusting administrations and data processing stage proficiently and viable. This paper
which was accessible on a website to the necessities endeavours to convey territories of pre-processing,
and the desires for an objective client, the dynamic including information purifying, session ID, client
client; the personalization undertaking by might recognizable proof. When the pre-processing stage is
profit by the information picked up from an all around performed, we can apply information
investigation of the client's navigational behaviour mining strategies like bunching, affiliation, order and
joined with different highlights which are curious to so on for utilization of web usage mining, for
a Web setting, to be specific its structure and content example, business insight, web-based business, e-
[10]. learning, personalization, and so on. Weblog mining
is one of the ongoing regions of research in Data
B. Session Identification mining. Web Usage Mining turns into an imperative
perspective in the present time in light of the fact
that the amount of information is ceaselessly
expanding. We manage the web server logs which

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


276
Veena Moharle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 273-277

keep up the historical backdrop of page demands Efficient Web Mining Algorithm for Web Log
Weblog record examination started with the reason Analysis: E-Web Miner ",1st International
to offer to Web webpage directors an approach to Conference on Recent Advances in Information
guarantee sufficient transfer speed and server ability Technology (RAIT) 2012
to their association. By examining these logs, it is [8] Xipei Luo, Jing Wang, Qiwei Shen, Jingyu
conceivable to find different sorts of learning, which Wang, Qi Qi ,"User Behavior Analysis Based on
can be connected behavior investigation of clients. User Interest by Web Log Mining ", 27th
International Telecommunication Networks
V. REFERENCES and Applications Conference (ITNAC),2017
[9] Zhen Liao, Yang Song, Yalou Huang, Li-wei
[1] G. Neelima , Dr. Sireesha Rodda, “Predicting He, Qi He,"Task Trail: An Effective
user behavior through sessions using the web Segmentation of User Search Behavior",IEEE
log mining”, Conference on Advances in Transactions on Knowledge and Data
Human Machine Interaction (HMI), R. L. Engineering,2014.
Jalappa Institute of Technology, Doddaballapur, [10] C., Ventura, S., Zafra, A., de Bra, P.: Applying
Bangalore, India, March 2016. Web usage mining for personalizing hyperlinks
[2] Anshul Bhargav , Munish Bhargav, “Pattern in Web-based adaptive educational systems
discovery and users classification through web (received January 8, 2009) (received in revised
usage mining”, International Conference on form May 4, 2009) (accepted May 4, 2009)
Control Instrumentation, Communication and
Computational Technologies, IEEE 2014.
[3] S. S. Patil, H. P. Khandagale, “Survey paper on
enhancing web navigation usability using web
usage mining techniques”, International journel
of modern trends inenginnering and
research.2016.
[4] Virendra R. Rathod and Govind V Patel,
“Prediction of user behaviour using web log
mining in web usage mining”, International
journel of computer application vol. 139- No. 8,
April 2016.
[5] . M. -S. Kao, T. Ozyer, R. Alhajj,"Hybrid
approach for predicting the behavior of Web
users", IEEE International Conference on
Information Reuse and Integration, Conf, 2005.
[6] Tsuyoshi Murata, Kota Saito,"Extracting Users'
Interests from Web Log Data ",
IEEE/WIC/ACM International Conference on
Web Intelligence (WI 2006 Main Conference
Proceedings) (WI'06)
[7] Mahendra Pratap Yadav, Pankaj Kumar
Keserwani, Shefalika Ghosh Samaddar,"An

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


277
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195169

Elimination of Impulse Noise using Mean Median filter for Image


Enhancement
Trupti Arun Jangale, Raj Kumar Paul
Department of Computer Science & Engineering, Vedica Institute of Technology, Bhopal, Madhya Pradesh,
India

ABSTRACT

In this method, we've got introduced a new technique for the improvement of gray scale images, when images
are corrupted by salt and pepper noise that's additionally referred to as an impulse noise. Our suggested
phenomena show a better output for Medium density impulse noise as compare to the opposite renowned
filters like standard Median Filter (SMF), a decision based mostly Median Filter (DBMF) and modified decision
based Median Filter (MDBMF), Nonlinear filter (NLF) and so on. Our projected technique worked on two steps,
within the beginning is that the detection of noisy pixels and within the second step is that the removal of
noisy pixels. For detection of noisy constituent apply condition pixels values lies in between 0 to 255 it noisy
it's noisy free pixels. In our second step that's the removal of noisy pixel recommended technique that's replaces
the noisy pixel by alpha trimmed mean median value. Different grayscale pictures are tested via proposed
technique. The experimental result shows higher Peak Signal to Noise ratio (PSNR) values and with higher
visual and human perception.
Keywords : Gray Scale, Impulse Noise, Trimmed Mean, Median, Unsymmetricness

I. INTRODUCTION degree economical de-noising technique to catch up


on such knowledge corruption.
Digital images play a very important role each in Image De-noising: -
existence applications like satellite television,
resonance imaging, laptop pictorial representation Image de-noising is a vital image process task, each as
also as in areas of analysis and technology like a method itself, and as a part in alternative processes.
geographical data systems and physical science. There ar some ways to de-noise a picture or a group
Knowledge sets collected by image sensors are of knowledge exists. the most properties of an honest
typically contaminated by noise. Imperfect image de-noising model ar that it'll take away noise
instruments, issues with the information acquisition whereas protective edges. historically, Image de-
method, and intrusive with natural phenomena will noising are often thought-about as a element of
all degrade the information of interest. what is more, process or as a method itself. within the 1st case, the
noise are often introduced by transmission errors and image de-noising is employed to enhance the
compression. Thus, de-noising is usually a necessary accuracy of varied image process algorithms like
and also the start to be taken before the image registration or segmentation. within the second case,
knowledge is analyzed. it's necessary to use associate

CSEIT195169 | Received : 01 Feb 2019 | Accepted : 10 Feb 2019 | January-February -2019 [ 5 (1) : 278-288 ]
278
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

the noise removal aims at rising the image quality for


visual examination.

The de-noising technique aims to attenuate noise


through 2 phases particularly,
1. Noise detection
2. Noise removal

Classification of De-noising Algorithm: -

As shown in Figure 2.2 below, there are two basic


approaches to image denoising, spatial filtering
Figure 1 (b): Introduction of noise into the clean
methods and transform domain filtering methods.
signal
• Spatial Filtering
• Non-Linear Filters
These embrace photographic grain noise, electrical
• Linear Filters
device noise, and transmission errors. Any globe
• Linear and nonlinear Filtering Approach
device is suffering from a particular degree of noise,
• Median Filter (M.F.)
whether or not it's thermal, electrical or otherwise.
• Mean Filter (M.F.)
Additive noise, in all probability the foremost
• LMS adaptative Filter
common kind, may be expressed as:
• Center Weighted Median Filter (CWM)
• Adaptative Mean Filters (AMF)
I (t) = S (t) + N (t)
• Progressive shift Median (PSM)
Where
• Switching Median Filter (SMF)
I (t) = ensuing knowledge measured at time t,
• Adaptative Weighted Median Filtering
• Adaptative Median Filter (AMF) S (t) = original signal measured, and
• Dision based Unsymmetrical Trimmed Mean N (t) = noise introduced by the atmosphere and
Filter (DBUTMF) different sources of interference.
• Modofied Dision primarily based
Unsymmetrical cut Mean Filter (MDBUTMF) kinds of noise
The types of Noise are following:-
Idea of noise:- I. Electronic equipment noise (Gaussian noise)
An image passes through many media and II. Impulse noise (Salt-and-pepper noise)
transmission channels before it reaches the process III.Shot noise (Poisson noise)
part. throughout this point, the image is also subject IV.Speckle noise
to interference from many sources. V.Brownian Noise

Image noise removal victimization Filters:-


Filtering whereas not Detection:-
this kind of filtering a window mask is affected across
the discovered image. The mask is usually of size
Figure 1 (a) : Introduction of noise into the system (2N+1)2, where N is also a positive range. usually the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


279
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

centre part is that the constituent of interest. once signal, the foremost obvious window is simply the
the mask is affected starting from the left-top corner primary few preceding and following entries,
of theimagetothe right-bottom corner, it performs whereas for 2nd (or higher-dimensional) signals like
some mathematics operations whereas not images, additional complicated window patterns are
discriminating any constituent. potential (such as "box" or "cross" patterns). Note that
if the window has associate degree odd range of
Detection followed by Filtering:- entries, then the median is easy to define: It's simply
This type of filtering involves two steps. In the center price on balance the entries within the
commencement it identifies screeching pixels and in window are sorted numerically. For an excellent
second step it filters those pixels. Here in addition a range of entries, there's quite one potential median.
mask is affected across the image and a number of
mathematics operations unit of measurement 1-D median Filter:-In 1D median filter, we have a
distributed to look at the screeching pixels. Then tendency to take into account the filtering window
filtering operation is performed exclusively on those slippy over a 1D array (either horizontal or vertical)
pixels that unit of measurement found to be of pixels. Assume the 5 pixels presently within the
screeching inside the previous step, keeping the non- windows are:
noisy intact.

Mean filter:-
Mean filtering may be a easy, intuitive and simple to Where the center picture element with price two
implement technique of smoothing images, i.e. hundred is associate isolated picture element and it's
reducing the quantity of intensity variation between out-of-range (noise). The median of those 5 values is
one pel and also the next. It's usually accustomed cut found by sorting the values (in either ascending or
back noise in images. The thought of mean filtering is falling order). the center price is that the median:
just to interchange every pel price in a picture with
the mean (`average') price of its neighbours, as well The original picture element price two hundred is
as itself. This has the impact of eliminating pel values replaced by the median a hundred and ten.
that square measure atypical of their surroundings.
Thearithmeticmean filter may be a terribly easy one 2-D median Filter:
and is calculated as follows: Replace every pel by the
typical of pels during a sq. window close this pixel

Median filter:-
The median filter is often accustomed scale back
noise in a picture, somewhat just like the mean filter.
However, it usually will a much better job than the
mean filter of protective helpful detail within the
image. the most plan of the median filter is to run Neighbouring pixel values = 115,119,120,123,
through the signal entry by entry, commutation 124,125,126,127,150
every entry with the median of close entries. The Median value = 124.
pattern of neighbours is named the "window", that Figure 2 (a) : Selection of filtering window of 2-D
slides, entry by entry, over the whole signal. For 1D

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


280
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

Input a 2-D Image (x by y) process simultaneously. In the suggested method


employ the smallest window dimension which
conserve the very well information of image.We
Preallocate another metrix of size (x+2 divide these pixels into matrix form and called it
by y+2) with zeroes window. Then we analyse every pixel by sliding that
window over the entire image. The window of size
Copy the input marix into the 3x3 choose for noise recognition and noise
Preallocate matrix elimination. The windowpane contains total 9
elements which are as follows:
Form a window matrix of size 3 x3
with the element of input matrix Z1, Z2, Z3, Z4, Z5, Z6, Z7, Z8, Z9.

Copy the window matrix (3 by 3) into Table 1 Filtering window of size 3x3
an array and sort it. Column 1 Column 2 Column 3
Row 1 Z1 Z2 Z3
Row 1 Z4 Z5 Z6
Find the median element. Here 5th
element (total elements are 9, middle Row 3 Z7 Z8 Z9
element will be 5).
Different blocks are shown for different steps. At the
starting there is filtering window of size 3x3 has been
Place 5th element into the output selected and calculated gray value of pixels according
matrix. Do the process for the to the strength of intensity of every pixel.
complete input matrix ALGORITHM

Convert the image into an image of 0- Step-1:-


255 colour range type.
In the first step select a image for image processing
than apply pre-processing task in this image byusing
Display the image without noise
pre-defined command in MATLAB. The prepressing
Figure 2 : Flow chart o 2-D Median Filter task are first convert our target image into a gray
scale level. Second convert image into a strand image
Another concern within the application of the that is 255X255. Now selected image is prepared for
median filter is that the form of the neighbourhood dealing out. For the pre doling out assignment and all
mask. A square-shaped mask will erode the corners of other things are implemented in MATLAB. In matrix
rectangular objects, whereas a cross-shaped lab. A large no. function file available for the
neighbourhood mask can leave corners intact. performing the initial task in the digital image
processing. Also done the padding of zeros and ones
II. PROPOSED METHOD in image.

We have develop the trouble-free algorithm in which Select a target image - [x] = imread(‘Test image.jpeg’)/
we execute the noise finding and noise removal for the read of image

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


281
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

Step-2:- Condition 1- If Pixels are between (0) Zero and two


In the second step apply impulse noise or can say salt five five (255) then they are noise free and shift to
and pepper noise in this image with the help Matrix reinstatement digital image pixels.
laboratory function. In this step apply noise with the
help of add-noise function in on the image. Condition 2- digital images pixel values are not
between in the range then they are moved to step 3.
[y] =imnoise(x,’Salt and Pepper’, Present of noise)
Step 3: In the step third check the window. Here two
Step-3:- cases are arise
In the third step for the preservation of corner, in 1. In a small 3X3 window contain all elements are
this step preserve the corner we the help of zeros and zeros (0’s) and two fifty five (255’s). In this
ones padding in the boundary side of the image. This condition apply mean filter. Take a mean of all
is use for the protection of corner of the digital image pixel in the small 3x3 windows all element aspect
pixels. centerpixel , and replace this value mean value by
the targeted pixel.
Step-4:-
If
W(ij) = [0] & [255] condition true
Noise elimination stage. In this stage first we divide a
{Mean (W (ij)) all pixels]=Replace by W(ij)
whole image into a small 3X3 window. Now relate
(Centervalue)
the trimming situation. This condition isknown
else
asnoise identification stage. In this stage check the
3x3 window pixel values combination of 0’s and
pixel values.In which pixels are stuck between 0 to
255’s. Now move to the case 2
255 ranges or not. Here two cases are generating.
end
If
2. In case 2, In a small 3x3 window pixels are the
X(ij) = 0<Y(ij)<255 / condition true Pixels Noise
combination of zeros and two fifty five. At this
Free
condition apply median filter for this case. Take a
else
median of all zeros and two fifty five and replace
X(ij) ≠ 0<Y(ij)<255 / condition Pixels are Noise
this value by centre pixel value.
End
If
Where X(ij) is the image small 3x3 windows pixels W(ij) = [0’s & 255’s ] / Combination in a 3X3
values.And Y(i,j) is the center pixel of targeted pixel window
of the small 3X3 window. {Median(W (ij) of all pixels}=Replace by W (i,j)’s
Center pixel value
Table 2 3x3 windows pixels value end
Step 4: Repeat all steps, all pixels in the whole
Column 1 Column 2 Column 3 255X255 image are 65,536 pixels are not
Row 1 Z1 Z2 Z3 processed.Hence a better de-noised image is obtained
Row 1 Z4 Z5 Z6 with improved PSNR MSE MAE and also shows a
Row 3 Z7 Z8 Z9 better image improvement. Therefore the overall
procedure of Modified unsymmetric trimmed median

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


282
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

Mean Filtering (MUTMMF), will next part of this the image, Y means the original image, 𝑌̂ shows the
report. Shows the result and simulation. de-noised image.
Graphic User Interface (GUI) in MATLAB:-
III. RESULT
In this preceding section, the image Tool GUI and
Mean Square Error (MSE): Measure GUI (Graphic User Interface) will be
introduced. Both of them were developed with the
The mean square error (MSE) dealings the value Matlab tool GUIDE (GUI Design Environment). In
modify or can say the changes to pixel intensity this GUI part shows the whole phenomena of
values between the original image (Xi)and de-noised proposed work like that image selection, impulse
image (Yj) and is given by noise addition, removal of noise and edges
preservation of the images in this research, in a user-
1 𝑁−1
𝑀𝑆𝐸 = ∑ (𝑋𝐽 ̅𝑗 )2
−𝑌 (5.1) friendly environment. First of all, the images are
𝑁 𝑗−0
selected by clicking the browse and Select Images. A
The MSE has been widely used to quantify image browser will appear afterwards to allow the user to
quality and when used alone, it does not correlate select the images, which can be any of the following
strongly sufficient with perceptual excellence. It file formats:
should be use therefore, with other quality check
parameter like SNR, PSNR, RMSEand visual Layout implementation of image enhancement based
perception. fusion techniques is done by following steps.

Peak signal-to-noise ratio (PSNR): ▪ First create simple Graphical User Interface (GUI)
window.
The PSNR is computed as - ▪ Second add impulse noise with the help of
𝑆2 imnoise command in this GUI.
PSNR=10log10 (𝑀𝑆𝐸 ) (5.2)
▪ Third step apply our proposed method that is
removal of noise.
Where 𝑆 is the maximum intensity in the original
▪ Fourth stepcalculated the image enhancement
image. The pear signal to noise ratio is higher for a
parameters like peak signal to noise ratio, mean
good resultant image and poorer for a reduced quality
square error, mean absolute error.
image. It calculate image reliability, that is, nearer
▪ Fifth step shows edges preservation of tested
the transformed image resembles the original
image.
image.In this proposed work on the source of our
▪ Sixth step graphical representation of PSNR and
image dimension 255×255, we mentioned peak signal
MSE values of the images. There are two
to noise ration and mean square error are as follows:
important push button design in this GUI these
are RESET and END.
The PSNR is expressed as:
(255)2
𝑃𝑆𝑁𝑅 = 10 𝑙𝑜𝑔10 (5.3)
(𝑀𝑆𝐸)
Where MSE acronym of Mean Square Error stands
for image enhancement factor, 𝑚 × 𝑛 is the size of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


283
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

Graphical representation of result of our proposed


method is shown in the blow the graph. Here in this
plot we have compare with different algorithms they
are - median filter,adaptive median filter, decision
based algorithms, modified decision based
unsymmetric trimmed median filter, Non-linear
Filter algorithm. This method is tested on the ‘Lena’
image of standard sizeshown plotsof PSNR values of
different noise. Our proposed method provide a good
result in low and medium noise density. Here
compression with different filters against noise
Fig. 3 – Shows the GUI of Proposed Method densities for Lena image is shown in Figure .

TABLE 4 CALCULATED PSNR OF PROPOSED In the figure 5 put side by side our proposed method
ALGORITHM FOR LENA IMAGE visual results compare with different algorithms they
are - median filter,adaptive median filter, decision
based algorithms, modified decision based
unsymmetric trimmed median filter, Non-linear
Filter algorithm. This method is tested on the ‘Lena’
image of size 256×256 in the 80% noise density ‘Lena’
image result with our planned algorithm. Shown
plotsof PSNR values of different noise. Our proposed
method provide a good result in low and medium
noise density.

(a)Original Image (b) 70% Noise (c) MF (d) AMF (e) PSMF (f) DBA

(g)MDBA (h )MDBUTMF (i) Proposed

Figure 4: Proposed outcome for ‘Lena’ test image

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


284
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

Table 5 Different values of Barbra image


Result of BARBRA Image
Noise Density PSNR MSE(dB) MAE (dB) Time (s)
3% 44.9155 2.0967 0.1486 2.24
5% 42.3732 3.765 0.2541 2.67
10% 38.5745 9.0288 0.5632 2.97
20% 35.1368 19.9249 1.1749 4.01
30% 32.5039 36.5334 1.9553 4.82
40% 30.3439 60.0745 2.8946 6.03
50% 28.7536 86.641 3.9566 6.17
60% 26.5334 144.4569 5.582 7.17
70% 24.1005 252.9446 7.8463 8.15
80% 20.8724 531.9078 12.599 9.63
90% 16.9621 1308.8 23.8651 11.49

Original Image Noise Image 10% Restored Image

Original Image Noise Image 20% Restored Image

Original Noise Image Restored


Image 30% Image

Figure-6 Shows the outcome of proposed method for ‘girl’ image

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


285
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

Table 6 different values of Girl image

Result of GIRL Image


Noise Density PSNR MSE(dB) MAE (dB) Time (s)
3% 44.2093 2.4669 0.1751 2.5
5% 43.2871 3.0505 0.2328 2.5
10% 40.7521 5.4685 0.3715 2.99
20% 37.2585 12.2244 0.7707 3.83
30% 35.1066 20.0642 1.1863 4.67
40% 32.5532 36.1213 1.7769 5.66
50% 30.2181 61.8405 2.5712 6.65
60% 28.1567 99.406 3.6035 7.6
70% 25.1084 200.5603 5.5773 8.89
80% 21.8142 428.2142 9.117 9.55
90% 17.5649 1139.2 18.7688 9.26

IV. CONCLUSION MSE and PSNR. Indifference to mean filter and


further accessible algorithms, the new algorithm
The new proposed algorithm has been introduced exploits a tiny 3x3 window containing only eight
to manage the problems, namely, poor image neighbors of the corrupted pixel that have
enhancement or we can say enhancement at high advanced bond this provides a good edge
noise density that is frequently enhanced in the protection , also better edge protection as well as
mean absolute error (MAE), mean square error more better person and illustration prescription.
(MSE) and peak signal to noise ratio. In this The proposed filter demonstrates trustworthy
proposed workModified unsymmetric trimmed and conventional performance among various
median Mean Filtering (MUTMMF) is used for range of noise densities increasing from 10%-
enhancing both the peak signal to noise ratio. 90%. The performance of the introduced
The performances of proposed Modified methodology has been tested at low and medium
unsymmetric trimmed median Mean Filtering densities on different level. In case of high noise
(MUTMMF) are quantitatively vies as well as the density levels the new proposed algorithm gives
individual perception show better outcome as better presentation as evaluate with other
compare last decade algorithms. Results reveal presented noise removal algorithms.
that the proposed filter exhibits improved
performance in comparison by means of Median V. REFERENCES
Filter (MF) [5], Adaptive Median Filter (AMF) [7],
Progressive Switching Median Filter (PSMF) [1] Digital Image Processing, Rafael C. Gonzalez,
[11],Decision Based Algorithms (DBA) [16], Richard E. Woods, Prentice Hall, Upper
Modified Decision Based Algorithms (MDBA), Saddle River, New Jersey 07458.
Modified Decision Based Unsymmetric Trimmed
Median Filter [29], Modified Non-Linear Filter
Algorithm (MNF)in terms of improved MAE,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


286
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

[2] Fundamentals of Digital Image Processing, S. [12] A. Ben Hamza, P. Luque, J. Martinez, and R.
Annadurai, R. Shammugalakshmi, Pearson Roman, “Removing noise and preserving
Education India. details with relaxed median filters,” J. Math.
[3] Castleman Kenneth R, Digital Image Imag. Vision, vol. 11, no. 2, pp. 161–177, Oct.
Processing, Prentice Hall, NewJersey, 1979 1999.
[4] T.A. Nodes and N.C. Gallagher, Jr., “The [13] V. Strela. “Denoising via block Wiener
output distribution of median type filters,” filtering in wavelet domain”. In 3rd European
IEEE Trans. Communication., 32(5): 532-541, Congress of Mathematics, Barcelona, July
1984. 2000. BirkhäuserVerlag.
[5] A.K.Jain,Fundamentals of digital image [14] S. Zhang and M. A. Karim, “A new impulse
processing. Prentice-Hall, 1989. detector for switching median filters,” IEEE
[6] David L. Donoho and Iain M. Johnstone., Signal Process. Lett., vol. 9, no. 11, pp. 360–
“Adapting to unknown smoothness via 363, Nov. 2002.
wavelet shrinkage”, Journal of the American [15] T. Yahagi. "A Random-valued Impulse
Statistical Association, vol.90, no432, NoiseDetector Using Level Detection", 2005
pp.1200-1224, December 1995. National IEEEInternational Symposium on Circuits
Laboratory, July 27, 2001. and Systems, 2005.
[7] H. Hwang and R. A. Hadded, “Adaptive [16] K. S. Srinivasan. "A New Fast and Efficient
median filter: New algorithms and results,” Decision-Based Algorithm for Removal of
IEEE Trans. Image Process., vol. 4, no. 4, pp. High-Density Impulse Noises", IEEE Signal
499–502, Apr. 1995. Processing Letters, 3/2007
[8] E. Abreu and S.K. Mztra,” A Signal- [17] JafarRamadhan Mohammed. "An Improved
Dependent Rank Ordered Mean (Sd-Rom) MedianFilter Based on Efficient Noise
Filter –ANew Approach For Removal Of Detection for HighQuality Image
Impulses From Highly Corrupted Images”, 0- Restoration", 2008 Second Asia
7803-2431- 5/95$ 4.000 1995 IEEE. InternationalConference on Modelling &
[9] R. Yang, L. Yin, M. Gabbouj, J. Astola, and Y. Simulation (AMS).
Neuvo, “Optimal weighted median filters [18] V.R. Vijaykumar. "Detail preserving median
understructural constraints,” IEEE Trans. basedfilter for impulse noise removal in
Signal Processing vol. 43, pp. 591–604, Mar. digital images",2008 9th International
1995. Conference on Signal Processing.
[10] Scott E Umbaugh, Computer Vision and [19] S. Saudia. "Salt & pepper impulse detection
Image Processing, Prentice HallPTR, New andmedian based regularization using
Jersey, 1998. Adaptive Median Filter", TENCON 2008 -
[11] Zhou Wang and David Zhang, “Progressive 2008 IEEE Region 10 Conference, 11/2008
Switching Median Filter for the Removal of [20] DENG Xiuqin, XIONG Yong PENG Hong , A
Impulse Noise from Highly Corrupted new kind of weighted median filtering
Images", IEEE Transactions On Circuits And algorithm used for image Processing,
Systems—Ii: Analog And Digital Signal International Symposium on Information
Processing, Vol. 46, No. 1, January 1999. Science and Engieering, 2008.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


287
Trupti Arun Jangale, Raj Kumar Paul Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 278-288

[21] Ashraf Aboshosha. "Image denoising based on


spatialfilters, an analytical study", 2009
InternationalConference on Computer
Engineering & Systems, 12/2009.
[22] V.R.Vijaykumar, P.T.Vanathi,
P.Kanagasabapathy and D.Ebenezer, “Robust
Statistics Based Algorithm to Remove Salt
and Pepper Noise in Images”, International
Journal of Information and Communication
Engineering 5:3 2009.

Cite this article as :

Trupti Arun Jangale, Raj Kumar Paul,


"Elimination of Impulse Noise using Mean Median
filter for Image Enhancement", International
Journal of Scientific Research in Computer
Science, Engineering and Information Technology
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue
1, pp. 278-288, January-February 2019. Available
at doi : https://doi.org/10.32628/CSEIT195169
Journal URL : http://ijsrcseit.com/CSEIT195169

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


288
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195145

E - Tax Information System


M. Florence Dayana*1, B. Vinithamani 2
1 Head & Assistant Professor, Department of Computer Applications, Bon Secours College for Women,
Thanjavur, Tamil Nadu, India
2M. Sc., Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India

ABSTRACT
For that purpose, we are developing this application, which can collect the tax information from different
authorities based on the house no’s and owner name by using Web Services and stores the data in our own
database. For getting the information, regarding taxes citizens need to register first. The bill is calculating on
different modes like annual, half-yearly, quarterly etc. By using the user, id and password citizens can log into
the system and collect the necessary information from the home page. For this every user need to provide his
house no and owner name. By clicking on the search button, they will get their own.
Keywords: E- TAX, Web Services

I. INTRODUCTION web service is only capable to get the information


from other sources.
Local bodies like municipalities need collect different
kind of taxes from the citizens. Every citizen needs II. METHODS AND MATERIAL
to pay Water Bills and Electricity Bills to the
particular authorities. Actually, a citizen wants to Local bodies are classified into Municipal
pay his municipal tax means he should go to Corporations, Municipalities of different grades
municipal office and collect the necessary bill, pay depending upon the population and income of the
there only. Same as water and electricity bills. This Urban Local Body. These Urban Local Bodies have
is not an easy job to get all this tax information in been constituted with the objective of discharging
different authorities now days. By using the user id certain.
and password citizens can log into the system and Obligatory functions like
collect the necessary information from the home Supply of Drinking Water
page. For this every user need to provide his house Providing Water connection to houses
no and owner name. By clicking on the search Sending water bills to houses
button, they will get their own Public Street Lighting
• House Details Calculating and sending Electric Bills
• Electricity Bill Details Maintaining sanitation and hygiene of public places
• Water Bill Details Municipal Library maintenance.
• Municipality Bill Details Calculate and sending Municipal tax and house taxes.
Web services play vital role here. Why because this The development of this new system contains the
system get the information from other sources. A following activities, which try to automate the entire
process keeping in the view of database integration

CSEIT195145 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 289-291 ]
289
M. Florence Dayana, B. Vinithamani. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 289-291

approach. User Friendliness is provided in the PROPOSED METHODOLOGY


application with various controls provided by system To debug the existing system, remove procedures
Rich User Interface. The system makes the overall those cause data redundancy, make navigational
project management much easier and flexible. It can sequence proper. To provide information about users
be accessed over the Intranet. The employee on different level and to reflect the current work
information can be stored in centralized database status depending on organization. To build strong
which can be maintained by the system. This can password mechanism
give the good security for user information because
data is not in client machine. Authentication is IV. RESULTS AND DISCUSSION
provided for this application only registered Users
can access. There is no risk of data management at
any level while the project development is under
process. The automated system will provide to the
employees for reliable services. The proposed system
using web services, a web service can get the
information from other sources also. There no burden
of calculating water, electric, house tax bills in this Administrator
system, by sending a request to the particular Citizen’s
authority via web service, we can get all the Web Services
information regarding these issues. Citizen’s Bill Information
Search
III. EXISTING PROCESS Reports
Authentication
The existing system is a manual system. Here the
employees needs to save the information in the form Administration
of excel sheets or Disk Drives. There is no sharing is
possible if the data is in the form of paper or Disk Administration is the main person of this system. He
drives. The manual system gives us very less security can have all the privileges to do anything in this
for saving data; some data may be lost due to is system. Admin can get the information regarding
management. It’s a limited system and fewer users citizen’s water bills, electric bills, and municipal taxes
friendly. Searching of particular information is very from various sources. For that administrator need to
critical it takes lot of time. Gathering information of send a request for that particular authority throw
different sources is not an easy job, data will be web services.
mismanaged. Calculating different bills manually
going to be as a process of mistake. There is no Citizens
interface to provide various bills in the existing People who are living in that municipality area are
system. In the existing system citizen get the bills called as citizens. Citizens need to pay water bill,
information manually, in various authorities like electricity bill, property tax for a certain of intervals.
water board, electricity department, property tax By using this system interface citizens know their
department etc. different bill amounts, pending bills, due dates etc.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


290
M. Florence Dayana, B. Vinithamani. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 289-291

Web Services taxpayers who are using the e-tax system. With these
Citizen Bill Information limitations the result of our study gives an insight to
The system interface is going to show the Bill e-tax quality evaluation criteria in an emerging
information of different authorities. The bills mainly economy like Nigeria and also provide the basis for
going to these types further research in these areas.
Water Bills
Electricity Bills VII. REFERENCES
Property Taxes
These interfaces also show the [1]. Productivity of the Nigerian Tax System: 1970-
Pending Bills 1990 AERC Research Paper 67. Nairobi: African
Penalties Economic Research Consortium. Black. (1999).
Due Dates Black’s Law Dictionary (Seventh Edition). West
Group. Centers for Medicare and Medicaid
V. CONCLUSION Services Office of Information Service. (2008,
March 27).
[2]. Selecting a development approach. United States
It has been a great pleasure for me to work on this
Department of Health and Human Services (HHS).
exciting and challenging project. This project proved
Clarke, J. M. (1998).
good for me as it provided practical knowledge of not
[3]. Automated test generation from a Behavioral
only programming in ASP.NET andC#.NET web-
Model. Pacific Northwest Software Quality
based application and no some extent Windows
Conference. IEEE Press. Federal Republic of
Application and SQL Server, but also about all
Nigeria. (1997).
handling procedure. It also provides knowledge about
[4]. Report of the Vision 2010 Committee. Abuja,
the latest technology used in developing web enabled Nigeria. Federal Republic of Nigeria. (2002).
application and client server technology that will be [5]. Personal Income Tax Act 1993(1993 No. 104):
great demand in future. This will provide better [6]. Operation of Pay as You Earn (PAYE) Regulations
opportunities and guidance in future in developing 2002 Official Gazette. Federal Ministry of Finance
projects independently. Abuja. Geoffrey, E. (2004).
[7]. Global Business Information Technology: An
VI. FUTURE ENHANCEMENT Integrated Systems Approach. Pearson Education.
Grady. (2005).
This paper examines the taxpayers’ e-tax system [8]. Unified Modelling Language User Guide the
implementation by validating the psychometric Second Edition. Addison-Wesley. International
properties and using the instrument validated for e- Organization for Standardization. (2015, February
tax service evaluation criteria in the context of 26).
Nigeria. It is concluded that all the six variables,
Cite this article as :
compatibility, complexity, reliability, affordability, M. Florence Dayana, B. Vinithamani, "E - Tax
ease of use and tangible are important in the Information System", International Journal of Scientific
evaluation of e-tax system implementation. The items Research in Computer Science, Engineering and
used should enable the tax authority to bench mark Information Technology (IJSRCSEIT), ISSN : 2456-3307,
their routine and plan for enhancement. The first Volume 5 Issue 1, pp. 289-291, January-February 2019.
limitation of these study is small sample size and the Available at doi : https://doi.org/10.32628/CSEIT195145
population been limited to only self-employed Journal URL : http://ijsrcseit.com/CSEIT195145

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


291
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195154

Children Tracking Classification Control Using Android


M. Juno Isabel Susinthra*1, S. Praveena 2
1Department of Computer Applications, Bon Secours College for Women, Thanjavur, Tamil Nadu, India
2M. Sc., Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India

ABSTRACT

Children tracking system based on android terminals is proposed. Recently, all over the world crime against the
children in the age of 14 to 17 years is more popular. Parent’s always worry about their children whenever they
are outside from the home. In this paper, the proposed system consists of two sides out of them one is parent
module and another is the child module. The child module consists of ARM7 Microcontroller (LPC 2148), GPS
(Global Positioning System), GSM (Global System for Mobile Communication) and voice chip where the parent
module consists of android mobile phone. This paper gives the information about missing child from school
campus. There are two android mobile phones for the safety of the both modules. The system tracking the child
from source to destination i.e. from home to school or anywhere.
Keywords : ARM Based Children Tracking System, arm7microcontroller, GSM, GPS and Android Mobile
Phone.

I. INTRODUCTION services of a smart phone which is GSM more


specifically SMS in one system.
Children tracking system is widely used all over the
world to assure their parent‘s that their wards are safe II. METHODS AND MATERIAL
from suspicious actions in today‘s world over 80% of
the world population in the age of seven or eight LITERATURE SURVEY
owns smart phone. The proposed system includes
tracking the child movements outside from the home. Monitoring mobile devices has been a concern for
This is due to reason one of them is the remarkable not only individual users, but also for organizations,
features and capabilities that new smart phone offers communities, and scientific researchers. The
especially android based smart phones. GPS offers challenge of using monitoring applications is about
outstanding capabilities in locating positions and this how the collected data will be handled and the how
can be used to developed resourceful application that to build a macroscopic pattern from the collected
helps in locating missing or lost children. This project data. This part of the survey discussed the summary
is designed to be used by parents and aimed to help of the selected research papers related to monitoring
locating missing or lost children. It takes advantages mobile devices. In, the authors discussed the
of the fact that many of today ‘s children bring smart importance of having mobile phone sensing to
phone which is convenient for this kind of situation. develop micro and macro views of mobile usage. This
In this paper GPS is combined with one of the basic importance increases as the mobile devices have
become the main computing and communication

CSEIT195154 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 292-295 ]
292
M. Juno Isabel Susinthra, S. Praveena Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 292-295

devices. Therefore, research areas on data collection, among different apps, which increase the difficulty in
analysis and patterns learning are becoming rich monitor single flow between an app and the
discussion area, especially with all the cheap features operating system. The proposed platform only tracks
and technological advances available on phones data flows but not control flows. Data flows are the
allowing for a variety of users experiences and usage. explicit flowing of data through the apps. Control
The framework architecture proposed in is built upon flows are the implicit flowing of data that require
using mobile phone sensing and the cloud computing. analytical analysis.
The paper discussed the different challenges with
gathering data techniques using mobile and EXISTING PROCESS
unexpected environment (mobile context), as well as,
the privacy issues on gathering personal information Android terminals have wireless LAN and Bluetooth
from a third-party application. The study discussed device. It results in lack of individual attention
the three different scales mobile sensing considered towards the children. It offers less security. Studies
beneficial on: individual, group, and community data conducted by Cyber Travel Tips, showed that in
gathering. The scalability feature of gathering Malaysia, missing children are basically classified into
individual behaviour to learn from community two categories. The first category is disappearance,
patterns may help improve communities in social, which includes running away from home. Children
healthy and environmental studies. The study also tracking system is also developed on mobile ad hoc
explained the two sensing paradigms: participatory networks. System developed in says that in GPS
sensing, which involves users who lunch sensors system and tag-based system, each parent cannot
manually, and opportunistic sensing, where sensors obtain group information on the vicinity of the child.
automatically collect data. The main focus of the A self -configurable new generation children tracking
introduced work in is about understanding how system. Hiroshima City Children Tracking System is
personal data travels through selected third-party a safety support system for children based on ad hoc
applications. The authors raised the issue about network technologies. Field experiments have been
violating privacy barriers when applications access conducted in cooperation with an elementary school
the personal information by only ask for users’ in Hiroshima. In this paper, propose a new
permissions to access the data with no explanation of generation child tracking system which is based on
how the data will be used. The research proposed a experiences and findings of the field experiments for
monitoring platform based on Android mobile phone Hiroshima children tracking system. Existing
to track the data flow through 30 selected technologies, however, are not powerful to prevent
applications. After testing the proposed the platform, crimes against children and helpful parent ‘s since it
the research concluded that 20 apps miss used the is difficult to take information of children as a group.
users’ private data and 15 apps used users’ locations to
support marketing services. The main challenges in PROPOSED METHODOLOGY
monitoring the use of mobile apps include the
shortage of Smartphone used data as a resource The children information is transmitted and received
constrained, as well as, the dynamic nature of data using GSM technology. The child module acts as a
flow using mobile devices. This dynamicity causes transmitter which includes ARM7 microcontroller
another context-based challenge because data could (lpc 2148), GSM module, GPS module and voice chip
be sent at any time and any place. Further, mobile as well as android mobile phone. The receiver
apps allow for across sharing data and information module includes Android mobile phone and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


293
M. Juno Isabel Susinthra, S. Praveena Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 292-295

monitoring database. The position of the moving defining the location of their children (indoor or
child is tracked by is tracked by GPS and is sent to outdoor) and to give the safety distance for each child.
ARM7 microcontroller. This controller forwards the
GPS data (latitude and longitude) to GSM board. Detection of Safety Distance When parent starts the
GSM will in turn send the position of the moving system, she/he should select where they stay; indoor
child to two receivers. It allows the parent to get or outdoor; in order to calculate the distance between
their child ‘s location. them and each child. Indoor option limit child range
to 3m by reducing child transmitter source so signal
III. RESULTS AND DISCUSSION attenuated will not be detected. But if parent selects
outdoor option, distance will be calculated by taking
IMPLEMENTATION RESULTS child and parent locations from GPS chips. Child
module will send its GPS reading to parent module
which calculates distance and compares it with
distance threshold; this threshold value should be
entered by parent after selecting outdoor option.

IV.CONCLUSION

All parent wishes to shield their children from real


dangerous they will inevitably undergo hard times.
Child Module The child module is attached to child. Normal behaviour of any child depends on the child's
Its primary role is to periodically receive messages age, personality, and at the end the physical and
and in response send messages to the parent module emotional development. A child's behaviour may be a
and alert them if the child is in danger. The child problem if it doesn't match the expectations of the
module also has a buzzer alarm that sounds whenever family or if it is disruptive. So Parents may face many
the child is in an alarmed state. This allows a parent problems with their children, when they let them
to more easily locate the troubled child. The sensors free without any observation especially if they have
that used with this module are described below. many children that may face many dangers.

Parent Module The parent module is the V. FUTURE ENHANCEMENT


computational brain of the entire system. It is
responsible for communication with all the children There are many tracking system available in present
and calculating their distance. It also analyzes time, these systems use different technologies but
incoming packets to see if any child has been in an each of them has one or more limitation such as not
alarmed state or not. Finally, if any child stops suitable for many children or monitor how far the
sending packets, the parent assumes it is out of range child from their parent without monitoring the
or in some other danger and also triggers an alarm. environment surrounded the children or can monitor
Every time the parent module cycles through all the only one state at a time. While the designed tracking
children, it packs the status of every child and system allows parent to monitor multiple children
displays information on the LCD. Also, parent and they will be alarmed if any child be in danger
module has a keypad which is used to let the parent state. The danger states that were taken in the
consideration are: missing far away from their parent,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


294
M. Juno Isabel Susinthra, S. Praveena Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 292-295

falling into a swimming pool, leaving in a hot car, Cite this article as :
leaving behind gas heater. The flexibility of the
designed system makes it easy to add new sensors for M. Juno Isabel Susinthra, S. Praveena, "Children
any child. Tracking Classification Control Using Android",
International Journal of Scientific Research in
VI.REFERENCES Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
[1] Druin A., "The role of children in the design of 5 Issue 1, pp. 292-295, January-February 2019.
new technology", Behavior and Information Available at doi :
Technology", vol21, No.1,pp1-25, 2002. https://doi.org/10.32628/CSEIT195154
[2] Morris Williams, Owain Jones, Constance Journal URL : http://ijsrcseit.com/CSEIT195154
Fleuriot and Lucy Wood, "Children and emerging
wireless Technologies", Conference on Human
Factors in Computing systems, 2005.
[3] Deborah L. Ruf, "Losing Our Minds", 2005, Great
Potential Press, Inc.
[4] Stephen S. Intill, James W. Davis and Aaron F.
Bobick, "Real-Time Closed-World Tracking",
Proceedings of the IEEE Computer Sciety
Conference on Computer Vision and Pattern
Recognition, pp. 697-703, June 1997.
[5] GregoryT.FRENCH "Understanding the GPS,
1996
[6] Katin Michael, "The Emerging Ethics of Human
Centric GPS Tracking and Monitoring",
International Conference on Mobile Business
(ICM'06), 2006.
[7] A. (Angela) Canny, Anne E. Green and Malcolm
J. Maguire, "Keeping Track", Published 2001, The
Policy Press.
[8] PIC16F87XA Datasheet 28/40-Pin Enhanced
Flash
[9] Lovine, John, “PIC Microcontroller Project
Book“, Mc Graw-Hill, 2004

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


295
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195163
Voice Based Information Sharing System for Visually Impaired People
R. Gomathijayam1, A. Jenifer2
1 Assistant Professor, Department of Computer Applications, Bon Secours College for Women, Thanjavur, Tamil
Nadu, India
M. Sc., Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India
2

ABSTRACT
In today’s world communication has become so easy due to integration of communication technologies
with internet. However, the visually challenged people find it very difficult to utilize this technology
because of the fact that using them requires visual perception. Even though many new advancements have
been implemented to help them use the computers efficiently no naïve user who is visually challenged can
use this technology as efficiently as a normal naïve user can do that is unlike normal users they require
some practice for using the available technologies. This paper aims at developing an email system that will
help even a naïve visually impaired person to use the services for communication without previous
training. The system will not let the user make use of keyboard instead will work only on mouse operation
and speech conversion to text. Also this system can be used by any normal person also for example the
one who is not able to read. The system is completely based on interactive voice response which will make
it user friendly and efficient to use.
Keywords : Visually Challenged People, IVR, Speech to Text Converter, Mouse Click Event, Screen
Reader

I. INTRODUCTION impaired it become necessary to make internet


facilities for communication usable for them also.
Internet is considered as a major storehouse of Therefore we have come up with this project in
information in today’s world. No single work can be which we will be developing a voice based email
done without the help of it. It has even become one system which will aid the visually impaired people
of the de facto methods used in communication. And who are naive to computer systems to use email
out of all methods available email is one of the most facilities in a hassle free manner. The users of this
common forms of communication especially in the system would not need to have any basic information
business world. However not all people can use the regarding keyboard shortcuts or where the keys are
internet. This is because in order to access the located. All functions are based on simple mouse
internet you would need to know what is written on click operations making it very easy for any type of
the screen. If that is not visible it is of no use. This user to use this system.
makes internet a completely useless technology for
the visually impaired and illiterate people. Even the II. LITERATURE SURVEY
systems that are available currently like the screen
readers TTS and ASR do not provide full efficiency to It is estimated that there are a total of 4.92 billion
the blind people so as to use the internet. As nearly email accounts existing in 2017 and there will be
285 million people worldwide are estimated visually approximately 5.59 billion accounts by the end of

CSEIT195163 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 296-298 ]
296
R. Gomathijayam, A. Jenifer Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 296-298

2019[4]. It is also estimated that there are a total of perception. Even though many new advancements
340.2 million smart phone users in India in the year have been implemented to help them use the
2017 [5]. This makes emails the most used kind of computers efficiently no naïve user who is visually
communication. The prevailing email systems don’t challenged can use this technology as efficiently as a
give any means of feedback or Talkback service. The normal naïve user can do that is unlike normal users
most common mail services that we tend to use in they require some practice for using the available
our day to day life cannot be used by visually technologies.
challenged people. This is as a result of they do not
offer any facility in order that the person in front will IV. PROPOSED METHODOLOGY
listen the content of the screen. As they cannot
visualize what is already present on screen they The proposed system relies on voice command based
cannot build out where to click in order to perform system unlike the existing mail systems. The most
the required operations. For a visually impaired important thing that has been kept in mind while
person employing a computer or smart phone system developing the proposed system is accessibility. A
for the first time isn't that convenient as it is for a system is accessible solely if it may be used
standard user even though it is user friendly. Though expeditiously by all varieties of individuals whether
there are several screen readers offered then also or not disabled. The current systems do not offer this
these individuals face some minor difficulties. Screen accessibility. Therefore the system we are developing
readers speak out whatever content is there on the is completely different from the current system.
screen and to perform the particular actions the Unlike current system that emphasizes a lot on user
person will have to use keyboard shortcuts because friendliness of traditional users, our system focuses a
mouse location cannot be detected by the screen lot on user friendliness of all varieties of individuals
readers. This means 2 things; one that the user cannot as well as normal people visually impaired individuals.
make use of mouse pointer as it is fully inconvenient The complete system is primarily based on speech to
if the pointer location cannot be derived and second text commands. Once using this system the
that user should be versed with the keyboard on application will be prompting the user to speak
wherever each and every key is placed. A user who is specific commands to avail respective services and if
new to computer will therefore not use this service as the user wants to access the respective services then
they're not conscious of the key locations. Also there he/she needs to speak that command.
are some difficulties faced by visually impaired
people when using smartphone systems. All these are V. IMPLEMENTATION RESULTS
some drawbacks of the present system that we are
going to overcome within the application we are
developing.

III. EXISTING PROCESS

In today’s world communication has become so easy


due to integration of communication technologies
Speech-To-Text: Input to the System is provided by
with internet. However the visually challenged
the means of speech-to text. Speech-to-Text API is
people find it very difficult to utilize this technology
used for conversion of speech-totext.HTML5’s Built
because of the fact that using them requires visual

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


297
R. Gomathijayam, A. Jenifer Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 296-298

in Speech Recognition API is used for this purpose keyboard usage is eliminated. The user solely has to
hence using this input is provided to different follow the directions given by the system and use
modules of the system as intended user cannot voice commands consequently to get the several
manually enter the data required. services offered. Other than this the user may have to
Text-To-Speech is used for conversion from text to be requested to feed info through voice inputs
speech. It is used to Speak menu to the user which whenever required.
can be used to provide list of available options a
particular user can do on a given module or page. It is VII.FUTURE ENHANCEMENT
also used to read message that is been received by the
intended user. As intended user cannot directly read. Voice could be extended to image attachments and
ResponsiveVocic.js is API that is used for conversion other options such as indentation, fonts etc., that are
of Text to speech. available with normal E-Mail.
Php Mailer and PhP-IMAP PhP Mailer is a library
which is provided by PhP which can be used to send VIII. REFERENCES
e-mail. PhP Mailer is used in proposed system which
is used to send emails. In order to extract the mails [1]. T. Shabana, A.Anam, A.Rafiya, K.Aisha, “Voice
from the system Php Imap is used to fetch the user’s based Email System for Blinds”, “IJARCCE”, Jan
mail from the IMAP server. 2015.
Knuth-Morris-Pratt Algorithm Pattern Matching is [2]. P. Ingle, Harshada Kanade, Arti Lanke, “Voice
very key component when it comes to searching mail based Email System for Blind”, “IJRSCSE”,
in inboxes. System provides three ways to search by 2016.
keyword, date and today’s date out of search by [3]. Jagtap Nilesh, Pawan Alai, Chavan Swapnil,
keyword requires KMP for pattern matching. Given a “Voice based System in Desktop and Mobile
string ‘S’, the problem of string matching deals with devices for blind people”, “IJETAE”, Feb 2014.
finding whether a pattern ‘p’ occurs in ‘S’ and if ‘p’ [4]. Prof Umesh A Patil, Pranouti Patil, “A survey
does occur then returning position in ‘S’ where ‘p’ on voice based mails system for physically
occurs. Knuth, Morris and Pratt proposed a linear impaired peoples”, “IJIRCCE”, Jan 2016.
time algorithm for the string matching problem as
compared to other approaches. Cite this article as:

VI.CONCLUSION R. Gomathijayam, A. Jenifer, "Voice Based


Information Sharing System for Visually Impaired
In this paper we have planned a system which can People", International Journal of Scientific Research
facilitate the visually impaired individuals to access in Computer Science, Engineering and Information
email services efficiently. This system can help in Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
overcoming some drawbacks that were earlier faced 5 Issue 1, pp. 296-298, January-February 2019.
by the blind individuals in accessing emails. We've Available at doi :
eliminated the thought of using keyboard altogether https://doi.org/10.32628/CSEIT195163
with screen readers which can help reducing the Journal URL : http://ijsrcseit.com/CSEIT195163
cognitive load of remembering keyboard operations.
Conjointly any user who will not grasp the location
of keys on the keyboard would like not worry as

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


298
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195149
A Review on Big Data Analytics Via Social Media
T. K. Anusuya*1, A. Preethi2
1 Head & Assistant Professor, PG Department of Department of Computer Applications, Bon Secours College for
Women, Thanjavur, Tamil Nadu, India
M. Sc., Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India
2

ABSTRACT

Analytics is very important in all fields in order to make decisions over certain facts. Social media analytics is
the process of collecting information from various social media platforms, websites and blogs. These analytics is
done to make effective business conclusions. The usage of social media has become the latest trend in today’s
world. Social data analytics is not about just collecting likes and comments shared by individuals but it has
become the platform for many trademarks to bring out promotion. Applications such as marketing, elections
widely used social data to make predictive decisions. Some of the approaches followed are forming hypothesis,
getting deep into the data, mapping events etc. These analytics can also be done in applications such as business,
Change in amendments, Education, Demonetization etc. The challenges faced are metrics formed by social
media should reach the right people, unstructured data being difficult to priestship paper discusses about the
model, theme, performance evaluation, advantages and disadvantages under literature survey.
Keywords : Social Media, Insights, Datasets, Analysis, Challenges

I. INTRODUCTION various applications in which Big data is being used.


The foremost is the Banking in which Securities
Big data is nothing but huge amount of data which Exchange Commission is being used to monitor
cannot be processed using traditional methods. The various activities. At present, they are using the
data to be processed may be structured, unstructured network analytics and natural language process as
or semi structured. These data come from various trend in the markets. Retail traders, big banks and
sources such as sensor data, transactional data, many others are using analytics for making decisions.
cameras, microphones and lot more of resources. On Moving on to the next application, it is the media.
these large data sets analytics is performed in order to Big data is widely used in media, since it has now
obtain undiscovered insights. Based on facts, reached the users through much hand held devices.
everyday petabytes of data are being generated and Here analysis is done on the consumer insights, social
now it has come up to zeta data. All social media media content and understanding pattern. Real time
platforms in which billions of users are connected examples of application using Big data are Spottily,
generate huge data which can be of any type such as Amazon Prime etc. Next widely used sector is
text, images, audio, video, gif, etc. Big data is highly Healthcare. This sector has access to huge sets of the
used in many applications because each and every data as they handle much number of patients and
field now generates tones of data which need to be their respective data.
processed at one point of time. Let us discuss the

CSEIT195149 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 299-302 ]
299
T. K. Anusuya, A. Preethi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 299-302

II. METHODS AND MATERIAL EXISTING PROCESS

LITERATURE SURVEY Big data is nothing but huge amount of data which
cannot be processed using traditional methods. The
A. Can Twitter save lives? A Broad scale study on data to be processed may bestructured, unstructured
Visual Social Media Analytics for Public Safety or semi structured. These data come from various
Theme: According to Dennis Thom et al. presents sources such as sensor data, transactional data,
that nowadays the social media is being used for cameras, microphones and lot more of resources. On
commercial purpose rather than safety. The paper these large data sets analytics is performed in order to
gives a clear understanding on the crisis intelligence obtain undiscovered insights. Based on facts,
field study via Twitter data during the flood on everyday terabytes of data are being generated and
German 2013. Scatter Blogs systems was introduced now it has come up to zeta data. All social media
to implement other techniques. The second phase platforms in which billions of users are connected
sketched out a system based on the feedback about generate huge data which can be of any type such as
Scatter Blogs and a comparison is made between both text, images, audio, video, gif, etc.
the systems. B. Big Data Analytics in Healthcare
Theme: Based on the field of healthcare Big Data PROPOSED METHODOLOGY
analytics and its benefits with the methodologies
used and its outcomes and the challenges faced are It has been observed that each paper has used
discussed here according to Wullianallur Raghupathi different techniques which have their own pros and
et al. C. Big Social Data Analytics for Public Health: corns. The techniques overall used are Scatter Blogs,
Face book Engagement and Performance Theme: SoLoMo, IBM Big Insights each used for a purpose
analysis of Facebook data using unsupervised learning such as analysis, storage, streaming real time data etc.
indicates recent tendency of user engagement in We have decided to take the positive features of all
public health has been increased as observed by the algorithms, tools and techniques used in the
Nadia Straton et al. D. SoLoMo analytics for Telco Big above discussed papers. The social Media analytics is
Data monetization Theme: The reference has the main resource to perform analysis. It is actually
observed that the world can be accessed using the the process of developing informative tools to collect
mobile internet. This has helped widely in Business. and summarize the data. The usage of social media
The algorithms and technologies used in this paper has been evolving more as the days pass by. It has
are used to find out the insights with the help of been used for many positive as well asnegative usage.
social, location and mobile data of the individual. E. It all depends upon the attitude of the users.
Impact of Demonetization on Indian Economy Facebook, YouTube and Twitter are in the first, third
Theme: India has the highest rate of currencies. 87% and tenth position on account of the usage. In general
of the currencies are rotating in the form of Rs.500 the analysis process comprises of three stages Capture,
and Rs.1000. These currencies become the source of Understand and Present. Topic modeling is also one
income either by working under organizations or of the important techniques being used in above
through illegal enterprise. This paper by Dr. Pratap papers. Hence Social media analytics is highly used in
Singh et al. presents the impacts after the process of the analysis that is performed on demonetization.
demonetization occurred in India.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


300
T. K. Anusuya, A. Preethi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 299-302

III. RESULTS AND DISCUSSION the API (Application Programming Interface) of the
respective platform. Sometimes it is requested to pay
in order to retrieve the data as they might keep their
resources secure. The next step processes it to
perform analysis on the retrieved data. This process
can also make use of R studio and flume platform to
store the data and perform analysis. The outcome is
derived in any of the formats such as Graph, line or
pie chart.

VI.REFERENCES

[1]. Dennis Thom, Robert Kruger, Thomas Ertl,


“Can Twitter Save Lives? A Broadscale Study on
Visual Social Media Analytics for Public
Safety”, IEEE Transactions on Visualization and
Computer Graphics, 2015.
IV. CONCLUSION [2]. Wullianallur Raghupathi and Viju Raghupathi,
“Big data Analytics in Healthcare: promise and
This paper presents about the introduction, potential”, Raghupathi and Raghupathi Health
characteristics, applications and limitations of Big Information Science and Systems, 2014.
data. Then about the introduction about social data [3]. Nadiya Straton, Kjeld Hansen, Raghava Rao
analytics which is the source for performing analysis. Mukkamala, Abid Hussain, TorMorten Grønli
The literature study includes the proposed work, Henning Langberg and Ravi Vatrapu, “Big
used tools and techniques in each respective paper. It Social Data Analytics for Public Health:
also contains the overall observations with the Facebook Engagement and Performance”, IEEE
futuristic solutions where it has been discussed about 18th International Conference on e-Health
the steps involved in analysis and the applications. Networking, Applications and Services
The conclusion and the reference for the above (Healthcom), 2016.
surveyed paper are also included. [4]. H. Cao, W.S. Dong, L.S. Liu, C.Y. Ma, W.H.
Qian, J.W. Shi, C.H. Tian, Y. Wang, D.
V. FUTURE ENHANCEMENT Konopnicki, M. Shmueli-Scheuer, D. Cohen, N.
Modani, H. Lamba, A. Dwivedi, A. A. Nanavati,
According the literature study that we did on the M. Kumar, “SoLoMo analytics for telco Big
paper, there were many techniques for analysis that Data monetization”, IBM J. RES. & DEV. VOL.
were implied on social data. This was especially to 58 NO. 5/6 Paper 9 September/November,
gather insights and make efficient decisions so that it 2014.
becomes useful for the users who wish to do research [5]. Dr. Partap Singh, Virender Singh, “Impact of
based on the problem. The first step process is to Demonetization on Indian Economy”, 3rd
collect data from the social networks such as Twitter, International Conference on Recent
Facebook. The data can be easily retrieved through Innovations in Science, Technology,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


301
T. K. Anusuya, A. Preethi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 299-302

Management and Environment, 18th


December, 2016
[6]. Weiguo Fan, Michel D. Gordan, “Unveiling the
Power of social Media Analytics”
[7]. Dr. Sunitha V Gangier, Ranganatha B,
“Demonetization and its impact on Social
Development”, INDIAN JOURNAL OF
APPLIED RESEARCH, January 2017

Cite this article as :

T. K. Anusuya, A. Preethi, "A Review on Big Data


Analytics Via Social Media", International Journal of
Scientific Research in Computer Science, Engineering
and Information Technology (IJSRCSEIT), ISSN :
2456-3307, Volume 5 Issue 1, pp. 299-302, January-
February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195149
Journal URL : http://ijsrcseit.com/CSEIT195149

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


302
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195147

Automatic Railway Gate Control System Using RFID with


High Speed Alerting System
M. Florence Dayana *1, R. Bhuvaneshwari 2
1 Head & Assistant Professor, Department of Computer Applications, Bon Secours College for Women,
Thanjavur, Tamil Nadu, India
2M. Sc., Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India

ABSTRACT

The main objective of this paper is to avoid the railway accidents happening at the level crossings. In India
railway is the major mode of the transportations and it is the cheapest way for travelling. So there are more
numbers of rail users. And it is not easy to stop railway anywhere to prevent accident or for any other reason.
Due to that there are major drawbacks of that. At present manned system is available at level crossings. Hence,
many accidents occur at such crossings, since there is no one to take care of the functioning of the railway gate
when a train approaches the crossing. The objective of this paper is to manage the control system of railway
gate using the microcontroller. This model includes IR sensors, RFID, LCD, buzzer, light led, motor driver and
microcontroller PIC16F877A. In the automatic railway gate control system, at the level crossing the arrival of
the train is detected by the IR sensor and RFID placed near to the gate. In case of RFID it detects only arrival of
train. Hence, the time for which it is closed is less compared to the manually operated gates and reduces the
human labour. As the entire system is automated errors occurring due to manual operation are prevented
because the accuracy of automated operation is more than the manned operation.
Keywords : Automatic Railway Gate, IR Sensors, RFID, PIC 16F877A, Motor Driver, Level Crossing.

I. INTRODUCTION the train by both sides. RFID is also used for the
detection of specific train only. RFID consisting of
We can’t take care of ours while in running by less tag and reader. Tag is given to the train and it will be
Railways being the cheapest mode of transportation read by that RFID reader only. When reader reads
are preferred over all the other means. But there are that tag which is placed on train then the gate will be
so many accidents happening at the manned control closed. When the last carriage of train passes having
level crossings. This is mainly due to the carelessness tag then the gate will be opened. LCD will display
in manual operations or lack of workers. This project information for the road users. This process Reduces
utilizes two pairs of IR sensor and RFID placed at the the traffic and time for which users have to for long
either sides of gate. Each pair of sensor consist of a time at level crossing.
transmitter and receiver, in our case light emitting
diode is used as a transmitter and a p-n junction
photodiode is used as receiver. It detects the arrival of

CSEIT195147 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 303-306 ]
303
M. Florence Dayana, R. Bhuvaneshwari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 303-306

Concept
At the present, India having world’s largest railway
network which having manned gate control system.
Over hundreds of railways running on track every
day. It is definitely impossible to stop the running
train at immediate in some critical situation or
emergency arises. Train accidents having serious
drawbacks in terms of loss of human life, injury,
damage to Railway property. This model uses
automatic control or operation of gate and reduces
the human labors. The concept of the model is to
control the railway gate using microcontroller
PIC16F877A. Figure 1. Block Diagram of System

Objective - To avoid an unwanted accident, Circuit Diagram


happening at the railway level crossing due to less It shows the circuit diagram of Automatic Railway
awareness of drivers and poor work of gate keeper Gate Control by using PIC 16F877A microcontroller.
and also time saves. This system includes the various components for the
operation of the system. This system includes two
II. METHODS AND MATERIAL components for the detection of the train i.e. IR
sensor and RFID. Initially IR emits the light. When
Block Diagram train cuts the IR sensor then it gives the input to the
microcontroller PIC16F877A, then microcontroller
Fig. 1 shows over all block diagrams for railway gate gives command to the gate driver so gate is closed
control system by using microcontroller PIC and buzzer will be ON. When the last carriage of the
16F877A. train cut the second IR sensor then gate is open and
buzzer is OFF. In case of RFID, the specific tag is
PIC1F877A performs some operations related to the given to the train. When that train is crosses the
system. IR and RFID detects the arrival the of train RFID, then that tag is read by the RFID reader. If the
and they gives the signal to the microcontroller. specific tag is read by train then and only
Accordingly microcontroller gives the command to microcontroller gives the command to the gate driver
the gate driver, LCD, buzzer. If train cut the IR then gate is closed and buzzer is ON. On the
sensor and RFID then microcontroller gives the departure of the train the gate is open and buzzer will
command to gate driver, it closed the gate and buzzer be OFF. Accordingly the red and green led is
is ON. After sometime, departure of train gate is operating. Also the LCD will display some message
opened and buzzer is OFF. Accordingly LCD will for the road users for informing the information
display the message to the road users. But in case of about the arrival and departure of the train.
RFID, it detects only the arrival of the train as
compared to the IR sensor. So RFID is the
advantageous than IR sensor for the detection of the
train.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


304
M. Florence Dayana, R. Bhuvaneshwari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 303-306

Algorithm large quantity. This system can contribute a lot of


• Step1: Start benefit either to the road users or to the railway
• Step2: Initialize timer, LED, interrupt management. Since the design is completely
• Step3: Display on LCD “WELCOME TO MET automated it can be used in remote villages where no
BKC IOE” station master or line man is present. The RFID
• Step4: Wait for 3sec placed at the either side of gate so they can detect the
• Step5: Clear LCD arrival and departure of train. Now a day’s automatic
• Step6: Display on LCD “AUTOMATIC system occupies each and every sector of applications
RAILWAY GATE CONTROL SYSTEM” as it is reliable and accurate.
• Step7: wait for 4sec
• Step8: if IR sensor1=1 or IR sensor2=1 or PROPOSED MODEL
RFID reader o/p=1 or RFID reader o/p=1
• Step9: if yes then display on LCD “Gate close
please wait”
• Step10: Relay1=ON
Relay2=OFF
Relay3=ON
Relay4=OFF
• Step11: Delay 1sec
• Step12: Red led on, green led off
• Step13: if IR sensor1=0 or IR sensor2=0 or
RFID reader o/p1=0 or RFID reader2=0
• Step14: Relay1=OFF
• Relay2=ON
• Relay3=OFF
• Relay4=ON
• Step15: Delay 1sec Figure 2. Proposed model of system
• Step16: Display on LCD “Gate open please go”
Step17: Stop. IV.REFERENCES

III.RESULTS AND DISCUSSION [1]. Hnin Ngwe Yee Pwint, ZawMyoTun,


HlaMyoTun, “Automatic Railway Gate Control
CONCLUSIONS System Using Microcontroller,” Volume 3,
Issue 5, May 2014.
Automatic gate control system using RFID is the [2]. Sayali R. More, Ruchira J. Raut, Rasika K.
efficient and easy way of reducing the accidents at Tandel,
the level crossings. This gives an secure model for the [3]. Snehal D. Yendhe, “Intelligent Railway
railway track system. Using RFID tag and reader, Crossing Gate Control with High Speed Anti-
microcontroller PIC16F877A detects only the arrival Collision Alerting System,” .
of the specific train so only at that time gate is open [4]. Krishna, Shakshi Yadav, Nidhi, “Automatic
and close. Due to that there is no any traffic jam and Railway Gate Control Using Microcontroller,”
wastage of time. This will reduce the accidents in

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


305
M. Florence Dayana, R. Bhuvaneshwari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 303-306

ISSN: 0974-6471 December 2013, Vol. 6, No.


(4).
[5]. R. BoopalaVenkatesh, K. Kamalanathan,
“Microcontroller Based Automatic Railway
Gate Control With Zero Accident,” Vol. 3,
Special Issue 2, March 2016.
[6]. Vikash Kumar, Prajit Paul, Nishantkumar,
Pratik kumar Sinha, Sumant Kumar Mahato,
“Automatic Railway Gate Controller with High
Speed Alerting System,” Vol. 4, Issue 5, May
2015.

Cite this article as :

M. Florence Dayana, R. Bhuvaneshwari, "Automatic


Railway Gate Control System Using RFID with High
Speed Alerting System", International Journal of
Scientific Research in Computer Science, Engineering
and Information Technology (IJSRCSEIT), ISSN :
2456-3307, Volume 5 Issue 1, pp. 303-306, January-
February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195147
Journal URL : http://ijsrcseit.com/CSEIT195147

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


306
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195141
Design of Robot for Scrambling on Tree and Clutching Fruit
Using Foundary Design
V. Manochithra1, S. Karthika2
1Head, Department of Information Technology, Bon Secours College for women, Thanjavur,
Tamil Nadu, India
2 M.Sc., (Computer Science), Bon Secours College for women, Thanjavur, Tamil Nadu, India

ABSTRACT

The advancement in nature has prompted the presentation of very effective organic systems. Emulating these
instruments offers gigantic possibilities for the change of our life and the apparatuses we utilize. Nowadays, the
robot is biggest mechanism that is used for each and every process as it has the gripper assembly designed uses a
flexible casing which grasps the tree trunk of almost all diameter within the predefined range, basically it uses
passive compliance. This mechanism can be used for climbing trees which are almost height like coconut trees
and poles. The main drawback is being slow and there is a possibility of damaging the tree trunk. Hence this
system propose a new tree climbing mechanism for multi tree climbing robot that can be used to climb trees
that are almost straight and height.
Keywords : Robot, Grippers, Climbing trees, Clutching

I. INTRODUCTION mechanism that can be used to climb trees that are


almost straight, like coconut trees and poles with
Most commonly used design for tree climbing is inch greater agility.
worm design. These models are very slow. The main
body of this type of robot is divided into two parts The main locomotion of the robot is caused due to
and each part has a gripper. These models using inch the two motors which facilitates simultaneous
worm mechanism are continuous in their motion and motion of two consecutive links. This robot grasps
can make over in complex tree environments the tree with the help of grippers which is acted by a
involving multiple branches. Next common type of spring. When the motor is ON it unclamps and
design is wheeled robot where in instead of grippers clamps while the motor is OFF by spring action while
wheels are used for climbing up. The robot engulfs harnessing the spring energy for gripping. When one
the tree and locks against it and as the motor rotates, gripper grasps the tree, the whole body makes a
it moves up. The sizes of these robots are comparable revolution such that the next gripper comes up and
to the diameter of tree. These models are suitable for grips while the other un grasps and the whole process
trees with straight and plain trunks. The main continues. Attachments like pesticide sprayer, weed
drawback is it being slow and there is a possibility of remover or some harvesting mechanism can be
them damaging the tree trunk. Hence most of the mounted on the Robot. With these attachments this
models are slow, less agile, have less load carrying robot can act as harvesting or maintenance
capacity. As a result, we propose a tree climbing equipment by the farmers.

CSEIT195141 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 307-311 ]
307
V. Manochithra, S. Karthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 307-311

These models using inch worm mechanism are pesticide sprayer, weed remover or some harvesting
continuous in their motion and can maneuver in mechanism can be mounted on the Robot. With
complex tree environments involving multiple these attachments this robot can act as harvesting or
branches. Next common type of design is wheeled maintenance equipment by the farmers.
robot where in instead of grippers wheels are used for
climbing up. The robot engulfs the tree and locks Coconut picking is a tough, arduous work. The
against it and as the motor rotates, it moves up. The farmers in popular coconut cultivating countries are
sizes of these robots are comparable to the diameter facing severe shortage of skilled workers for
of tree. These models are suitable for trees with harvesting. In recent years, the number of coconut
straight and plain trunks. tree climbers has declined sharply, as those men who
do this rigorous job of climbing 30 to 80 feet high
The main drawback is it being slow and there is a above the ground, bare handed or with the aid of a
possibility of them damaging the tree trunk. Hence rope, have broken away from their traditional role to
most of the models are slow, less agile, have less load take-up better paid jobs. Since the Green Revolution
carrying capacity. As a result, we propose a tree of the late 1960s, the harvesting of coconuts today
climbing mechanism that can be used to climb trees has little changed from how it used to be done for
that are almost straight, like coconut trees and poles generations. The job of coconut picking, which is
with greater agility. The main locomotion of the perceived as risky and unglamorous, does not attract
robot is caused due to the two motors which youngsters either, even though the remuneration is
facilitates simultaneous motion of two consecutive fairly good in local standards. Though there exist few
links. tree climbing machines today, by any measure,
climbing for coconuts is hard work. Not only does a
climber have to scale the tree and reach out among
the fronds and cut the coconut, but they also have to
instinctively learn which nuts are ripe enough to be
harvested. Besides, those few climbers who still
remain in the business of coconut picking, claims
higher wages.

III. PROBLEM STATEMENT


Fig.1. System Model

In traditional harvesting method on reaching the top


II. LITERATURE SURVEY
of a coconut tree, the climber taps the nut in the
lowermost bunch with its harvesting knife to test its
This robot grasps the tree with the help of grippers
maturity. If he is satisfied, he cuts the bunch at the
which is acted by a spring. When the motor is ON it
base of the stalk when it drops down to the ground.
un-clamps and clamps while the motor is OFF by
The climber also cleans the crown and removes the
spring action while harnessing the spring energy for
dry leaves, sheaths and spathes. In some places where
gripping. When one gripper grasps the tree, the
the trees are not tall, cutting the coconut bunches
whole body makes a revolution such that the next
with a knife, attached to a long bamboo pole, does
gripper comes up and grips while the other ungrasps
the harvesting. It seems quite simple to climb the
and the whole process continues. Attachments like
coconut palm in order to stay beneath the leaf crown

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


308
V. Manochithra, S. Karthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 307-311

to harvest fruits and cut old leaves. But it proves


much more challenging to climb up the coconut leaf
crown in order to reach the young inflorescences for
making controlled pollinations with bagging for
research purposes, and for harvesting the toddy (the
sweet sap from coconut inflorescences), which serves
to produce sugar, vinegar, wine and spirits. This
paved the way to the proposed robotic harvesting
method, where it can peep and reach out to any
remote nook and corner at the top of the tree, of the Fig.2. Robot Climbing Model
operator’s choice.

V. LOCOMOTION OF ROBOT
IV. METHODOLOGY
There are three types of motion. These are classified
In traditional harvesting method on reaching the top as continuous, discrete and serpentine. As we get
of a coconut tree, the climber taps the nut in the closer to continuum type, energy consumption is
lowermost bunch with its harvesting knife to test its reduced and speed is increased. But it is difficult to
maturity. If he is satisfied, he cuts the bunch at the achieve continuous motion, so discrete form is
base of the stalk when it drops down to the ground. followed in the construction.
The climber also cleans the crown and removes the
dry leaves, sheaths and spathes. In some places where First, the bot is attached to the tree trunk with the
the trees are not tall, cutting the coconut bunches help of grippers which grips on to the tree trunk and
with a knife, attached to a long bamboo pole, does provides adequate friction such that it does not slip.
the harvesting. The motors used for climbing overcome the moments
caused by its weight.
❖ Ease of use - They wanted use of the robot to be
something that could be easily learned, so it The movement takes place as follows:
would not slow down ALB inspection rate.
❖ Work safe - They wanted the system to have fail 1. First the arm with first motor will hold the tree.
safes implemented to ensure no potential harm 2. Then the link with second motor will release the
to USDA operators in the field. grip and the whole setup will rotate about motor
❖ Minimal damage to tree - They did not want the one as shown in figure.
robot to inflict any serious harm on the tree it 3. After half a rotation, the arm with second motor
was climbing. will grip the tree and the process continues. Once
❖ Navigation around branches - They needed to the robot reaches the topmost point of the tree
robot to be able to find its way around branches, trunk, the whole body is gripped on to the trunk
so inspections could be complete. without any further movement, i.e., the body is
grounded, thereafter miscellaneous action like
harvesting or maintenance work can be
performed.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


309
V. Manochithra, S. Karthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 307-311

device, hence conventional climbing robots cannot


be used for tree climbing applications.

Fig.3. Robot Climbing Model


Graph.1. Efficiency Analysis

This mechanism can be achieved by using a push


plate in the gripper which will be connected to the
shaft of linear motor, when the linear motor actuates
the push plate pushes the claws and unclamps from
the tree.

VI. RESULT AND DISCUSSIONS Graph.2. Robot View

The evolution in nature has led to the introduction of VII. CONCLUSION


highly efficient biological mechanisms. Imitating
these mechanisms offers enormous potentials for the In this work a tree climbing mechanism for coconut
improvement of our life and the tools we use. trees has been proposed this mechanism can also be
used to climb structures and other trees which are
We have devised this mechanism inspired from the almost straight like palm trees, poles etc. The
locomotion of a worm called Stomata pod which uses mechanism used in this robot is a bio inspired
its whole body as distributed foot facilitating it to mechanism which is derived from an organism called
move on any terrain and have high agility on a Stomata pod. This mechanism appears to be more
regular tree environment. The gripper assembly agile compared to other available mechanisms but the
designed uses a flexible casing which grasps the tree main limitation is that it can be used only for a
trunk of almost all diameter within the predefined smaller category of trees.
range, basically it uses passive compliance.
A prototype of robot was made and the efficiency of
This mechanism can be used for climbing trees which the slinky type gripper was tested on few trees. Also,
are almost straight like coconut trees and poles. software simulation was done for this mechanism.
Hence this paper presents a new climbing mechanism Our future works includes the addition of
for coconut tree climbing robot. It also requires agricultural equipment’s like harvester sprayer etc.
greater agility and high maneuverability to be used as also to develop upon the gripping materials as
a product. Also the bark of some trees may not be discussed.
strong enough to bear the weight of the climbing

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


310
V. Manochithra, S. Karthika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 307-311

VIII. REFERENCES [8]. Appeltants W and Ilsbroux W (1995)


Automatische sturing van een plukwagen
[1]. Juan Carlos Grieco,Manueul Prieto,Manuel (Autonomous guidance for a harvester). MA
Armada and Pablo Gonzalez de Santos, “A Six Thesis, Katholieke Hogeschool Limburg,
Legged Climbing Robot For High Payloads” Belgium
Proc. of the 1998 IEEE International [9]. Bulanon DM, Kataoka T, Okamoto H, Hata S
Conference on Control Applications ,Trieste, (2005) Feedback control of manipulator using
Italy,pp.446- 450,1998 machine vision for robotic apple harvesting. In
[2]. Yili Fu, Zhihai Li, Hejin Yang and Shuguo Proceedings of ASAE, Paper No. 053114,
Wang.” Development of a Wall Climbing Tampa, USA
Robotwith Wheel-Leg Hybrid Locomotion [10]. Bulanon DM, Kataoka T, Ukamoto H, Hata S
Mechanism”Proceedings of the 2007 IEEE (2004) Development of a realtime machine
International Conference on Robotics and vision system for the apple harvesting robot. In
Biomimetics, Sanya, China,pp.1879-1881,2007 SICE Annual Conference in Sapporo, pages
[3]. Michael P. Murphyand andMetin Sitti.” 595-598, Hokkaido Inst. of Techn., Japan
Waalbot: An Agile Small-Scale WallClimbing [11]. Hutchinson S, Hager G, Corke P (1996) a
RobotUtilizing Dry Elastomer Adhesives”, tutorial on visual servo controls. IEEE Trans. on
IEEE/ASME TRANSACTIONS ON Robotics and Automation, 12:651-670
MECHATRONICS, VOL. 12, NO. 3, pp.330- [12]. Jimenez A, Ceres R, Pons J (2000) a survey of
338,2007. computer vision methods for locating fruits on
[4]. Ozgur Unver, Ali Uneri, Alper Aydemir and trees. Transactions of the ASAE, 43(6):1911-
Metin Sitti,” Geckobot: A Gecko Inspired 1920
Climbing Robot Using Elastomer Adhesives.” [13]. Katypitiya J, Eaton R, Cole A, Meyer C, Rodnay
Proceedings of the 2006 IEEE International G (2005) Automation of an agricultural tractor
Conference on Robotics and for fruit picking. In IEEE Int. Conf. on Robotics
AutomationOrlando, Florida, pp.2329-2335, and Automation, pages 3212-3217, Barcelona,
2006. Spain
[5]. Tin Lun Lam and Yangsheng Xu, A Flexible [14]. Mateus D, Avi˜na G, Devy M (2005) Robot
Tree Climbing Robot: Treebot - Design and visual navigation in semi-structured outdoor
Implementation,’ 2011 IEEE International environments. In IEEE Int. Conf. on Robotics
Conference on Robotics and Automation and Automation, pages 4702-4707, Barcelona,
Shanghai International Conference Center, Spain
Shanghai, China,pp.5849-5854,2011 [15]. MVTec Software GmbH, M¨unchen, Germany
[6]. Y. Kushihashi, et al., “Development of Tree (2006) Halcon 7.1, Quick Guide, 2b edition
Climbing and PruningRobot,Woody-1- Cite this article as : V. Manochithra, S. Karthika, "Design
Simplification of Control using adjust Function of Robot for Scrambling on Tree and Clutching Fruit Using
of Grasping Power” .Proceedings of JSME Foundary Design", International Journal of Scientific
Research in Computer Science, Engineering and
Conference onRobotics and Mechatronics, pp.
Information Technology (IJSRCSEIT), ISSN : 2456-3307,
1A1- E08, 2006.
Volume 5 Issue 1, pp. 307-311, January-February 2019.
[7]. H. Kawasaki, et al., “Novel climbing method of
Available at doi : https://doi.org/10.32628/CSEIT195141
pruning robot”, Proceedings of the SICE Journal URL : http://ijsrcseit.com/CSEIT195141
Annual Conference, Tokyo, pp. 160-163,2008.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


311
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195142
Spinal Muscular Atrophy Type B Awareness and Symptoms
Prediction Using 3D Segmentation Process in MRI Images
V. Manochithra, G. Sumithra
1Head, Department of Information Technology, Bon Secours College for women, Thanjavur,
Tamil Nadu, India
2 M.Sc., (Computer Science), Bon Secours College for women, Thanjavur, Tamil Nadu, India

ABSTRACT

Spinal muscular atrophy (SMA) describes a group of disorders associated with spinal motor neuron loss. In this
review we provide an update regarding the most common form of SMA, proximal or 5q SMA, and discuss the
contemporary approach to diagnosis and treatment. Electromyography and muscle biopsy features of
denervation were once the basis for diagnosis, but molecular testing for homozygous deletion or mutation of
the SMN1 gene allows efficient and specific diagnosis. In combination with loss of SMN1, patients retain
variable numbers of copies of a second similar gene, SMN2, which produce reduced levels of the survival motor
neuron (SMN) protein that are insufficient for normal motor neuron function. Despite the fact that the
understanding of how ubiquitous reduction of SMN protein leads to motor neuron loss remains incomplete,
several promising therapeutics are now being tested in early phase clinical trials. This proposed model
investigates the symptoms and scans readings from the initial MRI scan images of babies with mutation
progress and SMN proteins formation benchmark values for this particular disorder SMA and further this
segmented parameters are acquitted into the K-means clustering technique that predict the report with the
disorder symptoms with MSE (mean square error) values that helps the babies in future to take prevention
measures to overcome this problem.
Keywords : Spinal Muscular, K-Means Clustering, Neuron, Disorder Symptoms, MRI Images.

I. INTRODUCTION legs first and then the arms and, in due time, they
will become quadriplegic and even more disabilities
Spinal Muscular Atrophy (SMA) is a progressive can follow including speech impairment. Assistive
neuromuscular disorder. Usually, this condition is Technology support for people with such disabilities
considered genetically induced with no known cure often requires identification of the best residual
to date. Children are born with the condition and muscular function so that this can be utilized as a
develop muscular weakness progressively as they means of voluntary control.
grow. The weakness ultimately encompasses the
whole muscular function rendering the limbs Spinal muscular atrophy (SMA) refers to a group of
dysfunctional or paralyzed. Many children with SMA, inherited diseases of the motor nerves that cause
if they do not have the weakness from the beginning, muscle weakness and atrophy (wasting). The motor
will start having the disease manifesting itself on the nerves arise from the spinal cord and control the

CSEIT195142 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 312-318 ]
312
V. Manochithra, G. Sumithra Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 312-318

muscles that are used for activities such as breathing, muscles at disease duration of >10 years compared to
crawling, walking, head and neck control, and disease duration of ≤10 years. Longitudinal follow-up
swallowing. SMA is a rare disorder affecting of four patients with two MRI studies eight years
approximately 1 out of every 10,000 individuals apart showed evidence of slow progression. Clinical-
worldwide. radiological comparison revealed a good correlation
for most of the muscles. This muscle MRI study in a
SMA affects muscles throughout the body. In the relatively large number of patients showed a specific
most common types, weakness in the legs is generally pattern of involvement in SMA 3b and revealed that
greater than in the arms. Sometimes feeding, muscle MRI can be an objective and powerful tool in
swallowing, and respiratory function (e.g., breathing, showing disease progression with implications for
coughing, and clearing secretions) can be affected. clinical trials.
When the muscles used for breathing and coughing
are affected and weakened, this can lead to an Clinical Features Of SMA
increased risk for pneumonia and other respiratory
infections, as well as breathing difficulty during sleep. Since only α motor neurons are lost progressively,
The brain’s cognitive functions and the ability to feel only motor function is compromised and sensory
objects and pain are not affected. People with SMA neurons are unaffected. This loss of function leads to
are generally grouped into one of four types (I, II, III, weakness and to progressive symmetrical atrophy of
IV) based on their highest level of motor function or the proximal voluntary muscles of the legs, arms and,
ability. sometimes, the trunk, as the disease progresses. A
number of unusual clinical features are observed in
There are only a few reported muscle magnetic SMA. One of these is the distribution of muscle
resonance imaging (MRI) studies of genetically- weakness, which is more compatible with a
confirmed spinal muscular atrophy (SMA). The aim myopathic disorder than with a neurogenic
of this study was to search for the presence of disorder.12 Proximal muscles are more involved than
selective muscle involvement and of disease distal muscles, legs are more affected than arms and
progression by MRI in SMA 3b. MRI of proximal arms are more affected than the face anddiaphragm.8,
upper and lower extremities was performed on a 1.5- 12 In other words, muscle weakness and atrophy does
Tesla MR scanner using conventional T1 and T2 not have a homogeneous distribution. The severity of
weighted axial images in twenty-two patients with muscle weakness is related to age at onset and
genetically-confirmed SMA. Patients had detailed children with the most severe form of the disease
clinical examinations within one month of the (Type I SMA) can appear normal at birth, but present
radiological investigation. MRI revealed a selective muscle weakness few months later. Furthermore, the
involvement of muscles. Triceps, iliopsoas and clinical course followed by SMA patients.
quadriceps femoris were the most affected muscles.
In contrast, biceps brachii, deltoid and gluteus Who survive beyond childhood demonstrates that
maximus were well-preserved. Gracilis, short head of loss of muscle strength is normally most evident at
biceps femoris, semimembranosus, and Sartorius and diseases onset and that, after this, residual muscle
adductor long us were also relatively spared. Slow, power can remain stable for months or years.
but definite progression with a predictable sequence
and magnitude was documented. Median MRI grade II. LITERATURE SURVEY
was higher (worse) usually by one grade in the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


313
V. Manochithra, G. Sumithra Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 312-318

One of the major challenges facing researchers poses an extra challenge precisely due to their
engaged in the study of neurodegenerative and complicated genetics. An attractive alternative would
neuron development diseases is to explain why therefore be to address these issues in diseases with a
defects in ubiquitously expressed proteins have such simple Mendelian inheritance pattern. Spinal
a selective effect on the nervous system and its muscular atrophy provides researchers studying
constituent cell types. Explaining the molecular neurodegenerative diseases with this opportunity.
mechanisms underlying the disease phenotypes has This review highlights why (1) SMA might be
been hampered by their multifactorial nature and considered a prototypical neurodegenerative disease
high incidence of sporadic cases. Alzheimer’s disease, in which to address certain general questions facing
Parkinson’s disease, and amyotrophic lateral sclerosis the field, and (2) why a thorough understanding of
(ALS) are among the most common this disorder might shed light on the mechanisms
neurodegenerative diseases in humans and good that have a specific effect on the development, health,
examples of disorders that exemplify these challenges. survival, and causes of motor neuron degeneration,
Two less common neurological diseases which are which ultimately leads to a disease phenotype.
monogenic in origin but are the result of defects in Proximal spinal muscular atrophy, commonly
ubiquitously expressed proteins involved in referred to as SMA, is a common autosomal recessive
housekeeping functions are Rett syndrome and 5q neuromuscular disease that affects the anterior horn
spinal muscular atrophy (SMA). cells of the spinal cord, resulting in atrophy of the
proximal muscles of the limbs and trunk. There are
The former is characterized by mutations in the gene numerous other forms of spinal muscular atrophy
encoding methyl – CpG binding protein 2 (MeCP2), which share certain characteristics with proximal
which likely regulates gene expression and chromatin SMA; however, they are genetically distinct and
structure; the latter by a deficit of the survival of often affect different subsets of neurons and muscle.
motor neurons (SMN) protein, whose best They include autosomal dominant forms of the
characterized function is in snRNP biogenesis and disease (Sambuughin et al., 1998; Van der Vleuten et
pre-mRNA splicing. Yet, despite the apparent need al., 1998), X-linked forms (La Spada et al., 1991;
for these proteins in all tissues, Rett syndrome, a Kobayashi et al., 1995), recessive forms that affect the
prototypical neurodevelopment disorder, is distal muscles (Viollet et al., 2002), and a severe form
characterized by a specific defect in maintaining of SMA (SMARD) with respiratory distress
proper function of post mitotic neurons in the (Grohmann et al., 2001) (Table 1).
forebrain, hippocampus, and brainstem, while SMA
has a particularly profound effect on lower motor After cystic fibrosis, SMA is the most common
neurons. autosomal recessive disorder in humans, with a
carrier frequency of approximately 1 in 35 and
This begs the question: do the proteins implicated in therefore an incidence of 1 in 6000 in the human
these diseases have functions in neurons other than population. It is also the most common genetic cause
their proposed housekeeping roles? Do they modulate of infant mortality. Despite the high incidence of the
downstream pathways specific to the tissues affected? disease in the human population, SMA has gained
Alternatively, are neuronal cells simply more relatively little attention among researchers studying
susceptible than other tissues to genetic, biochemical, neurodegenerative diseases. However, it has a
and environmental insults? Answering these fascinating biology, which includes two major players:
questions using multifactorial diseases as paradigms the highly homologous SMN1 and SMN2 genes.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


314
V. Manochithra, G. Sumithra Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 312-318

some labels cannot be adjacent to each other, then a


A splicing defect in SMN2 is a key factor in causing higher penalty is applied for the neighbouring pixels
the disease phenotype, while the SMN protein, that are assigned those labels for the given
expressed by both genes, is very likely neighbourhood system. Finally, the 3-D graph cut
multifunctional. However, low levels of the protein structure is formed and segmentation is performed.
have a particularly detrimental effect on one tissue
type, the lower motor neurons. One of SMN’s
functions is essential to cell survival. These
characteristics, coupled with a relatively simple
Mendelian inheritance pattern, make a compelling
case that study of this disorder in more detail will
provide a prototype that might shed considerable
light on motor neuron biology and disease. Such
studies may eventually lead to a better understanding
of SMA and other similar diseases and may accelerate
progress toward rational therapeutics.

III. METHODOLOGY

Magnetic Resonance Imaging is a fast and high


resolution technique that is commonly used to
investigate the anatomy and physiology of the body
in health and disease. High resolution imaging of the
important heart elements has become possible
through Multi-slice MRI technology. There has been
a very large body of studies using MRI scans to
segment body parts for research purposes and also to
increase the efficiency of the clinician’s work.

There are three major parts of the proposed approach


to segment 3-D MRI SMA Data with minimum user
interaction.

In the first step, a pre-processing operation is


employed to remove noise from individual slices.

Second, the user is required to provide soft


constraints to incorporate information about spatial
distance of regions to a reference point.

Fig.1. Work Flow


In the third step, the smoothness term can be
adjusted based on the input data. For example, if

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


315
V. Manochithra, G. Sumithra Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 312-318

employing new drugs and drugs that are already


STEPS INVOLVED: well-known. It cannot be expected that the
pharmacological treatments currently being studied
⚫ Respondents were invited by Cure SMA to can recover motor neurons or muscle cells that have
participate in the survey via email containing a already been lost to atrophy.
link to the survey.
Rather, the objective is to retard the progress of the
⚫ Responses were collected, tallied and analyzed
disease and improve residual muscle function. Sadly,
by Silicon Valley Research Group.
paralysis can be halted, but it cannot be reversed.
⚫ No PHI was collected from survey respondents However, through a combination of medical care and
through the survey or by any other means. rehabilitation, many patients with SMA can have
fulfilling and productive lives and often have normal
⚫ The survey was submitted to the IRB for life expectancy.
approval prior to data collection and is IRB
approved and compliant. Finally, there is one other point it is important to
emphasize, which is that since SMA is a recessive
⚫ The total sample size for the study was 298
genetic disorder, genetic counselling is an essential
completed responses, yielding a margin of error
component forthe families of these patients. Through
of 5.68%.
genetic counselling,
⚫ A possible explanation for the consistency in
most tolerable and least tolerable risks across all
treatments is that SMA sufferers and their
caregivers consider any of the benefits of
treatments presented as equally important, i.e.
they are not trading off risks for different
treatments.

⚫ Survey respondents also appeared to weigh risks


against their probabilities of occurrence.
Consistently, high probabilities of occurrence
made a risk less tolerable. Graph.1. Efficiency Analysis

IV. RESULT AND DISCUSSIONS

Both medical follow-up and palliative care are


important throughout SMA patients’ entire lives.
This care includes respiratory and nutritional support,
and orthopaedic and physiotherapeutic care to avoid
postural disorders.

In addition to this we can cite pharmacological


treatments that are still being studied, both

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


316
V. Manochithra, G. Sumithra Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 312-318

Graph.2. Diagnosed SMA of severity of spinal muscular atrophy. Am J


V. CONCLUSION Hum Genet. 2002;70:358‑68.
[6]. Chang JG, Hsieh-Li HM, Jong YJ, Wang NM,
It is obvious that significant challenges remain Tsai CH, Li H. Treatment of spinal muscular
toward an eventual treatment for SMA. These stem atrophy by sodium butyrate. Proc Natl AcadSci
from the many unanswered questions this review has U S A. 2001;98:9808-13.
raised. Readers may come away thinking that SMA [7]. Campbell L, Potter A, Ignatius J, Dubowitz V,
research is still in its infancy. However, it should be Davies K. Genomic variation and gene
noted that the tools to answer these questions, i.e., an conversion in spinal muscular atrophy:
excellent set of animal models and a basic implications for disease process and clinical
understanding of the disease, already exist. It should phenotype. Am JHum Genet. 1997;61:40-50.
also be reiterated that there are a number of [8]. Wang CH, Finkel RS, Bertini ES, Schroth M,
compelling reasons investigators outside the field Simonds A, Wong B,et al. Consensus statement
may find problems in SMA an attractive challenge to for standard of care in spinal muscular atrophy.
take on; for example, SMA is a neurodegenerative J Child Neurol. 2007; 22:1027-49.
disease in which a defect in a ubiquitously expressed [9]. Oskoui M, Levy G, Garland CJ, Gray JM,
protein affects a very specific tissue type. The O’Hagen J, De Vivo DC,et al. The changing
genetics of the disease are also relatively simple: SMA natural history of spinal muscular atrophy type
is autosomal recessive and involves a loss of function 1. Neurology. 2007;69:1931-6.
in an essential protein in certain cells. In essence, [10]. Wirth B, Brichta L, Hahnen E. Spinal muscular
SMA results from an insufficient amount of the SMN atrophy: from geneto therapy. Semin Pediatr
protein. Neurol. 2006;13:121-31. Review.
[11]. Zerres K, Rudnik-Schoneborn S. Natural
VI. REFERENCES history in proximal spinal muscular atrophy.
Clinical analysis of 445 patients and suggestions
[1]. Araujo AP, Ramos VG, Cabello PH. for a modification of existing classifications.
Dificuldades diagnósticas naatrofia muscular ArchNeurol. 1995; 52:518-23.
espinhal. Arq Neuropsiquiatr. 2005;63:145-9. [12]. Sumner CJ. Molecular mechanisms of spinal
[2]. Prior TW. Spinal muscular atrophy diagnostics. muscular atrophy. JChild Neurol. 2007;22:979-
J Child Neurol2007; 22:952-6. Review. 89.
[3]. Russman BS. Spinal muscular atrophy: clinical [13]. Swoboda KJ, Kissel JT, Crawford TO, Bromberg
classifications and disease heterogeneity. J MB, Acsadi G,D’Anjou G, et al. Perspectives on
Child Neurol. 2007; 22:946-51. clinical trials in spinal muscular atrophy. J
[4]. Shanmugarajan S, Swoboda KJ, Iannaccone ST, Child Neurol. 2007;22:957-66.
Ries WL, Maria BL,Reddy SV. Congenital bone [14]. Heier CR, Gogliotti RG, DiDonato CJ. SMN
fractures in spinal muscular atrophy: functional transcript stability: couldmodulation of
role for SMN protein in bone remodelling. J messenger RNA degradation provide a novel
Child Neurol.2007; 22:967-73. therapy for spinal muscular atrophy? J Child
[5]. Feldkotter M, Schwarzer V, Wirth R, Wienker Neurol. 2007;22:1013-8.
TF, Wirth B. Quantitative analyses of SMN1 [15]. Meldrum C, Scott C, Swoboda KJ. Spinal
and SMN2 based on real-time light Cycler PCR: muscular atrophy geneticcounseling access and
fast and reliable carrier testing and prediction

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


317
V. Manochithra, G. Sumithra Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 312-318

genetic knowledge: parents’ perspectives. J [24]. Oskoui M, Kaufmann P. Spinal


Child Neurol. 2007; 22:1019-26. muscularatrophy.Neurotherapeutics.2008;5:499
[16]. Lefebvre S, Bürglen L, Reboullet S, Clermont O, -506. Review.
Burlet P, ViolletL, et al. Identification and
characterization of a spinal muscular atrophy-
determining gene. Cell. 1995;80:155-65. Cite this article as :
[17]. Sumner CJ, Huynh TN, Markowitz JA, Perhac
JS, Hill B, CoovetDD, et al. Valproic acid V. Manochithra, G. Sumithra , "Spinal Muscular
increases SMN levels in spinal muscular Atrophy Type B Awareness And Symptoms
atrophy patient cells. Ann Neurol. 2003;54:647- Prediction Using 3d Segmentation Process In Mri
54. Images", International Journal of Scientific Research
[18]. Burghes AH, Beattie CE. Spinal muscular in Computer Science, Engineering and Information
atrophy: why do lowlevels of survival motor Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
neuron protein make motor neurons sick? Nat 5 Issue 1, pp. 312-318, January-February 2019.
Rev Neurosis. 2009;10:597-609. Available at doi :
[19]. Monani UR, Sendtner M, Coovert DD, Parsons https://doi.org/10.32628/CSEIT195142
DW, AndreassiC, Le TT, et al. The human Journal URL : http://ijsrcseit.com/CSEIT195142
centromeric survival motor neuron gene
(SMN2) rescues embryonic lethality in SMN (-
/-) mice and results in a mouse with spinal
muscular atrophy. Hum Mol Genet.2000;9:333-
9.
[20]. Burghes AH. When is a deletion not a deletion?
When it is converted. Am J Hum Genet.
1997;61:9-15. Review.
[21]. McAndrew PE, Parsons DW, Simard LR,
Rochette C, Ray PN, MendellJR, et al.
Identification of proximal spinal muscular
atrophy carriers and patients by analysis of
SMNT and SMNC gene copy number. Am J
Hum Genet. 1997; 60:1411-22.
[22]. Coovert DD, Le TT, McAndrew PE,
Strasswimmer J, Crawford TO,Mendell JR, et al.
The survival motor neuron protein in spinal
muscular atrophy. Hum Mol Genet.
1997;6:1205-14.
[23]. McWhorter ML, Monani UR, Burghes AH,
Beattie CE. Knockdown of the survival motor
neuron (SMN) protein in zebra fish causes
defects in motor axon outgrowth and path
finding. J Cell Biol.2003;162:919-31.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


318
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1183818
A Survey of Telemedicine Services using Datamining
T. K. Anusuya1, P. Maharajothi 2
Assistant Professor, Department of Computer Science, Bon Secours College for Women, Thanjavur, Tamil
1

Nadu, India
2M.Sc Computer Science, Department of Computer Science, Bon Secours College for Women, Thanjavur, Tamil
Nadu, India

ABSTRACT

The transfer of the medical care services to the patient, rather than the transport of the patient to the medical
services providers is aim of the project. This is achieved by using web-based applications including Modern
Medical Informatics Services which is easier, faster and less expensive. The required system implements the
suitable informatics and electronics solutions efficiently for the Tele-medicine care. We proposed an approach
to manage different multimedia medical databases in the telemedicine system. In order to be efficiently and
effectively manage, search, and display database information, we define an information package for both of
doctor and patient as a concise data set of their medical information from each visit. The methodology for
accessing various types of medical records will be provided, also we will design two web-based interfaces, high-
quality data and display for many medical service purposes.
Keywords: Datamining, Telemedicine, Medical Services, Databases.

I. INTRODUCTION can find that the integrated telemedicine system


consists of the following five subsystems:
There are shortages of medical resources in rural
areas or geographically isolated regions, so many 1) Acquisition subsystem;
physicians may be reluctant to serve in these areas. 2) Viewing subsystem;
Therefore, people who live there will receive lower 3) Teleconferencing subsystem;
medical care than those who live in urban areas. 4) Communication subsystem;
There is an important need to develop a telemedicine 5) Database management subsystem.
system to improve the quality of medical services
there and provide more educational opportunities to The first subsystem is the acquisition subsystem
the physicians in these areas [1] [4].Telemedicine can which collects multimedia information [12] then
be defined as the providing of medical services over a converts it to a standard format (e.g., DICOM 3.0
distance. The Archiving and Communication System [13]). The second one isthe viewing subsystem which
(PACS) will be used in the telemedicine process as displays and manipulates the images and other
this service requires patient history, medical images, medical information [14] [15]. The third one is the
and related information. By using PACS [5] [11], we teleconferencing subsystem which allows face-to-
face interactive conference between physicians in

CSEIT1183818 | Received : 05 Feb 2019 | Accepted : 15 Feb 2019 | January-February -2019 [ 5 (1) : 319-321 ]
319
T. K. Anusuya, P. Maharajothi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 319-321

rural areas and medical centers [16][18],this web management services, namely—
database
subsystem is not included in a PACS. fragmentation, data distribution, websites clustering,
distributed caching, and database scalability.
II. LITERATURE SURVEY
IV. PROPOSED METHODOLOGY
Recently, many researchers have focused on
designing web medical database management systems Our approach integrates three enhanced computing
that satisfy certain performance levels. Such services techniques namely, database fragmentation,
performance is evaluated by measuring the amount of network sites clustering and fragments allocation.
relevant and irrelevant data accessed and the amount We propose an estimation model to compute
of transferred medical data during transactions’ communications cost which helps in finding cost-
processing time. The medication reminder effective data allocation solutions. We perform both
synchronization system provides a patient with external and internal evaluation of our integrated
medications as prescribed by medical staff. In approach.
addition, medical staff can remotely send messages to
the system in order to change the medication In our proposed system we develop a fragmentation
schedules or device configuration settings embedded computing service technique by splitting
in the medication reminder. The proposed system telemedicine database relations into small disjoint
supports the OMA (open mobile alliance) DS (data fragments. This technique generates the minimum
synchronization) protocol, which was originally number of disjoint fragments that would be allocated
proposed as a DS standard protocol that synchronizes to the web servers in the data distribution phase. This
data between mobile devices and a central server. In in turn reduces the data transferred and accessed
this study, the OMA DS protocol is redefined and through different websites and accordingly reduces
extended to transmit the patient’ s medication status the communications cost.
data and the device configuration data.
V. IMPLEMENTATION RESULTS
III. EXISTING PROCESS

Recently, many researchers have focused on


designing web medical database management systems
that satisfy certain performance levels. Such
performance is evaluated by measuring the amount of
relevant and irrelevant data accessed and the amount
of transferred medical data during transactions’
processing time.

Several techniques have been proposed in order to


improve telemedicine database performance, • Doctor Registration
optimize medical data distribution, and control • Patient Registration
medical data proliferation. These techniques believed • Telemedicine services Activation
that high performance for such systems can be • Doctor Manage telemedicine services & their
achieved by improving at least one of the database setting

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


320
T. K. Anusuya, P. Maharajothi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 319-321

• Doctor Manage booking requests [3]. R. K. Taira, B. K. Stewart, a nd U. Sinha, PACS


• Patient add basic information & medical history database architecture and design, Comput.
• Patient makes booking &reviews his booking list Med.
[4]. Imag. Graph., vol. 15, no. 3, pp. 171-176, 1991.
VI. CONCLUSION [5]. S. Badaoui, V. Chameroy, andd F. Aubry, A
database manager of biomedical images, Med.
We perform both external and internal evaluation of Inform., vol. 18, no. 1, pp. 23-33, 1993 .
our integrated approach. In the internal evaluation, [6]. F. Pinciroli, C. Combi, and G. Pozzi,
we measure the impact of using our techniques on ARCADIA: A system for the integration of
WTDS and web service performance measures like angiocardiographic data and images by an
communications cost, response time and throughput. object- oriented DBMS, Comput. Biomed.Res.,
In the external evaluation, we compare the vol. 28, pp. 5-23, 1995.
performance of our approach to that of other [7]. Karmouch, Multimedia d istributed cooperative
techniques in the literature. The results show that system, Comput. Commun., vol. 16, pp. 568–
our integrated approach significantly improves 580, Sept. 1993.
services requirement satisfaction in web systems. [8]. R. E. Rakel, Textbook of Family Practice, 5th
This conclusion requires more investigation and ed. Philadelphia, PA: Saunders, 1995.
experiments. [9]. Digital Imaging and Com munications in
Medicine (DICOM) Version 3.0, A mer. College
VII. FUTURE ENHANCEMENT Radiologists/Nat. Elect. Manufacturers Assoc.,
1993.
As future work we plan to investigate our approach [10]. C. T. Liu, C. C. Lin, J. M. W ong, S. K. Chiou, R.
on larger scale networks involving large number of S. Chen, J. H. Chen, S. M. Hou, and T. Y. Tai,
sites over the cloud. We will consider applying [11]. Design and evaluation of a telediagnosis system,
different types of clustering and introduce search Biomed. Eng. Applicat., Basis Commun., vol. 9,
based technique to perform more intelligent data pp. 52-60, Apr. 1997.
redistribution. Finally, we intend to introduce [12]. H. S. Chen et al., Integrated medical
security concerns that need to be addressed over data informatics with small group teaching in
fragments. medical education

Cite this article as :


VIII. REFERENCES
T. K. Anusuya, P. Maharajothi, "A Survey of
[1]. T. Kitanosono, Y. Kurashita, M. Honda, T. Telemedicine Services using Datamining",
Hishida, H. Konishi, M. Mizuno, and M. A nzai, International Journal of Scientific Research in
The use of multimedia in patient care, Comput. Computer Science, Engineering and Information
Methods Programs Biomed., vol. 37, pp. 259– Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
263, 1992. 5 Issue 1, pp. 319-321, January-February 2019.
[2]. S. T. C. Wong and H. K. Huang A hospital Available at doi :
integrated framework for multimodaliity image https://doi.org/10.32628/CSEIT1183818
base management, IEEE Trans. Sy st., Man, Journal URL : http://ijsrcseit.com/CSEIT1183818
Cybern. A, vol. 26, pp. 455–469, July 1996.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


321
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195171

Medical Product Verify and Complaint Management System


for Ngo’s
R. Gomathijayam *1, J. Ayeesha Parveen2
*1Department of Computer Application, Bon Secours College for Women, Thanjavur, TamilNadu, India
2Department of Computer Science, Bon Secours College for Women, Thanjavur, TamilNadu, India

ABSTRACT

Today’s world belongs to information technology. Information is worthwhile only if, it can be accessed at the
right time, by the right person & is useful for the purpose defined. The purpose of Medical Product Complaint
Management System (MPCMS) is to provide their prescription details and complaint registration easier one.
This is an important tool to control and effectively manage the entire operation of patient details management
and complaint registration in a very efficient manner and take it to the path of growth. The System deals with
the collection of patient details and the prescription details and outputting it to the user through web
application. The prescription details are stored into the database and when the user logs into the system they
can view their prescription details. The user has been given a login to view the details and he can able to
register the complaint in case of severe problems occurred due to the wrong prescription of medicines. The
complaint will be submitted and it will be viewed by the NGO. The NGO is having separate login to view the
complaints and he will be given the authority to accept/reject the complaint.
Keywords : Medical Product, Complaints, NGO, Prescription, Patient, MPCMS (Medical Product Complaint
Management System).

I. INTRODUCTION governments, foundations, businesses, or private


persons. Some avoid formal funding altogether and
Effective complaint system can improve the are run primarily by volunteers. NGOs are highly
patient’s safety and leads to reduction in doing diverse groups of organizations engaged in a wide
mistakes. The complaint management systems are range of activities, and take different forms in
therefore important to help improving the safety different parts of the world. Some may have
and livelihood of patients. This system is for charitable status, while others may be registered for
complaint registering and acceptance of complaints tax exemption based on recognition of social
by the NGO’s therefore the users can be able to purposes. Others may be fronts for political,
register the case easily and therefore the necessary religious or other interest groups.
action be taken. A non-governmental organization
(NGO) is not the term commonly used for an The system includes separate module for user and
organization that is a part of neither a government NGO logins. Each login is given authentication as
nor a conventional for-profit business. Usually set the system involves handling medical data. In the
up by ordinary citizens, NGOs may be funded by routine functioning of a hospital, various types of

CSEIT195171 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 322-326 ]
322
R. Gomathijayam, J. Ayeesha Int J Sci Res CSE & IT. January-February-2019 ; 5(1) 322-326

investigations are carried out. Carrying out number his login. The online storage can be implemented
of tests and making the results available promptly is using cloud computing which provides a shared and
very crucial for assessing the patient's medical status secured access to all the resources shared. If a
and deciding on the further course of action. patient is currently admitted then that patient’s
Patient’s medical record data is critical for the current status (whether he is undertaking tests or in
analysis and research purposes and therefore it has OPD etc.). Doctors can have this customized version
to be stored and manipulated in an efficient manner. of application on their tablets based on their
This data includes patient history, observation, speciality that can give them all the patient details
diagnosis and therapeutic conclusions along with in hand which can help manage their schedule by
the tests details and serves as a source of prioritizing each patient.
information for any analysis and research. The
purpose for this storing the medical data is for Doctor can also send pictures and videos of a
utilizing the patient’s medical information and uses patient’s condition to other doctors for expert
it for analysis thereby improving patient care. opinion if required; doctors can give online
prescriptions to pharmacy directly specific to
When the user logs into the system, he can view all particular patient with their patient id. We can
his medical data through ‘medical prescription develop this application using android programming,
details module’. Since the data are stored in case of android programming being open source the entire
severe problems after the treatment, the user can OS is open for usage and extension. Every person
register the complaint as he has the electronic proof. can view the website and should be able to register
Further if the concerned doctor is not available and him/her providing all the necessary details required ,
if there comes the need to take care of the person by the system will provide a Patient’s ID to the patient
some other doctor he can make use of the stored and further actions can be taken based on the this
proof instead of searching the prescriptions and patient ID .A patient can take the appointments
medical data from the form. Even if third party tries online and know the availability of the doctor.
to login / alter the data he will be prompted to the Notifications about regular health check-ups and
error page as there were strong validations in the vaccinations can be sent to the patients and doctors.
system. Website also contains page for online journals
II. LITERATURE SURVEY which include all the latest medical advances and
also medical breakthroughs, unique cases for study
A. Functional Description of Online Medical and videos featuring several surgeries and research.
Management System Using Modern The user can also check for tips. Using web
Technologies: Priyanka Patil, Sruthi technology it can be easy to provide all the above
Kunhiraman, Rohini Temkar, VES Institute of functionalities and making the management of
Technology, Chembur, Mumbai patients efficient.

As many web services are made available online,


B. Handling e-complaints in customer complaint
Almost every field is made online, web based
management system using FMEA as a
applications can provide a boon to hospital
qualitative system:
management. The system should incorporate many
things online that include the following Most Negative customer perception often manifest as
importantly maintaining the patient’s records in formal companies; while unknown percentages do
details including his disease, history, reports etc. not become so visible. Within a customer complaint
which the doctor can access from anywhere using management system, it is possible through careful
Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com
323
R. Gomathijayam, J. Ayeesha Int J Sci Res CSE & IT. January-February-2019 ; 5(1) 322-326

analysis to identify the flaws in the service causes of online complaints; (2) seek effective ways
standards, in the complaint handling procedure of handling customer complaints by examining
itself, and the finer points which help to formulate different product types; and (3) provide guidelines
resolution methods. When major key problems for successful e-CRM. 1000 customer complaints
persist, complaints may spread across a range of from three different publicized e-business customer
company activity, from operational employee service centres and 500 complaints from online
performance to managerial handling of both policy feedback systems were analyzed in this study. The
and practice. A poorly managed and/or designed research findings suggest that e-businesses should (1)
complaint system will inevitably impact on the provide excellent online customer services because
company's profit and loss status, through excessive customer service is the most important factor in
expenditures, damage to market reputation and online customer satisfaction; (2) respond to
declining income. As the ultimate goal of a customers' requests/complaints quickly because the
company is to content customers, acquire an response speed is more important in online
efficient profit margin, and have a long-lasting customer satisfaction than offline; and (3) employ
relationship with customers and improving market strategies that are appropriate for the product
shares. They should handle the complaints within category in question.
complaint management system and this system
alone must be under the monitoring of customer D. A Multimedia Medical Report Management
relationship management system. The aim of this System:
study is to identify these irregularities and to
The goal of the MrBrAQue (Medical Report Browse
propose a strategy to minimize them. For improving
and Query) project is to develop a system that
reliability of the company we benefited from failure
enables effective management of a large collection
mode and effect analysis technique (FMEA) which
of multimedia documents. The tools for consulting
lets preventive analysis of failures that may
the report collection are presented in this paper.
compromise credibility of the company. As a
Our proposal is based on a paradigm, which we call
methodology model set of 30 questionnaires have
orthogonal navigation that allows a user to
been assigned for this reason and a case study of an
dynamically build a hypermedia network of
airline has been chosen. Ultimately, using, 5-point
documents by means of queries over the document
likert scale and SPSS software, we were managed to
collection
administer the data and conclude the results.

E. Design of Medical Management Information


C. An analysis of online customer complaints:
System Based On Soa:
implications for Web complaint management:
Medical retailers face the challenge of integrating
How businesses resolve customer-complaining
their legacy systems within their stores and at the
behaviour effectively has been considered a
enterprise (or corporate) headquarters. This
"defensive marketing" strategy or a "zero-
challenge becomes even more critical due to
defections" strategy, which diminishes customer
demanding customers who expect a rich experience
dissatisfaction. Handling customer dissatisfaction
both in the store and in other channels. In this
accompanies Web customer complaint management,
paper, a few common scenarios and use cases in
which might be the critical issue for online
retail have been discussed. The paper puts forward a
customer service solutions and e-CRM (electronic
new design method of medicine management
customer relationship management). In this paper;
information system based on SOA. This new system
the authors (1) investigate the current sources and
Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com
324
R. Gomathijayam, J. Ayeesha Int J Sci Res CSE & IT. January-February-2019 ; 5(1) 322-326

model makes full use of legacy applications and is • Less Time constraint for registering complaints.
flexible for future extensions. Experiment shows • Once the complaint is accepted, the patient can
that real-time, intelligent performances and register the case so that they can be given some
efficiency of operation are obviously improved. fund in return to the mistakes done.
F. Clinical Database Management Software
(CDMS) for medical, diagnostic and research
centres:

The Clinical Database Management Software


(CDMS) is a state-of-the-art concept in integrated
medical report recording and analysis for the
medical and research fraternity. It revolutionizes
the archiving and analysis of patient data and
replaces the traditional passive and manual method
of collecting patient data in text format with an
active, online, easy-to-use pop-up and pull-down
series of menus in an interactive mode in a very
structured format. This new concept is the outcome
of years of dedicated and continuous interaction
with leading medical practitioners, research
scientists and software specialists. Lacunae in the
collection of patent clinical and research data and Figure 1 : System Architecture
its management should now be a matter of the past.
Flexibility and cost-effectiveness are the most IV. CONCLUSION
important aspects of CDMS. The software is easily
adaptable to a variety of database requirements and, Thus the above mentioned system has been
with a minimum of modification of the source code, developed to satisfy all the requirements of the user.
it can be configured to meet the very specific needs Since it gives the user an easy way to view
of the medical and research fraternity. prescription details and to register complaints it
takes less time constraint to make the needful. The
III. METHODOLOGY system is simple and easy to implement. Since there
is high level of authentication, only right person can
In the proposed system, everything will be online. view and edit the details. By implementing this
The patients will have separate login to avail system, users can attain maximum benefit with less
services. The prescription details for that concerned time constraint and energy.
patient will be available to them in their Medical
Prescription Details module. The system provides V. REFERENCES
high security by granting access rights to change the
sensitive data. [1] An analysis of online customer complaints:
• The prescribed data will be secured since implications for Web complaint management
authentication has been provided to view it. -Yooncheong Cho ; Rutgers Univ.,
• Medical complaint forms are there for Piscataway, NJ, USA ; Il Im ; Hiltz, R. ;
registering the complaints in case of severe side Fjermestad, J.
effects.
Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com
325
R. Gomathijayam, J. Ayeesha Int J Sci Res CSE & IT. January-February-2019 ; 5(1) 322-326

[2] Handling e-complaints in customer complaint qualitative system, Faed, A., DEBII, Curtin
management system using FMEA as a Univ. of Technol., Perth, WA, Australia
qualitative system - Faed, A. ; DEBII, Curtin
Cite this article as :
Univ. of Technol., Perth, WA, Australia
[3] A multimedia medical report management
R. Gomathijayam, J. Ayeesha Parveen, "Medical
system ,Merialdo, P. ; Dipartimento di Inf. e
Product Verify and Complaint Management System
Autom., Rome Univ., Italy ; Sindoni, G.
for Ngo's", International Journal of Scientific
[4] Functional Description of Online Medical
Research in Computer Science, Engineering and
Management System Using Modern
Information Technology (IJSRCSEIT), ISSN : 2456-
Technology - Priyanka Patil, Sruthi
3307, Volume 5 Issue 1, pp. 322-326, January-
Kunhiraman, Rohini Temkar,VES Institute of
February 2019. Available at doi :
Technology, Chembur, Mumba
https://doi.org/10.32628/CSEIT195171
[5] Handling e-complaints in customer complaint
Journal URL : http://ijsrcseit.com/CSEIT195171
management system using FMEA as a

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


326
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195191

A Survey on Child Safety Wearable Device to Prevent Child


Trafficking Using Arduino
Elakiya M.1, S. Radhika2
1Master of Engineering (ME), Department of Computer Science and Engineering, RMK Engineering College,

Chennai, India
2 Assistant Professor, Department of Computer Science and Engineering, RMK Engineering College,
Chennai, India
ABSTRACT

Currently a child’s security is a crucial space of concern. This paper is developed to rectify the troubles parents
relating to their child security. In keeping with a story from 2014, about 135,000 kids are calculable to be
trafficked in India annually. Trafficked children are sold-out into slavery, domestic slavery, beggary, and
therefore the sex trade. This paper provides the conception of sensible wearable devices for our little ones. And
to stop kid trafficking. One amongst the most precedence over this device is that it provides us the tracing
details of our little ones. And it doesn’t require a smart phone or a technician person to know and use it. The
motive of this device is to assist find their kids with none difficulties. At the moment there are several tracking
device that helps us to note our children’s activity with ease and additionally notice the kid using Wi-FI and
Bluetooth. However Wi-Fi and Bluetooth seem to be unpredictable medium of communication between the
parent and kid. So the main focus of the paper is to possess an SMS text enabled communication between the
child’s wearable and also the parent because the GSM mobile communication is nearly present everyplace. The
parents will send a text with specific keywords like “LOCATION” “TEMPERATURE” “UV” “SOS” “BUZZ” ETC.
The tracking device can reply back providing the correct location of the kid and it will navigate through google
maps. So this paper provides parents with a way of security for his or her kid in today’s world.
Keywords : Child Safety Wearable Device, Arduino, Trafficked children, Lynceus2Market, RFID, MASS, RSSI,
DCC, ATNAC

I. INTRODUCTION the other describing the issues faced by these


fishermen in their day to day commute to the oceans.
Fishing is one of the chief sources of food and income In order to solve this issue, the governments of these
for almost all coastal lands irrespective of its countries decided to have a common territory of
geological location in the earth. Since it has an ocean as international waters and that it would act as
important role to play in the economy of a country, the region common as well as a separation between
there is no doubt that neighboring countries sharing the two lands. But even this did not prevent the
the same oceans frequently engage in disputes in fishermen from unknowingly wandering off into the
regards to ownership of the area. This has resulted in other country’s waters. Hence, there is a strong need
deep problems to the fishermen community residing to device methods to prevent this from happening
in the coastal regions of these countries. Often, we and save the fishermen from severe punishments and
hear in the news and see in the papers, one article or border disputes. Our model helps to ensure the

CSEIT195191 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 327-333 ]
327
Elakiya M., S. Radhika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 327-333

location of the fishing boats through a new Ultra low-power UWB-RFID system for precise
technology using radio waves. location-aware applications
Ultra low-power radio-frequency identification
(RFID) tag with preciseness localization is commonly
the sanctioning technology for location-aware
detector applications. The RFID tag could be a
transmitter comprising of a micro-controller board
and a UWB impulse radio board. Power saving and
precision localization is achieved by improvement of
the circuit style for immoderate short pulses further
as system design and operation.

Petrovic and R. Kanan, “Extremely low power indoor


localization system,” in IEEE 8th International
II. LITERATURE SURVEY Conference on Mobile Adhoc and Sensor Systems
(MASS), pp. 801–806, Oct. 2011.
“Lynceus2Market- An innovative individuals
localization system for safe evacuation of huge Extremely Low Power Indoor Localization System
traveler ships” In this paper style, implementation and testing of an
An innovative individual localization system for safe very low power and low price wireless detector
evacuation of huge traveler ships network for observing person or object position in an
interior atmosphere are mentioned. The localization
This paper is mainly based on the maritime disasters relies on the standard received signal strength
to control the cause, during emergency evacuation of intensity (RSSI) technique. The very low-power
large passenger ships. The Lynceus2Market project feature is achieved by employing a new Duty-Cycle
addresses this challenge through delivering a dominant (DCC) technique. The DCC technique
revolutionary operational system for safe evacuation relies on the frequency Identification (RFID)
supported innovative people localization principle to wake-up the mobile unit at the correct
technologies. time and save energy once it's not used.

This technology will be provided with SME –driven S. Nasrin and P. 1. Radcliffe, "Novel protocol allows
market segments of smoke alarm, fire detection DIY home automation," Telecommunication
systems, life saving equipment, emergency Networks and Applications Conference (ATNAC),
management decision support systems and assistive Novel Protocol Enables DIY Home Automation.
search and rescue equipment.
In this paper, we tend to describe a brand new design
Y. Zhou, C.L. Law and J. Xia, “Ultra low-power for a home automation system that is constructed on
UWB-RFID system for precise location-aware novel network protocols.
applications,” in Wireless Communications and
Networking Conference Workshops (WCNCW), pp. The new system permits a user or home-owner to
154–158, April 2012. buy off the shelf devices and management them by a

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


328
Elakiya M., S. Radhika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 327-333

computer or mobile device terribly} very easy Objective measures of response throughout these
manner. phases could facilitate determine kids in danger for
The system does not require expert configuration or a anxiety, we tend to propose an alternate technology,
central controller therefore the cost will be an inertial measurement unit (IMU), that permits
significantly reduced. The key enabling technology non-invasive activity of the movements related to
that makes this possible is an XML definition threat response.
Of the device capabilities, show needs, and
management signals. Internet of Things: a definition & taxonomy: Bruno
Dorsemaine, Jean-Philippe Gaulier, Jean-Philippe
Md Imdadul Hoque,1,* Abdullah Al Amin,1 Md Wary and Nizar Kheir
Jumshadur Rahaman,1 and Mohammad Hossam-E-
Haider1 1Electrical Electronic & Communication Internet of Things: a definition & taxonomy
Engineering Department Military Institute of Science
and Technology, Dhaka, Bangladesh The Internet of Things (IOT) has varied fields of
Designing and Performance analysis of GSM/GPS application including health care, resource
based mostly heavier-than-air craft following Device management, asset tracking, etc. depending on the
utilization case, varied technologies like RFID,
Low flying aircrafts are always powerful to trace Wireless sensing element Network (WSN) or smart
because of the radar horizon limitation. However an Objects is used. With each of these comes a specific
aircraft is often to be tracked for the security purpose vision of what the IoT and connected objects are and
.The combination of world Positioning System (GPS) – to our information – there's no global image of the
and international System for Mobile IoT.
Communications (GSM) also can perform the duty
satisfactorily. During this work, two GSM networks Jun Zheng David Simplot-Ryl Chatschik Bisdikian
were utilized to trace low flying aircrafts in several Hussein T. Mouftah
routes. The device style is projected basing on the
experimental results of chase. GUEST EDITORIAL – THE INTERNET OF THINGS

Ellen W. McGinnis*, Ryan S. McGinnis*, Member, The growth of the internet shows no signs of
IEEE, Maria Muzik, Jessica Hruschak, Nestor L. retardation down and it steady becomes the cause of
Lopez-Duran, Noel C. Perkins, Kate Fitzgerald, a brand new pervasive paradigm in computing and
Katherine L. Rosenblum communications. This new paradigm enhances the
standard internet into a smart net of Things (IoT)
Movements indicate threat response phases in created around intelligent interconnections of
children at-risk for anxiety numerous objects within the physical world, like
vehicles, cell phones, habitats, and habitat occupants.
Temporal phases of threat response similarly as It utilizes affordable information gathering and
Potential Threat (Anxiety), Acute Threat (Startle, dissemination devices, like sensors and RFID tags,
Fear), and Post-threat Response Modulation area unit that facilitate fast interactions among the objects
referred to as underlying markers of tension themselves as well because the objects and persons in
disorders. anyplace and at any time. IoT isn't a mere extension
of today’s net or an online of net systems.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


329
Elakiya M., S. Radhika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 327-333

Virtual Rehabilitation with Children. Challenges for


clinical adoption. By Stephanie Glegg.
Virtual Rehabilitation with Children III. COMPONENTS REQUIRED

The main draw of virtual rehabilitation is that the A. GPS


capability it allows clinicians to regulate the
The Global Positioning System (GPS) is a U.S. space-
characteristics of the virtual surroundings whereas
based global navigation satellite system.
modifying the degree of challenge to suit individual
patient desires. As an example, by reducing the
It provides reliable positioning, navigation, and
amount and speed of visual stimuli, removing visual
temporal order services to worldwide users on a
clutter and additive feedback, and directional stimuli
nonstop basis day and night, anyplace on or close to
to 1 aspect of the body only, therapists will modify a
the planet.
child with a brain injury to participate in virtual
reality in a way that optimizes success and recovery.
B. GSM MODEM
Augmented feedback (e.g., visual, auditory, and
haptic) and opportunities for motivating repetitive GSM electronic equipment may be wireless
practice produce this approach considerably electronic equipment that works with a GSM wireless
appealing for motor learning applications network.

ChildGuard: A Child-Safety Monitoring System A wireless modem behaves like a dial-up modem.
Zhigang Gao, Hongyi Guo, Yunfeng Xie, and Yanjun The main distinction between them is that dial-up
Luo Hangzhou Dianzi University Huijuan Lu and Ke electronic equipment sends and receives information
Yan China Jiliang University. through a set telephone line whereas wireless
electronic equipment sends and receives information
ChildGuard: A Child-Safety Monitoring System through radio waves.

The ChildGuard system exploits mobile devices to The operating of GSM electronic equipment relies on
observe the period of time movement of commands.
unsupervised children. It additionally sends
reminders and alarms to kids and notifies guardians C. LCD
of abnormities in a child’s routine. for instance, two
typical measures embody hiring manned guards or LCD (Liquid Crystal Display) screen is associate
using video surveillance, but guards aren’t practical electronic show module and notices a large vary of
or affordable when it involves monitoring kids, and applications.
video surveillance systems typically have blind zones.
to that end, ChildGuard may be a security A 16x2 {lcd liquid crystal display |LCD| digital display
methodology for observation kids that uses present |alphanumeric display} display is incredibly basic
computing devices, like smartphones or wearable module and is extremely unremarkably employed in
devices—which are growing in each quality and varied devices and circuits.
performance.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


330
Elakiya M., S. Radhika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 327-333

These modules area unit most popular over seven The sensors contain two in contact with an
phases and alternative multi segment LEDs. electrolyte.
The reasons being: LCDs are economical; easily The electrodes are usually invented by fixing a high
programmable. extent valuable on to the porous hydrophobic
membrane.

The perating conductor contacts each the solution


and also the close air to be monitored typically via a
porous membrane.

D. TEMPERATURE SENSOR - The LM35

The LM35 is associate degree computer circuit


detector which will be accustomed live temperature
with associate degree electrical output proportional
to the temperature (in oC)

The LM35 - An Integrated Circuit Temperature


Sensor
Why Use LM35s To Measure Temperature?
F. ARDUINO UNO
1.You can measure temperature more accurately than
a using a thermostat. Arduino/Genuino Uno is a microcontroller board
2.The sensor circuitry is sealed and not subject to supported the ATmega328P. It has fourteen digital
oxidation, etc. input/output pins (of that six will be used as PWM
outputs), six analog inputs, a sixteen MHz quartz, a
USB association, an influence jack, an ICSP header
and a reset button. It contains everything required to
support the microcontroller; merely connect it to a
laptop with a USB cable or power it with a AC-to-DC
adapter or battery to urge started.

G. ARDUINO IDE:

The ATmega328 on the Arduino/Genuino Uno comes


preprogrammed with a boot loader that allows you to
upload new code to it without the use of an external
hardware programmer. It communicates
victimization the first STK500 protocol (reference, C
header files).
E. GAS SENSOR

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


331
Elakiya M., S. Radhika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 327-333

radiation and SOS light with alarm buzzer for their


child's surroundings and the ability to locate their
H. EMBEDDED C child or alert someone in acting to rescue or comfort
the child. The smart child safety wearable can be
Embedded C may be a set of language extensions for
enhanced much more in the future by using highly
the C programming language by the C Standards
compact Arduino modules such as the LilyPad
committee to handle commonality problems that
Arduino which can be sewed into fabrics. Also a
exist between C extensions for various embedded
more power efficient model will have to be created
systems.
which will be capable of holding the battery for a
longer time.
Historically, embedded C programming needs
nonstandard extensions to the C language so as to
V. REFERENCES
support exotic options like fixed-point arithmetic,
multiple distinct memory banks, and basic I/O
[1]. "Lynceus2Market - An innovative people
operations.
localisation system for safe evacuation of large
passenger ships." Online]. Available:
I. Power Supply:
http://www.lynceus-project.eu/ .
This section describes how to generate +5V DC [2]. Y. Zhou, C.L. Law and J. Xia, "Ultra low-power
power supply UWB-RFID system for precise location-aware
applications," in Wireless Communications and
Networking Conference Workshops
(WCNCW), pp. 154–158, April 2012.
[3]. D. Petrovic and R. Kanan, "Extremely low
power indoor localisation system," in IEEE 8th
International Conference on Mobile Adhoc and
Sensor Systems (MASS), pp. 801–806, Oct.
2011.
[4]. N. Simon et al., "Indoor localization system for
emergency responders with ultra low-power
radio landmarks," in Instrumentation and
IV. CONCLUSION AND DISCUSSION Measurement Technology Conference (I2MTC)
,pp 309-314, may 2015
This paper briefly analyzes and reviews the related [5]. S. Nasrin and P.I Radcliffe, " Novel protocol
work carried out for child safety. When compared enables DIY home automation,"
with all the technologies and paper we finally prefer Telecommunication Networks and Applications
using SMS text enabled communication between the Conference (ATNAC) , 2014 Australasian,
child’s wearable and also the parent because the GSM Southbank, VIC, 2014, pp. 212-216.
mobile communication is nearly present everyplace. [6]. Internet of Things: a definition & taxonomy:
The usage of this can make us track the child’s Bruno Dorsemaine, Jean-Philippe Gaulier,
location easily. Only GSM/ GPS is used to track the Jean- Philippe Wary and Nizar Kheir.
child. It provides parents with the real-time
location, temperature around their child, UV

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


332
Elakiya M., S. Radhika Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 327-333

[7]. EDITORIAL- THE INTERNET OF THINGS:


Jun Zheng David Simplot- Ry1, Chatschik
Bisdikian and Hussein T.Mouftah
[8]. Virtual Rehabilitation with children.
Challenges for clinical adoption. By Stephanie
Glegg
[9]. ChildGuard: A Child-Safety Monitoring System
Zhigang Gao, Hongyi Guo, Yunfeng Xie, and
Yanjun Luo Hangzhou Dianzi University
Huijuan Lu and Ke Yan China Jiliang
University.
[10]. Guest Editorial: Mobile wearable
communications Hassnaa Moustafa, Holger
Kenn, Kamran Sayrafian, William Scanlon and
Yan Zhang.
[11]. Ellen W. McGinnis*, Member, IEEE, Maria
Muzik, Jessica Hruschak, Nestor L. Lopez-
Duran, Noel C. perkins, kate Fitzgerald,
Katherine L.Rosenblum "Movements indicate
threat response phases in children at risk for
anxiety"
[12]. Md Imdadul Hoque,1,* Abdullah Al Amin,1 Md
Jumshadur Rahaman,1 and Mohammad
Hossam-E-Haider1 1Electrical Electronic &
Communication Engineering Department
Military Institute of Science and Technology,
Dhaka, Bangladesh "Designing and
Performance Evaluation of GSM/GPS Based
Helicopter Tracking Device"

Cite this article as :

Elakiya M., S. Radhika, "A Survey on Child Safety


Wearable Device to Prevent Child Trafficking Using
Arduino", International Journal of Scientific Research
in Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 327-333, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195191
Journal URL : http://ijsrcseit.com/CSEIT195191

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


333
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195192

Bus Notification System Using Crowdsourcing


B. Harshini Reddy, K. Jhansi Rani, M.Vaishnavi
Information Technology, SNIST, Ghatkesar, Telangana, India
ABSTRACT

Most of the daily commuters of RTC face problems to know the exact time of arrival of a bus to a particular stop.
The time of arrival of bus depends on many factors like traffic, condition of a bus, climate etc. So it is very
tough for a person waiting for a bus without knowing whether it has left or not from a particular stop. So here
we are providing a solution for this problem by using latest technologies like an Android Application. The
solution is that we want to develop an ANDROID APP[1] which gives the notification of a bus arrival in a
particular route for a particular stop. The backbone of our application is using CROWDSOURCING. It is
nothing but taking the help of some people to help other people. The passengers who have seen or boarded the
bus can report the bus location along with an image of bus with the help of our app. Then this location will be
sent as a notification to all other commuters in that particular route. The commuters who have missed the bus
from the stops through which the bus has already passed, can go with an alternative option and the other
commuters waiting at the stops to which bus has to arrive can wait accordingly. This helps in saving our
precious time which many of us wasting it in waiting for a bus. In this way our application helps the daily
commuters to know the location of a bus and saves their precious time.
Keywords : Crowd Sourcing, Androidapp, Notification, Location

I. INTRODUCTION ordinates of the bus and send information to the


server.
Most of the daily commuters of RTC are facing
problems to know at which time bus will arrive or it The solution is that we want to develop an
has gone from particular stop. And the time of arrival ANDROID APP which gives the notification of a bus
of bus depends on traffic, condition of a bus, climate arrival in a particular route for a particular stop. The
etc..so, it is tough for a person to wait for a bus. backbone of our application is using
CROWDSOURCING[3]. It is nothing but taking the
So here we are with a solution, to this problem by help of some people to help other people. The
using Bus notification System with an android passengers who have seen or boarded the bus can
application. Our main motto is that the person report the bus location along with an image of bus
receives the exact location of the bus with the with the help of our app. Then this location will be
estimated departure time of the requested bus[7]. To sent as a notification to all other commuters in that
get into this process the person needs to register particular route.
himself in the application. Once the person gets
himself registered he can receive the notifications The commuters who have missed the bus from the
with the help of VMD(Variable Message stops through which the bus has already passed, can
Display)[2].This VMD is used to track the co- go with an alternative option and the other

CSEIT195192 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 334-339 ]
334
B. Harshini Reddy et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 334-339

commuters waiting at the stops to which bus has to Initial review which is organized is for proving that
arrive can wait accordingly. This helps in saving our problem regarding buses being late to bus stops exists.
precious time which many of us wasting it in waiting At the same time, we assessed the level of significant
for a bus. In this way our application helps the daily of the problem. We observed the expected arrival
commuters to know the location of a bus and saves time to each bus stop for a period of time for two
their precious time. weeks. The result of the observation is that most of
the time the bus failed to arrive at the bus stop as
II. LITERATURE SURVEY they are scheduled to arrive. Furthermore, we had
got a shocking observation result that is sometime the
For any enterprise or task, the initial phase would be bus skipped its trip and causing the passengers to wait
literature survey concerning with formerly issued for the next bus that might arrive an hour later.
journals. Prior to decision making, a thorough Sometimes the bus arrives late at the first bus stop for
patrons review regarding journals is done in a more than fifty minutes. It is causing the passengers
sequence manner. Complete knowledge and in that entire route to be delayed and arrive at late to
interpretation concerning the enterprise would be their workplaces. Also, the problem even happens
the main principle of this literature review. It also during peak hours where most of the passengers will
helps us lay a firm base for upcoming tasks. be rushing to their workplaces. These observation
Research which is undergone serves as guideline and results successfully proved that our problem
citation to the enterprise. Once the literature analysis statement is valid and these problems have to be
is done, we should be in a position to write to such a solved urgently for the benefit of the city and the bus
degree that manifests our experience and observation passengers. The second study is via survey. The main
regarding that particular domain and realize principal aim of this survey is to help understand the current
subject which are pertinent for the project. It’s demand and acceptance for our proposed solution in
important to have an idea about the one’s that can be the market. In order to understand the culture of the
ignored and must be in a position to predict end passengers, the problems faced by the passengers, and
result. the need of our proposed system, and the present
By summing up wholly, undeniably the intention of condition of public transportation we conducted a
literature survey is to find out the oversight survey by distributing hundred survey forms to the
pertaining to a task. The SMS Based Forecasting passengers who travel frequently through these
System of Public Transport Arrival Time[5]. public transportation.
After issue being identified and then coming up with According to our survey results the majority of the
suitable solving methodology we organized some commuters of the public transportation are
research concerning bus passengers stating our issue workforce and students who use this for travelling to
explanation, also explaining our recommended their respected workplaces daily. Also the most
solution. Three main segments are concentrated in common and major problem that is that the buses
our analysis. They are: usually do not arrive to the bus stops according to
1. Demonstrate as well as appraise the importance their scheduled time. Due this problem there is an
regarding issue description . inconvenience caused to the passengers and they are
2. The ultimatum and necessity of suggested end forced to wait for unpredictable frame of time.
result must be recognized and understood. Majority of the passengers who took part in survey
3.Providing the issue description demonstrated is reported that a SMS system which gives updates
correct, extract and analyse it. about the whereabouts and the arrival timings of the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


335
B. Harshini Reddy et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 334-339

bus would be a better solution for their problems. off. The methods used in this system are
Based on the this survey results and the information onPause(),onResume(),onKeyDown() and
obtained we are very positive that our proposed onDestroy().The Post-configuration process is
solution would be accepted by the bus passengers. managed by onDestroy() and onKeyDown() which
The current status of the problems faced by the makes the usage convenient.
passengers is very high. Thus we derived some
hypotheses to help us to understand the factors The implemented modules in this system are:-
which are affecting the bus’s arrival time. These
hypotheses are extremely important for us to 1. Bus Unit
understand the parameters involved in this system. 2. Central Control Unit
Based on the result of the hypotheses, we are able to 3. Client-Side Application
understand the possible causes of bus delays and this
The bus unit consists of GPS which tracks the co-
information will provide great assistant.
ordinates such as latitude, longitude and other
attributes of that particular location and send s it to
This report is a survey which helps the passengers to
the server. The GPS is free to use i.e.., no subscription
overcome the difficulties they are facing in their
fee is required. The GPS has the capability to
daily transportation. This helps to conserve time
receive information about these locations from three
which in very important. This not only saves the
satellites and calculates the position based on this
time but also it helps to form an advance transport
information. There are three different types of
system.
transceivers in GPS such as Data Loggers, Data
pullers and Data Pushers .Based on the application,
There are many resources present which gives us the
the appropriate transceiver will be chosen to gather
information regarding the bus numbers and its details.
the information and track the position. The GPS
In this review we get advance information like
attached to sends the information to the servers at
vacant seats, location of the bus, estimated time of
regular intervals of time. This information is received
arrival and departure to a particular place which can
by the device from the servers.
be viewed by the travellers.Gradual advancement of
bus systems reaches demands pertaining to various
The received data is analysed by the servers .The GPS
areas. Main motto of the application is to have a
antenna is connected to the right jack and fixes it’s
comfortable, simple and accessible bus facility that
position to receive signals in the appropriate places
uplifts the usage of public transport system.
.There is a slot allocated for a SIM card, the GSM
towers helps in receiving messages from different
III. PROPOSED METHODOLOGY
users and respond to them. The positive and negative
wires are connected to 12V or 24V vehicle power
This model consists of four parts. Each and every
system [4]. The tracker device is turned on to receive
individual model is divided based on their
the signals from the satellite hence thus enabling the
functionalities. The main activity of our application
device to receive the latitude and longitude values of
perform major tasks which include Pre-configuration
the position of the bus. At any point of time the GPS
and Instantiation methods. It consists information
receiver gives us the location values. The coordinates
about the total distance between the two locations
of the bus along with the timestamp is present in the
and also about the previous and current location. The
bus unit, these values are compared with the previous
status of the GPS is also examined whether it's on or

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


336
B. Harshini Reddy et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 334-339

coordinates if there is any distinction then the new and the elements which are in correspondence with
co-ordinate values are updated and will be sent to the the stack are integrated in a tight manner to provide
server over the Internet[4]. the convenient and proper execution frame for
mobile devices.
The location details of the bus are saved in a
particular format with few attributes such as As per the advancement of the technology the
Latitudinal position, longitudinal position, ID and Android is developed for the touch screen devices
timestamp etc. Every bus has a unique ID which like mobile devices and tablets. This android is an
helps to identify different buses easily. Every bus has operating system developed by Google. The android
a different identification number called as ID. Server is based on the Linux kernel[6] whereas the
in this system takes an important role and performs Android's user interface is based on direct
major functions. Here the server stores and maintains manipulation with the help of human touches that
the complete information. The server also performs help to perform actions on the screen and also to give
actions like a gateway between the user and the bus input[5]. In extension to the devices Google added
module. The overall information about the routes, Android TV, Android Wear etc. with an exclusive
arrival time, department time are stored in the new user interfaces. The complication of the mobile
database. world has completely been reduced with the use of
Android. The unimaginable features are implemented
The final user can only view the application which is and high quality phones are made easily with the
an interactive web based module. Here the help of this Android OS.
application provides different services to remote
users. The user module requires two inputs that is the
current location point and the end location point. As
the user requests the server, it sends the list the buses
available in that location. Ones the bus is selected it
shows the real time location and time of that
particular bus on screen.

IV. RESULTS AND ANALYSIS

The GPS of the system is built in such a way that it


tracks the exact location and provides dual role of
communication. The communication is done
between the driver and user. This process takes a
vital role in the development of the application. All
the information is provided below. Here we will Figure 1. Screen 1 of bus notification system
come across an operating system, run-time
environment, and software stack comprising
applications, libraries, middleware and services
which are a part of the Android Architecture.
Visually this architecture can be represented in its
best manner. The layers of the stacks available in it

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


337
B. Harshini Reddy et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 334-339

Figure 2. Screen 2 of bus notification system Figure 4. Login for bus notification system

Figure 3. Registration for bus notification system Figure 5. Location accessing for bus notification
system

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


338
B. Harshini Reddy et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 334-339

method of tracking the location makes the journey


easy.

Currently we do not have sufficient data to enable us


to draw prediction result with weather factor in
consideration. Therefore, we are very positive that
the inclusion of weather factor will increase the
accuracy of PredictBus system’s prediction
significantly. Next, we would like to improve the
system database’s capacity.

Last but not least, we hope to see this application to


be able to merge a few bus routes and predict the
arrival time. This is an important prediction for users
that needs to switches buses in order for them to
arrive at their destination.
Figure 6. Map page for bus notification system
VI. ACKNOWLEDGEMENT
V. CONCLUSION

We would like to thank our Research Guide Dr.


The present application relates to a system and Subhani Shaik, Associate Professor in Department of
methods for notifying passengers of an approaching Information Technology, SNIST, Hyderabad for their
vehicle.Utilizing this application passenger can save continue support and valuable suggestions
their time and safe in controlled environment.And all throughout carried this work. Authors are also
these are done by just installing our application grateful to the reviewer for the renovation of
android app and we can find the location of a bus. manuscript. We would also like to thank the
Department of information Technology providing us
The project is completely done on Android based with the facility for carrying out the simulations.
platform.Various attributes have been implemented
so that the system is very convenient to use with the VII. REFERENCES
best and advance features.the domain use in the
system is SQL.Since the GPS is added it displays maps [1]. https://www.webopedia.com/TERM/A/android_
automatically.The tracking method is also been made app.html
easy with this GPS. [2]. https://en.wikipedia.org/wiki/Variable-
message_sign
As the application is done using cloud soucing,it is [3]. https://en.wikipedia.org/wiki/Crowdsourcing
easily accessible for all the android usrs.the [4]. www.irjet.net
application is useful for both driver-traveller and [5]. www.studymode.com
tourists.This application can also be modified and be [6]. Submitted to University Malaysia Perlis
used for tracking any other means of transport.This [7]. Submitted to Asia Pacific University College of
Technology and Innovation.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


339
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195151

An Extension to Traditional Qwerty Cryptographic Method


V. Vennila*1, S. Umamaheswari2
*1Assistant Professor, Department of Information Technology, Bon Secours College for Women,
Thanjavur, Tamil Nadu, India
2Assistant Professor, Department of Information Technology, Bon Secours College for Women,

Thanjavur, Tamil Nadu, India

ABSTRACT

Cryptography comes from the Greek word kryptos which means hidden. The aim of our research is to provide
secrecy for the data during transmission. In this we discuss about the qwerty cipher algorithm, its merits and
demerits. The existing qwerty algorithm is based on the use of a keyboard which is not much secured. We have
proposed an enhancement to the existing algorithm by using a diagonalized qwerty algorithm which is more
secure and highly difficult for the intruder to understand or decrypt the cipher text, even if it is decrypted the
result would be gibberish.
Keywords : Transmission, Intruder, Decrypt, Cipher, Gibberish.

I. INTRODUCTION
Traditional ciphers consist of substitution or
Cryptography is the strongest tool used for secret transposition techniques [2] [3]. In Substitution
writing which is controlling against the security ciphers, replace letters in the plain text with other
threats [6]. The study of encoding and decoding is letters or symbols keeping the order as same. It maps
called as cryptography [1]. In cryptography, a cipher plain text elements into cipher text elements [4].
is an algorithm consists of well-defined steps for
performing encryption or decryption. In encryption
ABCDEFGHIJKLMNOPQRSTUVWXY
process, the original plain text is coded into the Z
cipher text and in the decryption process; the plain
ZYXWVUTSRQPONMLKJIHGFEDCB
text is restored from the cipher text [2]. We need a A
strong encryption algorithm in order to encrypt the
plain text into cipher text [7]. The sender and Fig 2. Substitution Cipher
receiver must have obtained the secret key in a
secure fashion and must keep the key secure. The In Transposition ciphers, keep all of the original
process of encryption and decryption are shown letters intact, but mix up their order [5]. It transposes
below systematically the position of plain text elements.
Transposition cipher hides the message by
rearranging the order. Transposition cipher, just
reorder the letters [8]. A character in the first
Fig 1. Encryption and Decryption process position of plaintext may appear at 12th position in

CSEIT195151 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 340-342 ]
340
V. Vennila, S. Umamaheswari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 340-342

cipher text, and a character at 6th position in the Decryption Technique:


plaintext may appear at 18th position in cipher text. Cipher text: MGY DHMCM HY EPC CLME
Key: QWERTY
Example of the transposition cipher is given bellow, Plain text: SUN RISES IN THE EAST

IV. PROPOSED DIAGONALIZED QWERTY


CIPHER

The proposed diagonalized qwerty cipher is the


improvement of the qwerty cipher for making the
communication more secure and it is difficulty for
Fig 3. Transposition Cipher the hackers to hack the plain text. Diagonalized
qwerty cipher is much secured when compared to
II. EXISTING QWERTY CIPHER the qwerty cipher.
Step 1. The model of a qwerty keyboard is taken.
QWERTY is the most common modern day keyboard
layout. The name comes from the first six keys
appearing on the top left row of the keyboard and
read from left to right: Q-W-E-R-T-Y. In the existing
qwerty cipher the top line is the keys on a keyboard
from top left to bottom right, and the bottom line is
the alphabet. It uses a seemingly random letters. The Fig 5. Model of a QWERTY keyboard
model of qwerty cipher is given bellow in which the
plain text uses the qwerty keyboard to cipher the Step2. The model of diagonalized qwerty keyboard is
plain text. This cipher is easy to crack. framed.

In the qwerty cipher the keys in the keyboard are


taken in an diagonal manner for framing an
diagonalized qwerty keyboard. Which is more
secured form of communication while transmitting
message from a source to an destination. The
Fig 4. Top box represent keyboard keys and the
mapping of diagonals is shown below.
bottom represent alphabets.
Q W E R T Y U I O P
III. EXAMPLE FOR QWERTY CIPHER

Encryption Technique:
A S D F G H J K L
Plain text: SUN RISES IN THE EAST
Key: QWERTY
Cipher text: MGY DHMCM HY EPC CLME Z X C V B N M

Fig 6. Mapping of qwerty keyboard.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


341
V. Vennila, S. Umamaheswari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 340-342

The diagonalized qwerty keyboard is represented as VII. CONCLUSION


13 keys in each row for the purpose of
encrypting/decrypting a plain text in a secured way. Finally through this paper we have pointed the
traditional qwerty cipher merits and demerits. In
order to overcome the demerits, we have proposed an
extension to traditional qwerty cipher which can be
used more efficiently while encrypting the plain text.
However the used diagonalized qwerty algorithm can
be improved for encrypting large amount of data.

VI. REFERENCES
Fig 7. Representation of qwerty keyboard
[1]. ”Transformation of plain Text with two
V. EXAMPLE FOR DIAGONALIZED QWERTY ciphers”,(Rajani Bala,Swapnika saxena,Sonal
CIPHER beniwal), Volume 3, Issue 5, May 2013.
[2]. “Implementation of Ceasar Cipher with Rail
Encryption Algorithm: Fence for Enhancing Data Security”,(Ajit Singh,
Plain text: CRYPTOGRAPHY IS MY FAVOURITE Aarti Nandal, Swati Malik), Volume 2, Issue 12
SUBJECT , December 2012.
Key: DIAGONALIZED QWERTY [3]. “An Extension to Traditional Playfair
Cipher text: IKZTPFQKBTWZ CN DZ OBLFXKCPJ Cryptographic Method”( Ravindra Babu, Uday
NXAEJIP Kumar, Vinay Babu, Aditya, Komuraiah),
Volume 17, No.5, March 2011.
Decryption Algorithm: [4]. “An Improved Ceasar Cipher
Cipher text: IKZTPFQKBTWZ CN DZ OBLFXKCPJ Algorithm”,(Ochoche Abraham,
NXAEJIP GaniyO.Shefiu), Volume 2, Issue-5,1198-1202.
Key: DIAGONALIZED QWERTY [5]. "Alpha-Qwerty Cipher: An Extended Vigenere
Plain text: CRYPTOGRAPHY IS MY FAVOURITE Cipher”( Khalid Imam Rahmani,Neeta
SUBJECT Wadhwa, Vaibhv Malhotra), Volume 3, No.3,
May, 2012.
VI. ADVANTAGES [6]. William Stalling “Network Security Essentials
(Applications and Standards)”, Pearson
This diagonalized qwerty cipher has various Education, 2004 .
advantages over simple cipher. It is more difficult to [7]. Atul Kahate (2009), Cryptography and Network
cryptanalyze. The result cannot be easily Security, second edition, McGraw-Hill.
reconstructed. Brute force attack cannot crack the [8]. William Stallings, Cryptography and Network
code. Overcome all the limitations of the qwerty Security, 5th impression, 2008.
cipher.
Cite this Article : V. Vennila, S. Umamaheswari, "An Extension to Traditional Qwerty Cryptographic Method", International
Journal of Scientific Research in Computer Science, Engineering and Information Technology (IJSRCSEIT), ISSN : 2456-3307,
Volume 5 Issue 1, pp. 340-342, January-February 2019. Available at doi : https://doi.org/10.32628/CSEIT195151
Journal URL : http://ijsrcseit.com/CSEIT195151

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


342
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195172

A Review on Hadoop Eco System for Big Data


Anushree Raj1, Rio D’Souza2
1M.Sc. Big Data Analytics Department, St Agnes College (Autonomous) Mangalore, Karnataka, India
2Computer Science and Engineering Department, St Joseph Engineering College Mangalore, Karnataka, India

ABSTRACT

In this era of information age, a huge amount of data generates every moment through various sources. This
enormous data is beyond the processing capability of traditional data management system to manage and
analyse the data in a specified time span. This huge amount of data refers to Big Data. Big Data faces numerous
challenges in various operations on data such as capturing data, data analysis, data searching, data sharing, data
filtering etc. HADOOP has showed a big way of various enterprises for big data management. Big data hadoop
deals with the implementation of various industry use cases. To master the Apache Hadoop, we need to
understand the hadoop eco system and hadoop architecture. In this paper we brief on the Hadoop architecture
and hadoop eco system.
Keywords: Big Data, Hadoop architecture, Hadoop eco system components, HDFS, MapReduce.

I. INTRODUCTION The proposed work the second sections gives you a


detailed review on Big Data, third section speaks of
Hadoop is a open source tool. It provides an Hadoop Ecosystem and explains all of its components
efficient framework for running jobs on multiple briefly and the fourth section concludes the paper by
nodes of clusters. Cluster means a group of systems providing the main advantages of Hadoop.
connected via LAN. Apache Hadoop provides parallel
processing of data as it works on multiple machines II. BIG DATA
simultaneously [7]. Big Data and Hadoop efficiently
processes large volumes of data on a cluster of Bigdata is a term used to describe a collection of data
commodity hardware [8]. Hadoop is for processing that is huge in size and yet growing exponentially
huge volume of data. Apache Hadoop software with time [1]. Big Data generation includes stock
library is a framework that allows for the distributed exchanges, social media sites, jet engines, etc. Big
processing of large data sets across clusters of Data could be Structured, Unstructured or Semi-
computers using simple programming models. It is structured. Volume, Variety, Velocity, and
designed to scale up from single server to thousands Variability are few characteristics of Big data [2]. It
of machines, each offering local computation and plays a major in Improved customer service, better
storage. Rather than rely on hardware to deliver high operational efficiency, Better Decision Making.
availability, the library itself is designed to depict and
handle failures at the application layer, so delivering There were two major challenges with Big Data [9]:
a highly available service on the top of a cluster of
computers, each of which may be prone to failures.

CSEIT195172 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 343-348 ]
343
Anushree Raj, Rio D’Souza Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 343-348

• Big Data Storage: To store Big Data, in a HDFS:


flexible infrastructure that scales up in a cost HDFS is a distributed file system. It supports a single
effective manner, was critical. file system name space, which stores data in a
• Big Data Processing: Even if a part of Big traditional hierarchical format of directories and files.
Data is Stored, processing it would take Across an instance, data is divided into 64MB chunks
years. that are triple-mirrored across the cluster to provide
To solve the storage issue and processing issue, two resiliency. It is optimized for streaming access of
core components were created in Hadoop – large files that are in the 100s of MB upwards on
HDFS and YARN. HDFS solved the storage issue as it HDFS and access them through MapReduce to
stores the data in a distributed fashion and is easily process them in batch mode [4]. HDFS files are write
scalable [8]. YARN solved the processing issue by once files. There is no concept of random writes.
reducing the processing time drastically. HDFS doesn't do random reads.
HBase:
III. HADOOP ECHO SYSTEMS Apache HBase is a database that stores it's data in a
distributed filesystem. The filesystem of choice
Hadoop Eco System is aligned in four different layers typically is HDFS owing to the tight integration
such as Data Storage, Data processing, Data Access between HBase and HDFS. HBase provides low
and Data Management. Each layer comprises of latency access to small amounts of data from within a
different Hadoop eco system components [7]. The large data set. We can access single rows quickly
figure shows the Hadoop eco system with its from a billion row table. It provides flexible data
components. model to work with and data is indexed by the row
key. It can perform fast scans across tables [5]. It
scales in terms of writes as well as total volume of
data.

2. Data Processing Layer in Hadoop


The data processing framework is a tool used to work
with the data itself.
The data processing components of hadoop eco
system are MapReduce and YARN.
Figure Hadoop Eco Systems
MapReduce:
1. Data Storage Layer in Hadoop MapReduce is the tool that actually gets data
Hadoop aims to execute the distributed processing processed. MapReduce runs as a series of jobs, with
with the minimum latency possible. this is achieved each jobs essentially a separate java application that
by executing Map processing on the node that stores gives out into the date and starts pulling out
the data, a concept known as data locality. information as needed [6]. Hadoop is not really a
The data storage components of hadoop eco system database. It stores data and we can pull data out of it,
are the HDFS and HBASE. but there are no queries involved. Hadoop is more a
data warehousing system so it needs a system like
MapReduce to actually process the data.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


344
Anushree Raj, Rio D’Souza Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 343-348

YARN: Apache Pig is a platform for managing large sets of


YARN stands for “Yet Another Resource Negotiator”. data which consists of high-level programming to
Yarn runs a non-MapReduce jobs within the Hadoop analyze the data. Pig also consists of the
framework. YARN is a key element of the Hadoop infrastructure to evaluate the programs [7].
data processing architecture that provides different The advantages of Pig programming is that it can
data handling mechanisms, including interactive SQL easily handle parallel processes for managing very
and batch processing [3]. It improves the large amounts of data. The programming on this
performance of data processing in Hadoop by platform is basically done using the textual language
separating the resource management and scheduling Pig Latin.
capabilities of MapReduce from its data processing
component. YARN allows different data processing Mahout:
methods like graph processing, interactive processing, Apache Mahout is an open source library of scalable
stream processing as well as batch processing to run machine learning algorithms that focuses on
and process data stored in HDFS. Therefore YARN clustering, classification, and recommendations [3].
opens up Hadoop to other types of distributed The primitive features of Apache Mahout are: The
applications beyond MapReduce [5]. YARN algorithms of Mahout are written on top of Hadoop,
architecture consists of Resource management, Node so it works well in distributed environment. Mahout
management, Application master and container. uses the Apache Hadoop library to scale effectively
in the cloud. Mahout offers the coder a ready-to-use
3. Data Access Layer in Hadoop framework for doing data mining tasks on large
Hadoop can usually hold terabytes or petabytes of volumes of data. Mahout lets applications to analyze
data to process, hence Data Access is an extremely large sets of data effectively and in quick time. It
important aspect in any project or product, especially Includes several MapReduce enabled clustering
with hadoop. implementations such as k-means, fuzzy k-means,
The Data access components of hadoop eco system Canopy, Dirichlet, and Mean-Shift, Supports
are: Distributed Naive Bayes and Complementary Naive
Hive: Bayes classification implementations, Comes with
Apache Hive is a component of Hortonworks Data distributed fitness function capabilities for
Platform (HDP). Hive provides a SQL-like interface evolutionary programming, Includes matrix and
to data stored in HDP [10]. Hive gives a SQL- vector libraries.
like interface to query data stored in various
databases and file systems that integrate with Avro:
Hadoop. Hive provides the necessary SQL abstraction Avro is a remote procedure call and
to integrate SQL-like queries (HiveQL) into the data serialization framework developed within
underlying Java without the need to implement Apache's Hadoop project [10]. It is a schema-based
queries in the low-level Java API. Since most data serialization utility, which accepts schemas as input.
warehousing applications work with SQL-based In spite of various schemas being available, Avro
querying languages, Hive aids portability of SQL- follows its own standards of defining schemas like
based applications to Hadoop type of file, location of record, name of the record,
fields in the record with their corresponding data
Pig: types.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


345
Anushree Raj, Rio D’Souza Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 343-348

Sqoop: Features of Oozie are: It execute and monitor


Apache Sqoop is a tool that is extensively used to workflows in Hadoop, provides periodic scheduling
transfer large amounts of data from Hadoop to the of workflows, Trigger execution of data availability
relational database servers and vice-versa [7]. Sqoop and uses HTTP and command line interface and web
can be used to import the various types of data from console.
Oracle, MySQL and such other databases.
Sqoop needs a connector to connect the different Ambari:
relational databases. Almost all Database vendors Apache Ambari is an open source administration tool
make a JDBC connector available specific to that deployed on top of Hadoop cluster and responsible
Database, Sqoop needs a JDBC driver of the database for keeping track of running applications and their
for interaction. status. Apache Ambari can be referred to as an open
source web-based management tool that manages,
Drill: monitors and provisions the health of Hadoop
Apache Drill is a low latency distributed query clusters [3]. It provides a highly interactive dashboard
engine for large-scale datasets, including structured which allows the administrators to visualize the
and semi-structured/nested data [3]. Drill is capable progress and status of every application running over
of querying nested data in formats like JSON and the Hadoop cluster.
Parquet and performing dynamic schema discovery. Its flexible and scalable user-interface allows a range
Drill does not require a centralized metadata of tools such as Pig, MapReduce, Hive, etc. [11], to be
repository installed on the cluster and administers their
performances in a user-friendly fashion.
Thrift:
Thrift is a driver level interface that provides API for Flumes:
client implementation. It combines a software stack Apache Flume is a distributed, reliable, and available
with a code generation engine to build cross- system for efficiently collecting, aggregating and
platform services which can connect applications moving large amounts of log data from many
written in a variety of languages and frameworks [11]. different sources to a centralized data store. Suppose
there are 100 servers with unstructured log files [11].
4. Data Management Layer in Hadoop We have to load data in HDFS and analyze the same.
Hadoop data management software(entity) is for We can use flume for this purpose. It has three tiers
managing enormous amounts of scattered data. It is Agents, Collector ad storage. Flume can capture real
part of the ecosystem of modern data management time data from the server. It can sink data from logs
[11]. into HDFS. Flume is used to move the log data
The Data Management components of hadoop eco generated by application servers into HDFS at a
system are: higher speed.
Oozie:
Oozie is a workflow scheduler system to manage Zookeeper:
Apache Hadoop jobs. Oozie is integrated with the Apache Zookeeper is a coordination service for
rest of the Hadoop stack supporting several types of distributed application that enables synchronization
Hadoop jobs such as Java MapReduce, Streaming across a cluster [3]. Zookeeper in Hadoop can be
MapReduce, Pig, Hive and Sqoop [10]. Oozie is a viewed as centralized repository where distributed
scalable, reliable and extensible system. The main applications can put data and get data out of it. It is

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


346
Anushree Raj, Rio D’Souza Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 343-348

used to keep the distributed system functioning HCatalog works like a key component of Hive and it
together as a single unit, using its synchronization, enables the users to store their data in any format
serialization and coordination goals. For and any structure.
simplicity's sake Zookeeper can be thought of as a file HCatalog is a table storage management tool for
system where we have znodes that store data instead Hadoop that exposes the tabular data of Hive
of files or directories storing data. Zookeeper is metastore to other Hadoop applications. It enables
a Hadoop Admin tool used for managing the jobs in users with different data processing tools (Pig,
the cluster. MapReduce) to easily write data onto a grid [7].
HCatalog ensures that users don’t have to worry
ZooKeeper is a centralized service for maintaining about where or in what format their data is stored.
configuration information, naming, providing This is a small tutorial that explains just the basics of
distributed synchronization, and providing group HCatalog and how to use it.
services. ZooKeeper is simple, distributed, reliable,
and fast [11]. IV. CONCLUSION
• maintaining configuration information: It
maintains cluster configuration info which is Apache Hadoop is the most popular and powerful big
shared across all the nodes in the cluster. data tool, Hadoop provides world’s most reliable
• naming: Zookeeper can be used as naming storage layer – HDFS, a batch Processing engine –
service, so that one node in the cluster can find MapReduce and a Resource Management Layer –
another node in the large cluster ex: 1000 node YARN [11]. HADOOP is a highly scalable storage
cluster platform, because it can store and distribute very
• providing distributed synchronization: We can large data sets across hundreds of inexpensive servers
also use zookeeper for solving distributed that operate in parallel. Unlike traditional relational
synchronization problems in the cluster by using database systems (RDBMS) that can't scale to process
locks, Queues etc. large amounts of data, Hadoop enables businesses to
• providing group services: Zookeeper also helps in run applications on thousands of nodes involving
group service with the selection of a master in thousands of terabytes of data. HADOOP also offers a
the cluster (Leader election process). cost-effective storage solution for businesses'
exploding data sets. Hadoop enables businesses to
ZooKeeper can work in replicated mode as well as easily access new data sources and tap into different
standalone mode types of data (both structured and unstructured) to
generate value from that data [13]. This means
HCatalog: businesses can use Hadoop to derive valuable
HCatalog is a table storage management tool for business insights from data sources. A key advantage
Hadoop. It exposes the tabular data of Hive of using Hadoop is its fault tolerance. When data is
metastore to other Hadoop applications. It enables sent to an individual node, that data is also replicated
users with different data processing tools (Pig, to other nodes in the cluster, which means that in the
MapReduce) to easily write data onto a grid. It event of failure, there is another copy available for
ensures that users don’t have to worry about where use.
or in what format their data is stored [3].

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


347
Anushree Raj, Rio D’Souza Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 343-348

V. REFERENCES [9] PekkaPääkkönen, DanielPakkala1, “Reference


Architecture and Classification of Technologies,
[1] Natalia Miloslavskaya ,Alexander Tolstoy, “Big Products and Services for Big Data Systems,”
Data, Fast Data and Data Lake Concepts” 7th [10] https://intellipaat.com/tutorial/hadooptutorial/introductio n- hadoop/
Annual International Conference on [11] Apache Hadoop. http://hadoop.apache.org/
Biologically Inspired Cognitive Architectures, [12] Kala Karun. A , Chitharanjan. K,” A Review on
Volume 88, 2016, Pages 300–305 Hadoop – HDFS Infrastructure Extensions”,
[2] B. Saraladevi, N. Pazhaniraja, P. Victer Paul, Conference on Information and
M.S. Saleem Basha, P. Dhavachelvan, “Big Data Communication Technologies,2013,IEEE
and Hadoop-A Study in Security Perspective,” [13] Naveen Garg , Dr. Sanjay Singla, Dr. Surender
2nd International Symposium on Big Data and Jangra, “Challenges and Techniques for Testing
Cloud Computing (ISBCC’15), Procedia of Big Data,” Procedia Computer Science 85 (
Computer Science 50 ( 2015 ) 596 – 601 2016 ) 940 – 948
[3] DT Editorial Services, “Big Data(covers
Hadoop2, Map Reduce, Hive, Yarn, Pig, R and Cite this article as :
Data Visualization)” by Dreamtech Press
[4] “Hadoop, MapReduce and HDFS:A developer Anushree Raj, Rio D'Souza, "A Review on Hadoop
perspective,”(Procedia Computer Science, Eco System for Big Data", International Journal of
Volume 48, 2015,Pages 45-50) Scientific Research in Computer Science, Engineering
[5] A Novel and efficient de-duplication system for and Information Technology (IJSRCSEIT), ISSN :
HDFS (Procedia Computer Science,Volume 92, 2456-3307, Volume 5 Issue 1, pp. 343-348, January-
2016, Pages (498-505) February 2019. Available at doi :
[6] Tharso Ferreira, Antonio Espinosa, Juan Carlos https://doi.org/10.32628/CSEIT195172
Moure, Porfidio Hern´andez, “An Journal URL : http://ijsrcseit.com/CSEIT195172
Optimization for MapReduce Frameworks in
Multi-core,” International Conference on
Computational Science, ICCS 2013, Procedia
Computer Science 18 ( 2013 ) 2587 – 2590
[7] Can Uzunkaya, Tolga Ensari, Yusuf Kavurucu,
“Hadoop Ecosystem and Its Analysis on
Tweets,” World Conference on Technology,
Innovation and Entrepreneurship, Procedia -
Social and Behavioral Sciences 195 (2015 ) 1890
– 1897
[8] Sachin Bende, Rajashree Shedge, “Dealing with
Small Files Problem in Hadoop Distributed File
System,” 7th International Conference on
Communication, Computing and Virtualization
2016, Procedia Computer Science 79 ( 2016 )
1001 – 1012

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


348
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195186

A Novel Approach on Prevention of a Mobile Exploration in


Machine Using AI
1J.Antony Daniel Rex, 2S.Dinakaran & 3S.Muthukumaran
1,3 Assistant Professor, PG and Reseach Department of Computer Science, St.Joseph College of Arts & Science
(Autonomous) Cuddalore, Tamilnadu, India
2Ph. D. Scholar, Anna University, Salem, Tamilnadu, India

ABSTRACT

This paper deals with a low cost solution to problem avoidance for a mobile machine using just a single Artifical
Intelligennce. It allows the machine to navigate smoothly in an unknown environment, avoiding collisions,
without having to stop in front of problems. The problem avoidance process is made up of three distinct stages -
the mapping algorithm, the core problem avoidance algorithm, and the steering algorithm. The mapping
algorithm takes the raw Artifical Intelligennce readings and processes them to create higher resolution maps
from the wide-angle Artifical Intelligennce. The problem avoidance algorithm is based on the potential field
theory which considers the machine to be a test charge that is repelled by all the problems around it, and
which moves in the direction of the resultant of the forces acting on it. An algorithm which steers a mobile
machine based on the differential drive system is also discussed.
Keywords : Mobile Machine, Problem Avoidance, Autonomous Navigation, Ultrasonic Range Sensor.

I. INTRODUCTION consist of a combination of an infrared light emitting


device and an infrared light sensor. These sensors are
Problem avoidance is a primary requirement for any merely proximity detectors, they cannot determine
autonomous mobile machine. The machine acquires the range of the problem in front. The range of these
information about its surroundings through sensors sensors is also limited to a maximum of 80cm.
mounted on the machine. Various types of sensors
can be used for problem avoidance, as detailed below. Ultrasonic range sensors, which determine the range
of the object in front of it. They work by sending a
A. Types of Sensors: short burst of ultrasonic waves, and measuring the
Bump sensors, which are micro-switches activated time taken for the echo to be received. They have a
when the machine touches an problem. This is a wide beam angle, typically 30. These sensors have
simple, inexpensive method of problem detection, ranges of up to 6m.
but operates only on contact, which makes it useful
only for slow moving machines. Laser range finders, which work on the same
principle as ultrasonic range sensors, except that they
Infrared proximity detectors, which detect the use LASER instead of ultrasound. Laser range finders
presence of an object in front of the sensor. They

CSEIT195186 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 349-356 ]
349
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

have a range of upto 30m, and are very accurate, Beyond the range value, there may be problems
having an angular resolution of upto 0.25. somewhere along the arc, but it is not certain at
which positions the problems are present.
However, they are very expensive compared to other Figure 1 shows the next sensor reading, which is
sensors of these, the ultrasonic range sensor was obtained after rotating the sensor through a certain
found most suitable for our requirement because of angle, say 5. This reading, with a range value of 1.2m
its low cost and ranging capability. covers angles -10 to +20. After obtaining this reading,
we are certain that there is no problem between
II. MAPPING these angles for a range of 1.2m. Thus, in the
overlapping region, the larger distance value is
An ultrasonic range sensor detects the range of an chosen over the smaller one.
object in front of it by sending a short burst of
ultrasonic waves, and measuring the time taken for If a subsequent overlapped reading is smaller than the
the echo to be received. These sensors typically have existing value, then we have to consider the ‘age’ of
a wide beam angle of 15 to 30 As a result, the cone the existing information. If the existing large range
gets wider with increasing distance from the value is very recent, then the overlapped area is taken
transmitter. This increases the uncertainty in the to have the larger range, because it can be assumed
spatial dimensions of the problem. Most mobile that no object moved into the intersecting area
machines which use ultrasonic range finders for between the two scans. If the existing large range
problem avoidance have 12 or 24 sensors in a ring value is old, then the smaller value is taken, because
around the machine, to cover all directions. This is the machine may have moved within that time, or a
not very cost-effective, and moreover, all of them moving problem may have appeared.
cannot be triggered at once due to crosstalk and
interference problems. A maximum of 4 sensors Using this technique, the accuracy of the ultrasonic
which are at 90 to each other can be triggered at a range information can be greatly improved upto the
time. Our solution was to use a single sensor mounted step size of the stepper motor. However, there are
on a stepper motor to pan the sensor across 180. This some cases where the information is still not accurate.
can be expanded to 4 sensors mounted at 90 to each Consider the case gap between problems A and B can
other on the stepper motor for maximum scanning never be detected by the sensor at that distance.
speed and full coverage. Such a setup also provides Once the machine moves closer to these problems,
the flexibility to perform fine scanning of an area for the gap can be detected.
a more accurate map of the environment, with an
angular resolution of the minimum step size of the IV. PROBLEM AVOIDANCE
stepper motor, as opposed to the 30 resolution of the
sensor. All mobile machines feature some kind of problem
avoidance. The algorithm could be as simple as
III. MACHINE SCANNING ALGORITHM stopping if an problem is detected in front of it. This
is common in track-following machines, where the
A single range reading from the sensor implies that machine follows a pre-determined track. Another
the area in a 30 sector in front of it with a radius of solution is wall-following, in which the machine
the range value contains no problems. A reading of a moves alongside a wall at a fixed distance, following
range of 1m. This reading covers angles -15 to +15. the contour of the wall till it reaches its destination.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


350
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

In the case of an exploration machine, the


environment is unknown and problem avoidance h as
to be dynamic in nature. Our approach is an
adaptation of the potential field method or VFF
method in which the machine is considered as a test
charge which is repelled by problems which are
considered as like charges. Section II explains our
mapping methods and an algorithm to obtain
accurate data from the wide angle ultrasonic range
sensor. Our potential field method adaptation is
discussed steering control for the differential drive
Figure 1 : The second reading from the ultrasonic
machine based on the output of the potential field
range finder overlapping the first
algorithm is described.
The machine is then steered in the direction of the
resultant of all the forces added up. This results in
Information about the ranges of problems around the
motion of the machine similar to a charge moving in
machine obtained from the mapping technique is
a potential field - avoiding like charges and moving
stored in a table which is continuously updated. This
towards the unlike charge. For the case of an
range information is then used to calculate a
exploration machine without a target, the algorithm
destination direction which the machine must move
has to be slightly modified. A threshold limit is set
towards. This calculation is performed using our
beyond which the charge assigned to the range is
problem avoidance algorithm based on the potential
positive and attracts the machine. Thus, problems
field method.In the potential field method, the
which are close by to the machine will repel it, and
machine is considered as a test negative charge. All
the free spaces attract the machine.
the problems are considered to be negative charges
which repel the machine, and the target is assigned a
positive charge which attracts the machine. The
forces of all these charges on the machine are This is the basic equation representing the force
calculated in accordance with Coulomb’s law, i.e. the exerted on the machine by a single problem at a
force is proportional to the charge and inversely distance r and an angle from the machine. k is a
proportional to the square of the distance. numerical constant selected to bring the force into
the required range. The charge q is not present in this
equation because different angles do not have
different charge values. This equation is modified to
make near ranges repel the machine and far ranges
attract, as shown in the following equation.

Here, t is a threshold distance at which the problem


neither attracts nor repels the machine, and ranges
beyond the threshold distance are attractive. The

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


351
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

total force on the machine is obtained by vectorial


addition of all the individual forces.

This is the expression for the resultant of all the


Figure 2.1 : Unit circle in which all possible
forces acting on the machine. The machine is steered
force vectors lie
in this direction in order to avoid problems. The
magnitude of the vector can be used to control the
On this circle, a point at 90 indicates that the
speed of the machine. This results in slow movement
machine has to move straight, a point at 0 indicates
in a cluttered environment, and faster motion when
that the machine has to turn right in-place, and at all
there are fewer problems surrounding the machine.
points in between, it has to exhibit varying degrees of
turning.
V. STEERING CONTROL FOR A DIFFERENTIAL
DRIVE SYSTEM

A mobile machine which uses differential drive has


separate speed and direction control for the left and
right sets of wheels. Such a system is very flexible for
a mobile machine because it does not have a
minimum turning radius, it can execute a turn
around its own center. Arc turning is achieved by
driving the left and right wheels at different speeds.
Driving one set of wheels forward and the other set
backward results in in-place turning.
Figure 2.2 : The unit circle shrunk into an ellipse

An algorithm was designed to convert the force


This unit circle is shrunk into an ellipse as shown in
vector obtained from the problem avoidance
Figure 2.2, by dividing the x coordinate by a constant.
algorithm into the left and right wheel speeds
This ensures that the wheel speeds of the machine
required for the differential drive system. The
are lower while turning and faster when moving
magnitude and direction of the force vector is plotted
straight. Turning at high speeds can result in slipping
as a point on a graph, after being normalized so that
and inaccurate turns, this prevents it from happening.
the maximum possible magnitude is 1. Thus, all
possible force vectors will lie in a unit circle around
the origin, as shown in Figure 2.1.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


352
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

The mobile machine we designed and constructed,


and implemented the problem avoidance algorithms
on. It uses a differential drive system for mobility, but
uses tracks instead of wheels, like a tank. The drive
motors used are stepper motors. It has a single
ultrasonic range sensor mounted on a stepper motor
in front which can pan across an angle of 180 in steps
as small as 0.45. The machine is controlled by a
computer through the parallel port connected to the
required driver circuits on the machine itself.

RTLinux was chosen as the software platform because


Figure 2.3 : Rotation of the force vector
real time functionality was required, especially to
measure the sonar range timing, and to trigger the
Now the existing force vector is rotated clockwise by
stepper motors. The default interval at which
45, as indicated in Figure 2.3. The new coordinates of
ultrasonic range readings are taken is 15, which
the vector are then taken as the left and right wheel
provides 12 readings over 180. Each reading overlaps
speeds respectively (after normalization). This result
the previous by 15. This provides for fast scanning of
in the machine turning in the required direction
the environment -- an average of 500ms to scan 180,
based on the direction of the original force vector.
and thus the machine can move at high speeds
This dynamic steering algorithm ensures that the
without danger of collision. If an area needs to be
machine doesn't have to stop in front of an problem
mapped more accurately, or in case of a highly
to go around it, instead it navigates smoothly to the
cluttered environment, the step size is reduced and
required direction.
the maps obtained are of higher resolution, which
can help in navigating precisely through smaller
The force vector is updated every time new range
pathways.
information is obtained,which is approximately every
20 or 30 milliseconds, and the wheel speeds are also
VII. CONCLUSION
correspondingly modified every time. Modifications
in the wheel speeds will not be abrupt, because every
This traveler machine can explore in any unknown
reading changes the range associated with only a
environment smoothly avoiding problems. Thus it
narrow angle, resulting in a very small change in the
can be used even in exploring other planets using
resultant force vector. Since the wheel speeds vary
very precise sensors. This method of avoiding
smoothly according to the sensor information, and
problems is very useful in army, military applications.
the speeds are updated almost in real time, the
These kinds of machines hold the technology for the
machine follows a smooth course around the
coming future.
problems present in the environment, while still
being able to respond immediately to changes in the
VIII. REFERENCES
environment like dynamic problems.

[1] “Sensors Handbook”-Solaman, Sabrie.


VI. EXPERIMENTAL RESULTS

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


353
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

[2] “Sensory based motion planning”-IEEE Computer Networking, Wireless and Mobile
transactions on Machineics & Automation- Communications. Vol No – 3, Issue No - 1, pp.
Kamon Ishay. 245-256.
[3] A.S.Syed Navaz, C.Prabhadevi & [11] A.S.Syed Navaz, H.Iyyappa Narayanan &
V.Sangeetha”Data Grid Concepts for Data R.Vinoth.” Security Protocol Review Method
Security in Distributed Computing” January Analyzer (SPRMAN)”, August – 2013,
2013, International Journal of Computer International Journal of Advanced Studies in
Applications, Vol 61 – No 13, pp 6-11. Computers, Science and Engineering, Vol No –
[4] A.S.Syed Navaz, V.Sangeetha & C.Prabhadevi, 2, Issue No – 4, pp. 53-58.
“Entropy Based Anomaly Detection System to [12] A.S.Syed Navaz, G.M. Kadhar Nawaz &
Prevent DDoS Attacks in Cloud” January 2013, B.Karthick.” Probabilistic Approach to Locate a
International Journal of Computer Mobile Node in Wireless Domain”, April –
Applications, Vol 62 – No 15, pp 42-47. 2014, International Journal of Computer
[5] Y. Kim, H. Shin, and H. Cha, “Y-mac: An Engineering and Applications, Vol No – 6, Issue
energy-efficient multichannel mac protocol for No – 1, pp.41-49.
dense wireless sensor networks,” in IPSN ’08, [13] S. Gandham, Y. Zhang, and Q. Huang,
April 2008, pp. 53–63. Distributed time-optimal scheduling
[6] A.S.Syed Navaz, A.S.Syed Fiaz, C.Prabhadevi, forconvergecast in wireless sensor networks,”
V.Sangeetha & S.Gopalakrishnan “Human Computer Networks, vol. 52, no. 3, pp. 610–
Resource Management System” Jan – Feb 2013, 629, 2008.
International Organization of Scientific [14] A.S.Syed Navaz & G.M. Kadhar Nawaz, “Ultra
Research Journal of Computer Engineering, Vol Wideband on High Speed Wireless Personal
8, Issue 4, pp. 62-71. Area Networks” August – 2014, International
[7] A.S.Syed Navaz, S.Gopalakrishnan & R.Meena Journal of Science and Research, Vol No – 3,
“Anomaly Detections in Internet Using Issue No – 8, pp.1952-1955.
Empirical Measures” February 2013, [15] A.S.Syed Navaz & Dr.G.M. Kadhar Nawaz &
International Journal of Innovative Technology A.S.Syed Fiaz “Slot Assignment Using FSA and
and Exploring Engineering, Vol 2 – Issue 3. pp. DSA Algorithm in Wireless Sensor Network”
58-61. October – 2014, Australian Journal of Basic and
[8] “Problem detection based on qualitative Applied Sciences, Vol No –8, Issue No –16,
&quantitative 3d reconstruction-IEEE pp.11-17.
transaction on pattern analysis and [16] A.S.Syed Navaz & A.S.Syed Fiaz, “Load
intelligence”-Zhang-Zhonggei. Balancing in P2P Networks using Random
[9] A.S.Syed Navaz, M.Ravi & T.Prabhu, Walk Algorithm” March – 2015, International
“Preventing Disclosure of Sensitive Knowledge Journal of Science and Research, Vol No – 4,
by Hiding Inference” February 2013, Issue No – 3, pp.2062-2066.
International Journal of Computer [17] Qiao Li, Baihai Zhang, Lingguo Cui, Zhun Fan,
Applications, Vol 63 – No 1. pp. 32-38. V. Vasilakos Athanasios, Epidemics on small
[10] A.S.Syed Navaz , T.Dhevisri & Pratap worlds of tree-based wireless sensor networks,
Mazumder “Face Recognition Using Principal Journal of Systems Science and Complexity
Component Analysis and Neural Networks“ December 2014, Volume 27, Issue 6, pp 1095-
March -2013, International Journal of 1120.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


354
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

[18] A.S.Syed Navaz, J.Antony Daniel Rex, S.Jensy and Research, Vol No - 5, Issue No - 1, pp.1581-
Mary. “Cluster Based Secure Data Transmission 1584.
in WSN” July – 2015, International Journal of [26] A.S.Syed Fiaz, N.Asha, D.Sumathi & A.S.Syed
Scientific & Engineering Research, Vol No - 6, Navaz “Data Visualization: Enhancing Big Data
Issue No - 7, pp. 1776 – 1781. More Adaptable and Valuable” February –
[19] Ricardo Severino, Nuno Pereira, Eduardo 2016, International Journal of Applied
Tovar, Dynamic cluster scheduling for cluster- Engineering Research, Vol No - 11, Issue No -
tree WSNs, Springer, Vol.3, No.493, 2014. 4, pp.–2801-2804.
[20] A.S.Syed Navaz, J.Antony Daniel Rex, P.Anjala [27] W. Song, F. Yuan, and R. LaHusen, “Time-
Mary. “An Efficient Intrusion Detection optimum packet scheduling for many-to-one
Scheme for Mitigating Nodes Using Data routing in wireless sensor networks,” in MASS
Aggregation in Delay Tolerant Network” ’06, pp. 81–90.
September – 2015, International Journal of [28] A.S.Syed Navaz & Dr.G.M. Kadhar Nawaz
Scientific & Engineering Research, Vol No - 6, “Flow Based Layer Selection Algorithm for
Issue No - 9, pp. 421 – 428. Data Collection in Tree Structure Wireless
[21] A.S.Syed Navaz, P.Jayalakshmi, N.Asha. Sensor Networks” March – 2016, International
“Optimization of Real-Time Video Over 3G Journal of Applied Engineering Research, Vol
Wireless Networks” September – 2015, No - 11, Issue No - 5, pp.–3359-3363.
International Journal of Applied Engineering [29] A.S.Syed Navaz & Dr.G.M. Kadhar Nawaz
Research, Vol No - 10, Issue No - 18, pp. 39724 “Layer Orient Time Domain Density Estimation
– 39730. Technique Based Channel Assignment in Tree
[22] M.Ravi & A.S.Syed Navaz. “Development of Structure Wireless Sensor Networks for Fast
Mail Tracking System” October – 2015, Data Collection” June - 2016, International
International Journal of Innovative Research in Journal of Engineering and Technology, Vol No
Computer and Communication Engineering, - 8, Issue No - 3, pp.–1506-1512.
Vol No - 3, Issue No - 10, pp.9757-9762. [30] M.Ravi & A.S.Syed Navaz "Rough Set Based
[23] S.Jensy Mary, A.S Syed Navaz & J.Antony Grid Computing Service in Wireless Network"
Daniel Rex, “QA Generation Using Multimedia November - 2016, International Research
Based Harvesting Web Information” November Journal of Engineering and Technology, Vol No
– 2015, International Journal of Innovative - 3, Issue No - 11, pp.1122– 1126.
Research in Computer and Communication [31] A.S.Syed Navaz, N.Asha & D.Sumathi “Energy
Engineering, Vol No - 3, Issue No - 11, Efficient Consumption for Quality Based Sleep
pp.10381-10386. Scheduling in Wireless Sensor Networks”
[24] A.S Syed Navaz & A.S.Syed Fiaz “Network March - 2017, ARPN Journal of Engineering
Intelligent Agent for Collision Detection with and Applied Sciences, Vol No - 12, Issue No - 5,
Bandwidth Calculation” December – 2015, pp.–1494-1498.
MCAS Journal of Research, Vol No – 2, pp.88- [32] N. Lai, C.King, and C. Lin, “On Maximizing the
95, ISSN: 2454-115X. throughput of convergecast in wireless sensor
[25] A.S Syed Navaz & K.Durairaj “Signature networks,” in GPC ’08, pp. 396– 408.
Authentication Using Biometric Methods” [33] A.S.Syed Fiaz, K.S.Guruprakash & A.S.Syed
January – 2016, International Journal of Science Navaz “Prediction of Best Cloud Service
Provider using the QoS Ranking Framework”

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


355
J.Antony Daniel Rex et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 349-356

January – 2018, International Journal of InternationalJournal of Computer Science


Engineering & Technology, Vol No - 7, Issue Engineering and Information Technology
No -1.1, pp.– 486-488. Research, Vol No: 3; Issue No: 1, pp. 257-266.
[34] A.S.Syed Navaz, Asha.N, Vanmathi [41] A.S.Syed Fiaz, N.Devi, S.Aarthi "Bug Tracking
Chandrasekaran & J.Jayashree “Resourceful and Reporting System", March 2013,
Investigate Encryption Method Using Data International Journal of Soft Computing and
Hunt in Mobile Cloud Service”, August – 2018, Engineering, Vol No: 3; Issue No: 1, pp. 257-
ARPN Journal of Engineering and Applied 266.
Sciences, Vol No - 13, Issue No - 15, pp.–4543- [42] M. Usha, J. Akilandeswari and A.S.Syed Fiaz
4549. “An efficient QoS framework for Cloud
[35] Asha.N, A.S.Syed Navaz, J.Jayashree & Brokerage Services”, Dec. 2012, International
J.Vijayashree “RFID Based Automated Gate Symposium on Cloud and Service Computing,
Security System”, November – 2018, ARPN pp: 76-79, 17-18, IEEE Xplore.
Journal of Engineering and Applied Sciences,
Vol No - 13 , Issue No - 22, pp.– 8901 – 8906.
[36] A Xiaojun Zhu, Xiaobing Wu, Guihai Chen, An
exact algorithm for maximum lifetime data
gathering tree without aggregation in wireless
sensor networks, Springer, Wireless Networks
2014.
[37] .S. Syed Navaz, N. Asha & Kuna
Venkateswararao “Face Recognition and
Detection with New Image Enhancement
Techniques” December – 2018, Journal of Cite this article as :
Advanced Research in Dynamical and Control
Systems, Vol No -10 , Issue No – 14 (Special J.Antony Daniel Rex, S.Dinakaran, S. Muthukumaran,
Issue), pp.– 760-764. "A Novel Approach on Prevention of a Mobile
[38] A.S.Syed Fiaz, I.Alsheba & R.Meena “Using Exploration in Machine Using AI ", International
Neural Networks to Create an Adaptive Journal of Scientific Research in Computer Science,
Character Recognition System”, Sep 2015, Engineering and Information Technology
Discovery - The International Daily journal, (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
Vol.37 (168), pp.53-58. 349-356, January-February 2019. Available at doi :
[39] A.S.Syed Fiaz, M. Usha and J. Akilandeswari “A https://doi.org/10.32628/CSEIT195186
Brokerage Service Model for QoS support in Journal URL : http://ijsrcseit.com/CSEIT195186
Inter-Cloud Environment“,March 2013,
International Journal of Information and
Computation Technology, Vol.3, No.3, pp 257-
260,
[40] A.S.Syed Fiaz, R.Pushpapriya, S.Kirubashini &
M.Sathya “Generation and allocation of
subscriber numbers for
telecommunication“,March2013,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


356
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT183857

Tuition Plus+ For Managing Tuition Class Data


Bharat Dhak*1, Prachi P. Paunikar2, Simran S. Khan3, Snehal S. Ramteke4, Yashkumar R. Jais5, Priya R. Gupta6
*1Lecturer, Computer Science and Engineering Department, Priyadarshini J. L. College of Engineering, Nagpur,
Maharashtra, India
2-6 Computer Science and Engineering Department, Priyadarshini J. L. College of Engineering, Nagpur,
Maharashtra, India

ABSTRACT

Tuition Plus+ is a website and a web-based application provides a simple interface for maintenance of a student
records and information. It is used by educational institution like tuitions or colleges which can manage
student’s records easily. Tuition Plus website deals with different modules like registration, organization,
Subjects, Course, Batches, Student, Account, Enquiry and Report Module. It tracks all the details of a student
from a day to day course. It will also track all the students fees structure like installments. These will be
available through a secure, online interface embedded in the Tuition Plus website. It will also have registration
details, batch execution details, students’ details in all aspects, the various academic notifications to the staff and
students updated by the tuition organisation
Keywords : Tuition Plus+, DFD, OTP

I. INTRODUCTION the Tuition administrator and ensures highest


possible level of security. The system features a
[1]The design and implementation of a complex logging system to track all users access and
comprehensive student information system and user ensure conformity to data access guidelines and is
interface is to replace the current paper records. expected to increase the efficiency of the Tuition
Tuition Staff are able to directly access all aspects of a record management thereby decreasing the work
student’s academic progress through a secure, online hours needed to access and deliver student records.
interface embedded in the Tuition Plus+ website. The
system utilizes user authentication, displaying only One of the major reasons of developing Tuition Plus+
information necessary for an individual’s duties. is to replace the existing system with computerized
Additionally, each sub-system has authentication system. The staff of the organisation can directly
allowing authorized users to create or update access the details of the student, regarding their fees
information in that sub-system. All data is structure and payment using online interface. The
thoroughly reviewed and validated on the server website consists of authentication system such as
before actual record alteration occurs. In addition to a student login and merchant login. Using merchant
staff user interface, the system plans for student user login, tuition administrator can access the
interface, allowing users to access information and information of student. The data will be securely
submit requests online thus reducing processing time. saved in the SQL server which is managed by the
All data is stored securely on SQL servers managed by tuition administrator. Only the head administrator

CSEIT195197 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 357-362 ]
357
Bharat Dhak et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 357-362

will have the authority to update or edit the easy and interactive manner which provides facilities
information of student as well as the information of like online registration and profile creation of
their organisation. Student login will help the student’s thus reducing paper work and automating
students to see their own data, fees structure, etc. the record generation process in an educational
organization, which will also save the time of a
[1]Previously, the Tuition relied heavily on paper student. This website also provides a common
records for this initiative. While paper records are a platform for registering their organization online that
traditional way of managing student data there are is why the need of making separate website by
several drawbacks to this method. Like when a investing money will reduce.
student's wants to register for a particular course then
it must have to visit the particular tuition class and III. PROPOSED SYSTEM
register themselves. It takes a very long time to
convey the information to the student. Paper records The purpose is to design a Tuition Plus+ website,
are difficult to manage and track. The physical which contains up to date information of an
exertion required to retrieve, alter, and re-file the educational organization that should improve
paper records are all non-value added activities. This efficiency of a tuition record management and about
system provides a simple interface for the a different courses that they will provide to the
maintenance of student information. It can be used student.
by educational institutes or colleges to maintain the
records of students easily. Achieving this objective is OBJECTIVES
difficult using a manual system as the information is
scattered, can be redundant and collecting relevant • Provide the online interface for a student’s
information may be very time consuming. All these information, courses, fees structure, etc.
problems are solved using Tuition Plus+. The paper • Increase the efficiency of tuition record
focuses on presenting information in an easy and management.
interact manner which provides facilities like online • Decrease time required to access and
registration and profile creation of student’s thus delivered records.
reducing paper work and automating the record • To make a system more secure.
generation process in an educational institution. • Decrease time spent on non-value added
tasks.
II. RELATED WORK

SYSTEM DESIGN
The existing system is web based student information
management is also good but that have certain
A. DATA FLOW DIAGRAM
drawback like these existing website is only for a
Data Flow Diagram shows the graphical
college but our system is for all educational
representation of the flow of Tuition Plus+. It shows
institution .Also the existing system does not provide
the visualization of processing of data. It shows how
any information about a fees structure but that is
users of the Tuition Plus+ will interact with the
included in our system .
system. DFD contains the different modules that can
be viewed as a function that is used to transform the
The purpose is to design a Tuition Plus+ website that,
these website focuses on presenting information in an

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


358
Bharat Dhak et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 357-362

given input into the desired output. Different


modules are as follows:

a. Registration Module
b. Organization Setting Module
c. Subject Module
d. Course Module
e. Batch Module
f. Student Module
g. Account Module
h. Enquiry Module
i. Report Module
There are two different login because both the users
have different level of authorization to the system.
Merchants have the authority to access the data of
the students in their organization whereas users have
the authority to access their own data.

A. Registration Module

Merchant User will fill this registration form on


submit it will send OTP to user mobile. Users have to
fill the registration form the email and password
entered in this form will be used further for login in
B. Detailed Structure
Tuition Plus+ software. Users have to enter OTP
which sent by Tuition Plus+ on entered mobile
The system has two logins i.e. User Login and
number in input given in above screen. Users can
Merchant Login. Merchant Login is the owner of the
create their own workspace it will generate their
company, which do overall setup of the organization
own unique URL, which they can use further to
and User Login, is the individual login of the student
access Tuition Plus+. Users have to enter their
to check their personal details.
workspace to login in Tuition Plus+ to access their
organization login.

B. Organization Setting Module

This organization setting is done by the merchants.


Organization must setup their information like they
fill their complete details. They have to put their
financial approach. The branch is specifies if it has
number of branches. After login user have to setup
their organization by putting all essential information
in form user can upload logo of their organization it
will appear in students receipt. User can also create

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


359
Bharat Dhak et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 357-362

branches of their organization by clicking manage batch for subjects one or more subjects all students in
branch tab. It will set company as default primary that batch will show in batch list.
branch user can change primary branch. User can allocate batch to one student from student’s
profile.
C. Subject Module
It has two options batch for course and batch for
Merchant Login have the authentication to add subjects.
subject that are taught in their organization. User can 1) Batch for course:
create subjects from add button in subject It will show all batches, which have all
module.Once subjects are created user can create subjects of that course.
course through course panel. User can edit or delete 2) Batch for subjects:
subjects. Add subject for particular branch, which are It will show batch for every subjects of that
already created from organization setting. User can course.
select more than one branch to assign same subject in
multiple branches. F. Student Module

The student is of centre focus, because in every


D. Course Module
organization student plays the very important role.
Once subjects are, created user can create course User Login is the authorization, which is given to the
through course panel. Add course for particular student. In this module student can view their own
branch that are already created from organization information but cannot edit it. The information,
setting. User can select more than one branch to which is shown in this module, is the personal detail
assign same course in multiple branches. of the students as well as their payment detail.

E. Batch Module G. Account Module

Batch module has two steps to create batch In account, module user can manage fees, duration,
1) Create batch: tax and discount for courses.
2) This tab includes basic information about In this student, can able to pay in three different
batch name, batch capacity, start date, end ways:
date, start time, end time. A. Onetime payment
3) Batch subjects B. Instalment payment
In batch subjects user have to assign one or C. Monthly payment
more subjects in batch with trainer name In Payment Management, tab user can manage
subject start and end time, start and end date. course payment methods, instalment amount and
User can edit batch or can view batch. schedule of instalment.
1) One Time :
View batch In Onetime, payment method complete fees
It will show all details fill during batch creation. User should be paid at a time.
can add more subjects after creating batch from 2) Installment:
assign subject’s button. In Batch student, tab user can In installment user can create any number of
assign multiple students to single batch for separate installments and can schedule duration foe
subjects or for all subjects. After assigning students to that installment

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


360
Bharat Dhak et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 357-362

3) Monthly: more things automatically in less time. Tuition Plus+


In monthly installment user can create allows you to manage staff data and student data in
installment for per month. an observatory format that ensures data completeness,
accurate reporting and quick access to information. It
Payment details will show on print and on submit
is the Better, Smarter, Efficient and Very Simple Way
registration students will get conformation through
to Manage the amount of data at one Place. Tuition
SMS and email with registration receipt attachment
Plus is designed to help you and your Student to
in email.
develop the way of working or the use of Tuition
Plus makes you more smart for managing the
H. Enquiry Module
thousand of data of students which none can do.
This module will help organization to maintain all
V. REFERENCES
enquiries in records so that no any enquiry will miss
by any user in organization so the enquiry module
[1]. Web based student information management
will help to connect with the different interested
system S.R.Bharamagoudar1, Geeta R.B.2,
students. After filling enquiry form user can save
S.G.Totad3 Assistant Professor, Dept. of
enquiry it will reflect in enquiry list.
Electronics & Communication Engg,
Basaveshwar Engg. College, Bagalkot,
After saving enquiry user can add next follow-up for
Karnataka1 Associate professor, Department of
that same enquiry and can add next follow-up for
IT, GMR Institute of Technology, RAJAM,
that Enquiry.
Andhra Pradesh2 Professor, Department of
Enquiry record generated by
Computer Science & Engineering, GMR
1. Website
Institute of Technology, RAJAM, Andhra
2. Newspaper
Pradesh 3
3. Social Media
[2]. G.Satyanarayana Reddy,Rallabandi
4. Friends
Srinivasu,Srikanth Reddy Rikkula,Vuda
Sreenivasa Rao," Management Information
I. Report Module
System To Help Managers For Providing
Report module gives the answer of following Decision Making In An Organization",
questions: International Journal of Reviews in Computing,
1. How much student came for enquiry? ISSN: 2076-3328, EISSN: 2076-3336
2. How much student is interested to take [3]. Bishnu Prasad Gautam, Shree Krishna
admission? Shrestha,"Effective Campus Management
3. From where mostly the crowd came. through Web Enabled Campus-SIA (Student
4. Mode of attraction. Information Application)", Proceeding Of the
International MultiConference of Engineers
IV. CONCLUSION and Computer Scientists 2012 Vol I, IMECS
2012, March 14-16, 2012, Hong Kong
We have developed this product to managing all data
[4]. Design and implementation of student's
with less paper work and to maintain all details easily.
information management system(SIM) based on
In Tuition Plus+ our main objective is that tuition
Active Server Page(ASP) College of Electrical
classes can see their graph and they can able to do
Engineering ,Northwest University for

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


361
Bharat Dhak et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 357-362

Nationalities, Lanzhou Jiatong University,


Chaina yangchenghui36@163.com
[5]. Unnati A. Patel, Dr. Swaminarayan Priya R. ,"
Development of a Student Attendance
anagement System Using RFID and Face
Recognition: A Review", International journal
Of Advance Research In Computer Science and
Management Studies,Volume2, ISSN: 232
7782,Issue 8,August 2014
[6]. Unnati A. Patel," Student Management System
based on RFID Technology", International
Journal of Emerging Trends & Technology in
Computer Science (IJETTCS), Volume 2, Issue
6, ISSN 2278-6856, November – December
2013
[7]. Mayuri Kamble et al, / (IJCSIT) International
Journal of Computer Science and Information
Technologies, Vol. 6 (2), 2015, 1341-1344

Cite this article as :

Bharat Dhak, Prachi P. Paunikar, Simran S. Khan,


Snehal S. Ramteke, Yashkumar R. Jais, Priya R. Gupta,
"Tuition Plus+ For Managing Tuition Class Data ",
International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 357-362, January-February 2019.
Journal URL : http://ijsrcseit.com/CSEIT195197

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


362
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195195

Heikin-Ashi Transformation and Vix Index data for Stock Market


Index Prediction and It’s Effects
Nishchal Sharma*1, Dr. Chaman Singh Chauhan2
*1Research Scholar, School of Engineering and Technology, Career Point University, Kota, Rajasthan, India
2Assistant Professor, Dept. of Computer Science & Application, Govt. College, Chamba, Himachal Pradesh,
India

ABSTRACT

This paper performs a comprehensive analysis of Vix Index data with Heikin Ashi Transformation of stock
index Neural Network Learning. It has been demonstrated that Heikin Ashi Transformation can improve the
learning effect of Neural Network and the effect can also be filter out if volume weights are also considered.
This paper introduces another improvement beside using volume-weighted data. Instead volatility index is used
as an input and its effect for neural network learning process is analyzed.
Keywords : Heikin Ashi Transformation, Vix Data, Neural Network, Stock Market Analysis, Time series
filtering

I. INTRODUCTION like human beings. If a large training data is available


then neural network can learn it properties and
Predicting uncertain events have been a curious make prediction based on the learning [1] . One of
subject for human beings. we were able to make the most desired fields of such research includes stock
better prediction framework with the computation market. Stock market is a very dynamic place with
power we have. The science of examining tons of information. Stock market data creates
information and its effects is extremely valuable. The patterns, that when harvested can provide better
information itself is of no use if it is raw and decision making opportunity. But stock market data
unprocessed. The process of locating hidden patterns is full of unwanted entropy. Such noise can be a short
or data inside can be very useful. High power term fluctuation in stock market data or may be long
machine can turn this data into recognizable patterns term disturbence. Such Noise is undesired within the
far easily now. Identification of patterns inside given data and may or may not be related to the underlying
data is a key variable in modern research. Since we stock data trend. We have already demonstrated that
have abundance of computational power available to such noise can interfere with learning of neural
us, we would formulate techniques that enable network and hence reduces the effectiveness of
machine to analyse data patterns and discover learning and its effectiveness. The Heikin-Ashi
underlying information easily. transformation has worked remarkably well reducing
noise and its effect can be further improved using Vix
Neural network are system that mimic human brain Index data.
and allow machine to learn underlying patterns just

CSEIT195195 | Received : 09 Feb 2019 | Accepted : 18 Feb 2019 | January-February -2019 [ 5 (1) : 363-365 ]
363
Nishchal Sharma, Dr. Chaman Singh Chauahn Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 363-365

Some Researchers [2] have demonstrated the HA-Close = (Ocurrent + Hcurrent +Lcurrent+Ccurrent)/ 4
importance of availability of favorable methods for HA-Open = (HA-OpenPrev + HA-ClosePrev) / 2
finance managers and incorporating many systems HA-High = Maximum of the Hcurrent , HA-Open or HA-Close
for portfolio management. Such connection of HA-Low = Minimum of the LCurrent , HA-Open or HA-Close
selection of technology and company is vital [3] for a
finance driven system. Artificial Neural network has Using Vix Data:
been a very important learning system [4] and it has
provided a definite edge for it’s user. The effect of Vix is a volatility index and tracks volatility measured
neural network and genetic algorithm has been as per the options of stock index. Since the value of
researched [5] and found to be very effective. Neural Vix may very between some fraction value, it is
network for trend determination [6] and optimized converted between [0,1.0] using inversion.
neural network [7] are a few key methods to work
with. Many researchers have incorporated hybrid A 2 hidden layer ANN is used with configuration
models [8] and had early success. Generating buy and 5,6,3,1. The choice of structure of ANN is quite
sell signal using deep neural network [9] as well as a subjective. This setup given relatively stable results.
self organizing rule based system [10] perform very Again test data is converted into 2 sets. One is for
well. Effect and result of ANN on NASDAQ learning and other for test results.
Composite [11] and results are inline with
expectations. III. RESULTS AND DISCUSSION

II. METHODS AND MATERIAL A Plot of Prediction by ANN and its original value
provides us with an absolute error. This absolute
Heikin-Ashi is widely used in candlestick charts. In error is plotted against Vix data and results are
Japanese, Heikin means “average” and “Ashi” means compared.
“pace”. Since it is based on candlestick patterns. We
give a brief intro to candlesticks. A candlestick Mean Average Error 11.87130668
Mean error when Vix is above 19 14.00708243
describes trade information of a given time period.
Mean Error when Vix is below 16 8.658732212
Every candlestick has 4 values Open, High, Low and
Close. These values represent the value at which
stock or index candle opened, highest value of stock,
lowest value of stock and closing value of the stock in
that time period. Heikin-Ashi transforms these 4
values using following formulas to create new
candlesticks. New Candlestick produces simpler and
smoother candlestick for analysis.

Transformation rules:
Let Ocurrent, Hcurrent , Lcurrent, Ccurrent represents current
open, high, low, close values.
Let OPrev, HPrev, LPrev, CPrev represents Previous
day/period open, high, low, close values. These results clearly show an increase in error rate
Then Heikin-Ashi values (HA) are calculated as:-

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


364
Nishchal Sharma, Dr. Chaman Singh Chauahn Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 363-365

when Vix is High. A vix above 19 causes a spike in [8]. "A hybrid stock trading framework integrating
error rate. Also a lower level of vix causes low error technical analysis with machine learning
rate. However variation in lower level of vix is not techniques - ScienceDirect."
linear with the error rate. This phenomenon can’t be [9]. "A Deep Neural-Network Based Stock Trading
explained and a further study might be needed. System Based on Evolutionary Optimized
Technical Analysis Parameters - ScienceDirect."
IV.CONCLUSION [10]. "Establishing Rules for Self-Organizing
Systems-of-Systems - ScienceDirect."
[11]. "Stock market index prediction using artificial
It is clear that Vix data may be an indication of
neural network - ScienceDirect."
volatility of underlying system and the noise may
increase during high period of volatility which affect
the learning capability of ANN but it is also clear that Cite this article as :
there is a still a disconnect between vix and stock
market movement and a further research might be Nishchal Sharma, Dr. Chaman Singh Chauhan,
needed. "Heikin-Ashi Transformation and Vix Index data for
Stock Market Index Prediction and It's Effects ",
V. REFERENCES International Journal of Scientific Research in
Computer Science, Engineering and Information
[1]. "Stock Market Volatility and Learning - ADAM Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 363-365, January-February 2019.
- 2016 - The Journal of Finance - Wiley Online
Available at doi :
Library."
https://doi.org/10.32628/CSEIT195195
[2]. "Data | Free Full-Text | Current Core Journal URL : http://ijsrcseit.com/CSEIT195195
Competencies Trend of Small to Medium
Enterprises (SMEs) in China—A Concurrent
Comprehensive Evaluation and Active Learning
Study of Newly Listed Chinese Stocks from
2015 through 2017."
[3]. "The contribution of technology selection to
core competencies - ScienceDirect."
[4]. "Using artificial neural network models in stock
market index prediction - ScienceDirect."
[5]. "Application of artificial neural network for the
prediction of stock market returns: The case of
the Japanese stock market - ScienceDirect."
[6]. "Predicting stock and stock price index
movement using Trend Deterministic Data
Preparation and machine learning techniques -
ScienceDirect."
[7]. "Predicting the Direction of Stock Market
Index Movement Using an Optimized Artificial
Neural Network Model."

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


365
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195185

Factors Affecting the Successful Implementation of Cloud and


IoT - A Study on IT and Business Leaders
Jipson George Thoomkuzhy*, Dr. Mohammed Nazeh
Department of Management Studies, Limkokwing University of Creative Technology, Cyberjaya, Selangor,
Malaysia
ABSTRACT

The Internet of Things (IoT) is turning into the following Internet-related transformation. It enables billions of
gadgets to be associated and speak with one another to share data that enhances the nature of our day-by-day
lives. Then again, Cloud Computing gives on-request, helpful and versatile system get to which makes it
conceivable to share computing assets, without a doubt, this, thus, empowers dynamic information integration
from different information sources. There are numerous issues hindering the fruitful implementation of both
Cloud and IoT. The integration of Cloud Computing with the IoT is the best path on which to defeat these
issues. The huge number of assets accessible on the Cloud can be greatly helpful for the IoT, while the Cloud
can acquire attention to enhance its impediments with true protests in a more powerful and dispersed way. This
paper gives to the various factors that affect the integration of the Cloud into the IoT by featuring the
integration advantages and implementation challenges. This paper mainly focus on a study which was
conducted on various IT and business leaders.
Keywords : Cloud, IoT, CIO, SLA, QoS

I. INTRODUCTION conditions of under-used IT framework. With its


particularly autonomic character, cloud-computing
Cloud empowers its dynamic accessibility takes out a colossal bit of the time largely required to
applications and framework, paying little respect to demand and plan IT assets. Cloud computing likewise
locale. Moreover, fast hierarchical advancement yields basic cost adventure sponsors in the land
works out as intended in perspective of the required for server utilization and besides power and
mechanical and psychological capacity to deal with cooling costs. As a result of virtualization and the
the capacity to make or arrange a game plan and fuse cloud's ability to tap assets (either through a private
enrolling power in its assistance and business benefits cloud or tapping direct open cloud assets), server
more quickly than would be conceivable with the estates can manage the chose weight to expand their
present rising organizations. Cloud computing in like physical impression.
way advances IT streamlining with the target those
IT assets is planned for most essential money saving II. LITERATURE REVIEW
favoured outlook. This is conceivable in light of the a) Reliability
way that cloud computing underpins gigantic Enduring quality is one of the key elements to be
adaptability to manage times of intrigue while considered in coursed computing condition. Decided
keeping up an imperative partition from extended quality is depicted as the likelihood that a given thing

CSEIT195185 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 366-396 ]
366
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

will play out its typical farthest point concerning a like way identified with fearless quality: an
given time distribution under a given strategy of organization that is on 24x7 yet goes continually
conditions. A cloud will be more dependable on the isolated is futile. As cloud organizations and
off chance that it is more denounce tolerant and applications wind up being even more amazing and
more adaptable to propelling conditions. It is difficult more dependent on the covered cloud mastermind, it
to have a cloud that is totally free from sorrows or twists up continuously hard rapidly perceiving and
dissatisfaction safe [Garrison, 2012]. Different kinds managing issues. Weights can ascend out of an
of dissatisfactions are interleaved in the passed on individual organization, and furthermore from the
computing condition, for example, flood bafflement, collaboration of various parts and mechanized
timeout dissatisfaction, asset missing dissatisfaction, structures over spread frameworks and server farms,
coordinate dissatisfaction, equipment dissatisfaction, acknowledging issues that set aside a long
programming frustration, and database opportunity to be settled.
disillusionment.
c. Scalability:
The boundless idea of cloud computing is Lee and Kim showed programming centred
uncommonly fundamental yet hard to isolate on the procedures for ensuring the high versatility of
grounds that cloud computing is comprised of the administrations in Cloud Computing. [Lee and Kim,
blend of different factors, for example, wide-zone 2006] High flexibility under high administration
frameworks, heterogeneous programming/adapt parts, loads seeks nothing out of pocket; it very well may be
and so on. There are different confused trades among ensured by grasping some versatility confirmation
the various parts of cloud computing. Thusly, the designs. The most broadly perceived conventional
standard quality models that are depicted for arrangement is to simply incorporate the required
unadulterated programming/outfit or regular resources. Nevertheless, as they propose in the paper,
frameworks cannot be connected with the surveying diverse plans can ensure high adaptability or
of cloud unwavering quality [Pocatine, 2010]. The adaptability.
downtime of the cloud server develop negatively
affects the cloud advantage enduring quality. There is always some cost drew in with running
Surveying and re-designing undeterred quality in versatility ensuring plans, for instance, the expense of
coursed computing is not so much direct errand. additional CPU and memory. The adaptability
Notwithstanding these, two or three contraptions are expanded through such plans should be with respect
utilized to overview and overhaul the cloud to the expense of applying the plans. That is, cost-
advantage consistency. reasonability should be considered in ensuring
flexibility. Administrations should give the level of
b) Availability the esteem decided in their SLAs. Administrations
High transparency is, at last, the colossal vessel of the with palatable flexibility should not to encounter the
cloud. It encapsulates the probability of wherever evil impacts of significant degradation of QoS.
and at whatever direct access toward organizations, Adaptability affirmation designs should ensure that
mechanical congregations, and information and is the administrations satisfy the confinements of meeting
connecting with the effect of dreams of a future with the immaterial edge of their QoS properties
organizations with no physical workplaces, generally [Collinchau, 2013]. From this, two convincing
speaking organizations with totally arranged and programming centreed plans can be resolved:
joined IT frameworks. Transparency or openness is in advantage replication and administration

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


367
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

development. Administration replication is a picked here is the number of dynamic sessions or


methodology for cloning administrations that are logon sessions in each web application.
successfully running on interchange centres to propel
the administration stack over the centres without In perspective of the moving typical of the scaling
impacting errands ahead of time. Rehashed pointer, a dynamic scaling computation is used to
administrations secure additional advantages gave by trigger a scaling event to the provisioning subsystem.
the new centre points for dealing with greater Dependent upon the revived bits of knowledge,
administration stack. Figuratively speaking, movement to scale up or down may be begun. Scaling
advantage replication enhances advantage flexibility up or down suggests that an event will be set off that
and diminishes the threat of QoS corruption by instructs the provisioning subsystem to start or close
dealing with greater advantage loads. To exhibit a down web-server virtual machine events in the cloud.
sufficiency of administration replication, they imitate Right when the web application is scaled up, the as of
the advantage replication plot for the seventeen late started virtual machine event will run the web
extraordinary volumes of administration stack application. After the web application events are
[Collinchau, 2013]. On each advantage stack, they readied, the front-end stack balancer configuration
take a gander at general administration system with record is then invigorated and resuscitated to place
advantage replication plot the extent that typical them into dynamic administrations. As determined
response time. heretofore, the scaling estimation is executed in the
Service Monitor sub-system and is used to control
Administration migration is an arrangement for and trigger the scale-up or down in the Provisioning
setting an administration on a choice centre when a sub-structure on the quantity of virtual machine
particular centre can't give high QoS in view of a cases in light of the estimations of the scaling marker
physical issue or programming issue. After [Frey, Reich, et al. 2013].
administration migration, the moved administration
has out a vague influence from it did on the insecure d) Security
centre, and the volatile centre point is ousted from Like PC security programs, cloud security
the once-over of administration centre points. consolidates a relative general concern: keeping up
Components, advantage stack and machine of the validity of data, guaranteeing access is limited to
adaptability ensuring plan, are same with advantage underwrite clients, and keeping up the availability of
replication re-enactment. In this re-enactment, they data and organizations (EDUCAUSE, 2010).With
have a circumstance that an administration is drawn cloud computing, the data and organizations are
to a closer centre from clients. To enable this, they outer to the grounds and in like way controlling
acknowledge that the response time is in respect to moreover, securing these points of interest changes
the partition. Appropriately, an administration is into an impressively more unpredictable and testing
moved to the speediest centre point to the extent that proposition. Data encryption, e-exposure, rehash and
the response time. They developed a web application dependability of data fortresses and recuperation of
for controlling the provisioning and de-provisioning data, the entire arrangement sensibility of the cloud
of web-server VM events, a dynamic scaling seller and laws concerning farthest point and access
estimation in perspective of the huge point of to data all breeze up fundamental issues. It is central
confinement or scaling marker of the web application that individual and controlled data are sincerely
[Chen and Kunz, 2016]. The scaling marker that is ensured and that they are appropriately overviewed.
Fratto (2009) called attention to that the security of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


368
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

data remains the commitment of the data a cloud setting, this test is befuddled by the circled
proprietor/outsourcer and scattered computing thought of fogs and the possible nonattendance of
professional organizations are not plainly subject for supporter care over where data is secured and who
data misfortune or fines or other lawful. Like PC has or can approach.
security programs, cloud security consolidates a
practically identical general concern. Data Framework Integrity: the cloud requires security
orchestrated in an open cloud and applications against deliberate subversion or mischief of the value
running in an open cloud may encounter specific of a cloud. Inside a cloud, there are accomplices:
security exposures than would be the situation in an endorsers, providers, and a combination of chiefs.
on region energized condition. The capacity to fragment get to rights to each one of
these social occasions, while monitoring harmful
Right when an association purchases into a cloud, attacks, is a key normal for keeping up cloud
each one of the data created and arranged will reliability. In a cloud setting, any nonappearance of
physically abide in premises had and worked by a porousness into a cloud's frameworks makes it more
provider. In this extraordinary situation, the troublesome for endorsers of checking the
fundamental issue is whether a supporter can obtain uprightness of cloud-encouraged applications. Cloud
an attestation that a provider is executing the computing administrations use the Internet as a
equivalent or proportionate controls as for what the transmission medium and change data development
endorser would have completed. The going with resources into administrations for end-customers,
issues develop when an endorser is endeavoring to including programming administrations, figuring
ensure scope for these controls: Compliance stage administrations, change organize
necessities, regarding data that an endorser is a administrations, and crucial structure leasing.
significance to move to a cloud, may call for specific
levels and granularities of audit logging, the time of Vaquero and Lindner prescribed that cloud
cautions, activity declaring, and data support. For computing could be described as the joining of virtual
encryption of data still, the quality of the encryption resources according to customer necessities,
figuring suite, the key administration plots a provider adaptably solidifying resources including hardware,
reinforces, and the number of keys for each datum change stages, and diverse applications to make
proprietor (individual or shared keys) should be administrations [Faisal, 2011]. The novel features of
known by the data proprietors. Data arranged in an cloud computing join the capacity of customer data in
open cloud and applications running in an open the cloud and the nonattendance of any prerequisite
cloud may experience particular security exposures for programming foundation on the client side. For
than would be the circumstance in an on-area whatever timeframe that the customer can interface
encouraged condition. with the Internet, most of the gear resources in the
cloud can be used as a client-side structure. When in
Data Privacy: Security watches out for the protection doubt, cloud-computing applications ask for driven,
of data for specific components, for instance, giving distinctive administrations as demonstrated by
endorsers or others whose data is taken care of in a customer requirements, and expert communities
structure. Assurance passes on authentic and chance charge by metered time, instances of usage, or
concerns and should be viewed as a particular test portrayed period.
and also a legal and good concern. Guaranteeing
insurance in any handling system is a specific test; in

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


369
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

The use of passwords as an affirmation strategy is in a remote server requires a costly expense of
even more outstanding to general customers, yet transmission and capacity. To alleviate the expense of
messages sent by the customer are weak against the transmission and capacity, it is required to partition
covert record by software engineers who might then IoT data into two sorts of data; one is put away in
have the capacity to use the data in the message to edge systems and the other is put away in cloud
sign into the administration as the customer. systems. The impact of Edge computing is uncovered
with the treatment of IoT data in edge systems.
Additionally, the One-Time Password (OTP)
approval system fluctuates from a great many people As of not long ago, most system hardware, for
gatherings' beginning of a secret word. By far, most example, switches, passages, and switches simply
grasp a mystery word to be a watchword picked by forward data conveyed from other system gadgets,
the customer to be critical and can be used repeatedly. not to peruse the substance or change them. In light
The complement of OTP, in any case, is the single- of end-to-end communication, data is recognized and
use nature of the mystery key. In the wake of getting continue at a last relating hub. This is a common use
affirmation from the customer, the structure side of cloud computing a customer service
must make an ensured transmission channel to communication. Be that as it may, in the IoT
exchange data with the customer. The Secure environment, some IoT data will be exchanged to a
Attachments Layer (SSL) is a run of the mill strategy cloud system and some IoT data will be conveyed to
for building secure channels, primarily using RSA an edge hub/haze hub. The primary purpose behind
encryption to transmit the secret keys required for this partition is to give constant preparing and
the opposite sides to encode and unravel data security upgrade. Despite the fact that there are
transmitted between them [Zheng, 2009]. Consistent numerous new advancements to diminish the
methodologies for guaranteeing customer data join postpone time and transmission time, it is difficult to
encryption going before a capacity, customer ensure ongoing handling. The commonplace utilize
approval strategies before capacity or recuperation, instance of this necessity is the Industrial Internet
and building secure channels for data transmission. and savvy processing plant. What's more, despite the
These security systems commonly require fact that, there are control capacities to give security,
cryptography figurings and propelled stamp the more fundamental decision is that not to uncover
techniques, as illuminated underneath. Ordinary data the private data to open systems. On the off chance
encryption systems fuse symmetric and Hilter kilter that we separate IoT data into private data and non-
cryptography computations. private data and keep private data inside an edge
organize, not to uncover them in an open system, it
e) Capacity will diminish numerous feeble purposes of security.
As huge IoT sensors, IoT actuators, and IoT gadgets
are associated with the Internet, IoT data volume In the event that we consider new difficulties of IoT
from these things are required to increment administrations, the enormous volume of IoT data, as
dangerously. Also, it is normal that a significant well as the gigantic number of IoT things, can be a
number of these high volumes of IoT data are created basic issue. Despite the fact that we recognize this
and additionally devoured inside edge systems, not to future issue, Internet design initially has the capacity
navigate through cloud systems. As of recently, of scalability and it will alleviate the scalability issue
primarily IoT data produced IoT things are moved in the IoT environment. In any case, we can't assess
and gathered in a remote server and to store IoT data the quantity of IoT things later on and we can't

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


370
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

ensure the Internet engineering still manage the interviewees avowed that execution issues are caused
scalability issue in the IoT environment. Edge generally by the low web association the adopters
computing will isolate the scalability area into edge have in there. In like way, little organizations should
systems and outside system (e.g., cloud systems) and ensure to have a snappy Web association which
this partition of scalability space can give a more would reduce inaction (Marston et al., 2011). As
effective approach to handle the enormous number of demonstrated by Thurman (2008), execution is one of
IoT things. Since Edge computing can deal with IoT the troubles of cloud advancement execution
data in an edge zone and store the IoT data in an contiguous data security, data affirmation, and
edge/haze hub, and continue IoT data on the off availability. There are challenges in grasping
chance that it is required, it can likewise isolate the conveyed computing advancement. Key execution
administration area into two sections. Edge obstacles are application quality (i.e. application
Computing can focus on the administration of IoT execution likewise, versatility), application faithful
things in an edge zone and collaborate with the quality (i.e. persevering administrations) and data
administration of other outside systems. security and insurance concern (i.e. illegitimate
access of organization data, security bursts) (Benlien
f) Throughput et al., 2009). In this manner, execution is a factor for
Throughput implies the execution of errands by a organizations who wish to get Cloud courses of
figuring administration or contraption over a specific action and use them capably.
period. It checks the proportion of completed clash
with time exhausted and may be used to evaluate the g) Lower costs
execution of a processor, memory and also In cloud computing, setting up an organization as a
framework trades. Throughput was considered to utility, and giving figuring as an organization to
evaluate the benefit of PC processors. This was, clients is a piece of the real work. One of the
generally, registered with respect to bunch beguiling and fundamental cost issues in cloud
occupations or assignments consistently and computing is the capacity to pay for the favorable
countless consistently. A couple of backups measure a position, in the viewpoint of necessities, keeping
structure's general throughput by evaluating the total away from the significant costs for PC frameworks
and unusualness of work, the number of synchronous purchases (Armbrust, 2009). Consequently, cost
customers and application/system responsiveness. sufficiency is another basic factor bearing
Additionally, for arranging trades, throughput is organizations should note while tolerating spread
evaluated by discovering the proportion of data computing. The organizations of today dependably
traded between zones in the midst of a foreordained filter for approaches to manage cut expense or
period, overall occurring as bits consistently (bps), additional cash. Organizations need to spend a critical
which has created to bytes consistently (Bps), piece of their wage on the IT foundation, while
kilobytes consistently (Kbps), megabytes consistently under 10% of their servers can be incredibly used,
(Mbps) and gigabytes consistently (Gbps). accomplishing an essential maltreatment of cash. Also,
these servers should be supplanted respectably at
The physical zone of the data can influence the standard intervals and ought to be kept up and
execution of the system if it is arranged in another bit administrated; expanding its aggregate cost practices
of the world. In this way, Cloud plans may significantly (Marston et al., 2011). Cost perfect
comprehend the postponement in the administration conditions are the most grounded driver influencing
(Kim et al., 2009). Concurring the result consider the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


371
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

IT chiefs' impression of SaaS openings (Benlian et. al, Organizations can in like manner immensely
2011). diminish their help costs (Miller, 2009). This implies
both gear and programming upkeep.
Above all, cloud computing enables organizations to
decrease their gear costs (Miller, 2009). While using According to Miller (2009 p. 26), organizations are
cloud administrations, organizations never again "never again compelled to what a singular work
require intense and high-evaluated PCs to run region PC can do" and now can "perform
applications inside the cloud. This starts with the supercomputing – like assignments utilizing the
lessened requirements for a dealing with power and vitality of countless and servers." Thus they can
storage space (Miller, 2009). Not in any way like perform more conspicuous errands and augmentation
standard programming, for running cloud has by and large their benefit and capability. On the
applications had PCs required less memory. They other hand, from a provider`s perspective, the basic
moreover can be with minimal hard plates in light of cash sparing favorable circumstances for the vendors
the way that there isn't foundation programming. are the foreseen employments and the more broad
Therefore, organizations can diminish costs by customer base (Goncalves and Ballon, 2011).
procuring lower-esteemed PCs. Since using cloud Disseminated computing cuts down legitimate
computing, organizations don't have to do high expenses and empowers an association to offload
interests in the IT system. This especially concerns three sorts of organization (Rosenthal et al., 2009).
the greater organizations (Miller, 2009). Instead of
putting a colossal proportion of trade out an extensive III. SURVEY AND TOOLS USED FOR THE SURVEY
number of extreme servers, the IT divisions of those
organizations can use the cloud` figuring vitality to The survey was conducted using online methods. The
supplant or improve the internal preparing resources. survey attempted to explore the major challenges
Plant administrator (2009) states, as Armbrust at al. faced by CIOs in cloud computing and IoT based
(2009), that organizations don't have to manage the organizations from various industries. This study was
zeniths and the nonpeak times in figuring power asks conducted on IT and business leaders/professionals
for any more drawn out by getting new equipment, from various countries across the globe. A sample size
the apex figuring demands are dealt with by cloud of 405 was used for this survey. Since the target
servers. Cloud computing furthermore prompts cut population is unknown the baseline was of 385 has to
down programming costs. be maintained. The sample size of 405 was used in
order to have an effective result after excluding the
Organizations never again need to buy disconnect errors. A website named e-mailmeform.com was used
programming packs for each PC (Miller, 2009). Or to create an online survey. It is a paid services based
maybe, a particular application is gotten to simply by portal which provides survey setup services. This has
the laborers using that application. Moreover, this in security features like one response from one
like manner infers saved the expense of presenting computer and one response from one IP address.
and keeping up that item on each PC. Another item These sort of security features would help to avoid
related cash sparing preferred standpoint is that duplicate submissions resulting in improved quality
organizations don't have to pay for an item refresh in responses. A total of 40 questions were asked to the
order to have the latest variations of the applications respondents and the results were captured and
(Miller, 2009). As all applications are in the cloud, correlated using the IBM SPSS statistical tool.
they are redesigned thusly by the provider.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


372
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

IV. SURVEY RESULTS & DISCUSSIONS


In all, the majority of the respondents are well above
Altogether, about 346 male and 59 female forty years of age while the least frequency of the
respondents were surveyed to explore the major respondents falls between ages 36 and 40. From the
challenges faced by CIOs in cloud computing and IoT pool of respondents, it’s a single participant that has
based organizations from various industries purely his highest form of education as at then being
from the healthcare, manufacturing and service line. PUC/XII, representing a minute 0.2% of the total
Thus the totality of the participants equals about 405 respondents. The number of graduate-respondent
in number. The feminine gender has the less valid elevated as high as 127 people in total representing
percent of participation as compared to the male 31.4% of the total respondents. A whooping 192
counterparts. participants claimed to be holders of a post-graduate
degree or qualification (47.4 % of the population)
We can therefore infer that there is likelihood of while, about thirty-five of them already has a PhD
more males functioning as IT business leaders than degree in relevant disciplines. Only forty-seven of
their female counterparts. This singular fact was the respondents studied a professional course
considered in the distribution of the survey accruing to about 11.6 percent of the total population.
questionnaires. Out of the 405 respondents to the The remaining 0.7% of the total respondents, that is
online questionnaire, only 55 people (both males and only three people, had a different educational
females) falls within the age range 21 to 25 years of background from the listed.
age; eighty-six participants similarly fell within the
age bracket of 26 to 30; whereas fifty-two
respondents that participated in the online survey
falls between age 31 to 35. The number of
participants between the age ranges of 36 to 40 is
forty-eight while the rest participant, about one-
hundred and sixty-four, are all over 40 years of age.

Fig 3 : Education

According to the population studies, about forty-one


team leaders responded to the online survey
representing a significant percentage of 10.1 percent.
Fig 1 : Gender
Only eighteen of the participants that responded
were junior managers in rank while thirty-one
participants were senior officers or managers within
their respective organizational domain. The number
of respondents that were vice-presidents or general-
managers amounted to thirty-five in all whereas a
whooping one hundred and eighty-two of the
participant were actually chief information officers,
Fig 2 : Age chief executive officers and/or chief functional

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


373
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

officers-which accounts for almost half percent of the


population. The remnant of the 405 respondents
occupied other organizational positions not reflected
in the listed designations.

Fig 5 : Overall Experience

Most of the volunteers who responded, about a


whooping population numbering about 197 people,
only have had between one to five years of
experience as a chief information officer (CIO) in
Fig 4 : Designation their present position which may range from junior
manager to senior managers, etc. Only a handful of
With respect to their overall experience, most of sixty-three persons who participated in the online
the respondents had less than ten years of experience survey have only had less than a year of experience as
and exposure concerning the job description of a related to the job competencies of a chief information
chief information officer; that is, about 169 officer in any position they may be holding. Then it
respondents had lesser that ten years’ experience of was calculated that about seventy-four of the total
the subject matter accounting for 41.7 percent of the 405 respondents had only served between five to ten
total population. years in the capacity of a chief information officer in
any capacity or position. The rest seventy-one
In the same vein, ninety-six of the participant had an respondents have had over ten years of experience in
overall experience and exposure of greater than or their present position with respect to the subject
equal to ten years, but lesser than 20 years at the matter.
same time. Also, ninety-nine of the participants had
already an overall experience worth more than
twenty years but below thirty years of age. The
remainder forty-one respondents already had
experiences of the subject matter over accumulated
years, typically above thirty years. Thus, the overall
percent of respondents with the highest frequency
suggests those of years below ten while those with
Fig 6 : Experience in Present Position
years above thirty had the lowest overall percentage.

The number of respondents within the


telecommunication, service and information
technology (IT) industries that participated in the
online survey was 226 in number representing a
whopping 55.8 percent of the total population.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


374
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

About twenty-five participants interviewed are


currently working in certain manufacturing
industries accounting for 25% of the totality of
respondents. Forty-seven of the respondents are
working in the healthcare sector of the economy
whereas thirty-three of the respondents are
domiciled in the finance, banking or the insurance
companies, with the rest seventy-four of them
practicing in other spheres of discipline accruing a
whole 18.3 of the percentage of the whole population.

Fig 9.

As showed in the data computed above, exactly 142


respondents out of the total number of people who
responded, which is 405 in number, strongly agreed
with the fact stating CIO should be able to answer
Fig 7 : Industry sector of your present organization how an IT solution is beneficial to the company.
Following this wise, a whopping number of 115
From the table provided above, majority of the CIO respondents, representing 28.4 percent of the total
respondents interviewed are currently working in respondents also did agree though not strongly with
either an IT/telecommunication and service firms the fact stated above. Meanwhile, very few
accounting for about 226 respondents in all. In respondents, about 32 in number and accounting for
addition, 25 of the remainder are from the 7.9 percent of the total respondents, disagreed with
manufacturing industries; 47 from healthcare settings that: Do you think that the changed role of CIOs has
while another 33 of the remaining respondents are resulted in successful implementation of Cloud
currently on a job with a finance/insurance computing and IoT based technologies? Only four of
companies. The others could not fit into any of the these respondents who disagreed, strongly expressed
afore-mentioned categories, thus representing 18.3 their own disagreement while about 112 of the
percent of the total number of respondents. remainder participants went neutral and
indeterminate on the matter.

a. Reliability and Availability

Fig 8 : Do you think that the changed role of CIOs


has resulted in successful implementation of Cloud
computing and IoT based technologies.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


375
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

FIG 10: Our cloud and IoT platforms have minimum


likelihood, collapse, or failure

Fig 13

As showed in the data computed above, exactly 66


Fig 11
respondents out of the total number of people who
responded, which is 405 in number, strongly agreed
As showed in the data computed above, exactly 57
with the fact stating CIO should be able to answer
respondents out of the total number of people who
how an IT solution is beneficial to the company.
responded, which is 405 in number, strongly agreed
Following this wise, a whopping number of 215
with the fact stating CIO should be able to answer
respondents, representing 53.1 percent of the total
how an IT solution is beneficial to the company.
respondents also did agree though not strongly with
Following this wise, a whopping number of 181
the fact stated above. Meanwhile, very few
respondents, representing 44.7 percent of the total
respondents, about 21 in number and accounting for
respondents also did agree though not strongly with
5.2 percent of the total respondents, disagreed that as
the fact stated above. Meanwhile, very few
a CIO, in case of cloud/ IoT breakdown, we have
respondents, about 32 in number and accounting for
adequate arrangements to recover. Only four of these
7.9 percent of the total respondents, disagreed that as
respondents who disagreed, strongly expressed their
a CIO, our cloud and IoT platforms have minimum
own disagreement while about 99 of the remainder
likelihood, collapse, or failure. Only seven of these
participants went neutral and indeterminate on the
respondents who disagreed, strongly expressed their
matter.
own disagreement while about 128 of the remainder
participants went neutral and indeterminate on the
matter.

Fig 14: We us quick and effective tools for addressing


Fig 12: In case of cloud/ IoT breakdown, we have issues in our cloud and IoT platforms
adequate arrangements to recover

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


376
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 17
Fig 15

As showed in the data computed above, exactly 74


As showed in the data computed above, exactly 69
respondents out of the total number of people who
respondents out of the total number of people who
responded, which is 405 in number, strongly agreed
responded, which is 405 in number, strongly agreed
with the fact stating CIO should be able to answer
with the fact stating CIO should be able to answer
how an IT solution is beneficial to the company.
how an IT solution is beneficial to the company.
Following this wise, a whopping number of 226
Following this wise, a whopping number of 225
respondents, representing 55.8 percent of the total
respondents, representing 55.6 percent of the total
respondents also did agree though not strongly with
respondents also did agree though not strongly with
the fact stated above. Meanwhile, very few
the fact stated above. Meanwhile, very few
respondents, about 9 in number and accounting for
respondents, about 15 in number and accounting for
2.2 percent of the total respondents, disagreed that as
3.7 percent of the total respondents, disagreed that as
a CIO, users can confidently use our cloud and IoT
a CIO, we possess quick and effective tools for
platforms continuously. Only three of these
addressing issues in our cloud and IoT platforms.
respondents who disagreed, strongly expressed their
Only four of these respondents who disagreed,
own disagreement while about 93 of the remainder
strongly expressed their own disagreement while
participants went neutral and indeterminate on the
about 93 of the remainder participants went neutral
matter.
and indeterminate on the matter.

Fig 18: We have taken enough measures to backup


and store data to be easily retrieved during
Fig 16: Users can confidently use our cloud and IoT emergencies
platforms continuously

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


377
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 21
Fig 19
As showed in the data computed above, exactly 79
As showed in the data computed above, exactly 96 respondents out of the total number of people who
respondents out of the total number of people who responded, which is 405 in number, strongly agreed
responded, which is 405 in number, strongly agreed with the fact stating CIO should be able to answer
with the fact stating CIO should be able to answer how an IT solution is beneficial to the company.
how an IT solution is beneficial to the company. Following this wise, a whopping number of 209
Following this wise, a whopping number of 201 respondents, representing 51.6 percent of the total
respondents, representing 51.9 percent of the total respondents also did agree though not strongly with
respondents also did agree though not strongly with the fact stated above. Meanwhile, very few
the fact stated above. Meanwhile, very few respondents, about 31 in number and accounting for
respondents, about 18 in number and accounting for 7.7 percent of the total respondents, disagreed that as
4.4 percent of the total respondents, disagreed that as a CIO, the cloud/ IoT platform is up and working at
a CIO, we have taken enough measures to backup all the times. Only five of these respondents who
and store data to be easily retrieved during disagreed, strongly expressed their own disagreement
emergencies. Only five of these respondents who while about 81 of the remainder participants went
disagreed, strongly expressed their own disagreement neutral and indeterminate on the matter.
while about 76 of the remainder participants went
neutral and indeterminate on the matter.

Fig 22: Our cloud platform experiences minimum


downtime

Fig 20: The cloud/ IoT platform is up and working at


all the times

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


378
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 23
As showed in the data computed above, exactly 98 Fig 25
respondents out of the total number of people who
responded, which is 405 in number, strongly agreed As showed in the data computed above, exactly 84
with the fact stating CIO should be able to answer respondents out of the total number of people who
how an IT solution is beneficial to the company. responded, which is 405 in number, strongly agreed
Following this wise, a whopping number of 192 with the fact stating CIO should be able to answer
respondents, representing 47.4 percent of the total how an IT solution is beneficial to the company.
respondents also did agree though not strongly with Following this wise, a whopping number of 213
the fact stated above. Meanwhile, very few respondents, representing 51.6 percent of the total
respondents, about19 in number and accounting for respondents also did agree though not strongly with
4.7 percent of the total respondents, disagreed that as the fact stated above. Meanwhile, very few
a CIO, Our cloud platform experiences minimum respondents, about 11 in number and accounting for
downtime. Only four of these respondents who 2.7 percent of the total respondents, disagreed that as
disagreed, strongly expressed their own disagreement a CIO, our cloud and IoT platforms are prepared to
while about 92 of the remainder participants went withstand server failures. Only two of these
neutral and indeterminate on the matter. respondents who disagreed, strongly expressed their
own disagreement while about 95 of the remainder
participants went neutral and indeterminate on the
matter.

Fig 24 : Our cloud and IoT platforms are prepared to


withstand server failures

Fig 26 : Our cloud and IoT platforms are resilient


enough to endure power outages, network failure and
environmental damage.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


379
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 29:

Fig 27: As showed in the data computed above, exactly 98


respondents out of the total number of people who
As showed in the data computed above, exactly 93 responded, which is 405 in number, strongly agreed
respondents out of the total number of people who with the fact stating CIO should be able to answer
responded, which is 405 in number, strongly agreed how an IT solution is beneficial to the company.
with the fact stating CIO should be able to answer Following this wise, a whopping number of 211
how an IT solution is beneficial to the company. respondents, representing 52.1 percent of the total
Following this wise, a whopping number of 210 respondents also did agree though not strongly with
respondents, representing 51.9 percent of the total the fact stated above. Meanwhile, very few
respondents also did agree though not strongly with respondents, about 9 in number and accounting for
the fact stated above. Meanwhile, very few 2.2 percent of the total respondents, disagreed that as
respondents, about 13 in number and accounting for a CIO, we have taken measures for automated
3.2 percent of the total respondents, disagreed that as backups in case the cloud fails. Three of these
a cio, OUR CLOUD AND IOT PLATFORMS ARE RESILIENT respondents who disagreed, strongly expressed their
ENOUGH TO ENDURE POWER OUTAGES, NETWORK own disagreement while about 93 of the remainder
FAILURE, ENVIRONMENTAL DAMAGE. only one of these participants went neutral and indeterminate on the
respondents who disagreed, strongly expressed their matter.
own disagreement while about 88 of the remainder
participants went neutral and indeterminate on the b. Scalability and Security
matter.

Fig 30 a : Our cloud and IoT platform can


Fig 28 : We have taken measures for automated accommodate a large number of users in its full
backups in case the cloud fails capacity

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


380
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 30b
Fig 32

In accordance with the computed data computed in


In accordance with the computed data computed in
Fig.30a and 30b, the number of people who
Fig.31 and 32, the number of people who responded
responded to the notion stating that as CIOs our
to the notion stating that as CIOs our cloud and IoT
cloud and IoT platform can accommodate a large
platform does not collapse or slow down while being
number of users in its full capacity, was 405. About
used by multiple users, was 405. About 98 of these
115 of these respondents representing 28.4 percent of
respondents representing 24.2 percent of the total
the total respondents strongly agreed with this notion,
respondents strongly agreed with this notion, while
while 211 respondents accounting for a whopping
187 respondents accounting for a whopping 46.2
52.1 percent of the total respondents did agree but
percent of the total respondents did agree but not
not strongly with this notion. However a few number
strongly with this notion. However a few number of
of respondents, not more than eleven in number did
respondents, not more than twenty-one in number
disagree with the notion with one more respondents
did disagree with the notion with four more
strongly disagreeing, with about 67 respondents
respondents strongly disagreeing, with about 95
representing 16.5 percent of the total respondents
respondents representing 23.5 percent of the total
were neutral as concerns this notion
respondents were neutral as concerns this notion

Fig 31: Our cloud and IoT platform does not collapse
Fig 33: We keep making improvements to the IoT
or slow down while being used by multiple users
and cloud platform to ensure higher scalability

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


381
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 34 Fig 36

In accordance with the computed data computed in In accordance with the computed data computed in
Fig.33 and 34, the number of people who responded Fig.35 and 36, the number of people who responded
to the notion stating that as CIOs We keep making to the notion stating that as CIOs our systems can
improvements to the IoT and cloud platform to handle multiple application requests arising at the
ensure higher scalability, was 405. About 112 of these same time, was 405. About 104 of these respondents
respondents representing 27.7 percent of the total representing 25.7 percent of the total respondents
respondents strongly agreed with this notion, while strongly agreed with this notion, while 213
203 respondents accounting for a whopping 50.1 respondents accounting for a whopping 52.6 percent
percent of the total respondents did agree but not of the total respondents did agree but not strongly
strongly with this notion. However, a few number of with this notion. However, a few number of
respondents, not more than eleven in number did respondents, not more than eleven in number did
disagree with the notion, with about 79 respondents disagree with the notion, with only two more
representing 19.5 percent of the total respondents respondent strongly disagreeing. About 75
were neutral as concerns this notion respondents representing 17.5 percent of the total
respondents were neutral as concerns this notion

Fig 35: Our systems can handle multiple application


requests arising at the same time
Fig 37: Our cloud and IoT platforms have been
designed to match up with the changing business
objectives

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


382
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 40
Fig 38 In accordance with the computed data computed in
Fig.39 and 40, the number of people who responded
In accordance with the computed data computed in to the notion stating that as CIOs We have adequate
Fig.37 and 38, the number of people who responded safeguards for different interconnected devices to
to the notion stating that as CIOs Our cloud and IoT avoid data theft or breach of privacy, was 405. About
platforms have been designed to match up with the 139 of these respondents representing 34.3 percent of
changing business objectives, was 405. About 102 of the total respondents strongly agreed with this notion,
these respondents representing 25.2 percent of the while 182 respondents accounting for a whopping
total respondents strongly agreed with this notion, 44.9 percent of the total respondents did agree but
while 202 respondents accounting for a whopping not strongly with this notion. However a few number
49.9 percent of the total respondents did agree but of respondents, not more than 9 in number did
not strongly with this notion. However, a few disagree with the notion, with only three more
number of respondents, not more than thirteen in respondent strongly disagreeing. About 72
number did disagree with the notion, with only four respondents representing 17.8 percent of the total
more respondent strongly disagreeing. About 84 respondents were neutral as concerns this notion
respondents representing 20.7 percent of the total
respondents were neutral as concerns this notion

Fig 41: We have adequate safeguards for different


Fig 39: We give utmost importance to network
interconnected devices to avoid data theft or breach
security and data security in our cloud and IoT
of privacy
platforms

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


383
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 42a Fig 43

In accord to the data computed in Fig.41 and 42a, the In accord to the data computed in Fig.42b and 43, the
number of people who responded to the notion number of people who responded to the notion
stating that as CIOs We have adequate safeguards for stating that as CIOs We constantly monitor our cloud
different interconnected devices to avoid data theft and IoT platforms to identify potential security
or breach of privacy, was 405. About 116 of these threats, was 405. About 121 of these respondents
respondents representing 28.6 percent of the total representing 29.9 percent of the total respondents
respondents strongly agreed with this notion, while strongly agreed with this notion, while 185
176 respondents accounting for a whopping 43.5 respondents accounting for a whopping 45.7 percent
percent of the total respondents did agree but not of the total respondents did agree but not strongly
strongly with this notion. However, a few number of with this notion. However, a few number of
respondents, not more than 10 in number did respondents, not more than 16 in number did
disagree with the notion, with only three more disagree with the notion, with only three more
respondent strongly disagreeing. About 100 respondent strongly disagreeing. About 80
respondents representing 24.7 percent of the total respondents representing 22.0 percent of the total
respondents were neutral as concerns this notion respondents were neutral as concerns this notion.

Fig 44 : We regularly update and improve our data


safety mechanisms to counter novel means of breach
Fig 42b : We constantly monitor our cloud and IoT to network security.
platforms to identify potential security threats

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


384
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 47
Fig 45
In accord to the data computed in Fig.46 and 47, the
In accord to the data computed in Fig.44 and 45, the number of people who responded to the notion
number of people who responded to the notion stating that as CIOs Our cloud and IoT platforms are
stating that as CIOs We regularly update and improve designed to make optimal use of existing resources,
our data safety mechanisms to counter novel means was 405. About 116 of these respondents representing
of breach to network security, was 405. About 114 of 28.6 percent of the total respondents strongly agreed
these respondents representing 28.1 percent of the with this notion, while 202 respondents accounting
total respondents strongly agreed with this notion, for a whopping 49.9 percent of the total respondents
while 190 respondents accounting for a whopping did agree but not strongly with this notion. However,
46.9 percent of the total respondents did agree but a few number of respondents, not more than 7 in
not strongly with this notion. However, a few number did disagree with the notion, with only one
number of respondents, not more than ten in number more respondent strongly disagreeing. About 79
did disagree with the notion, with only two more respondents representing 19.5 percent of the total
respondent strongly disagreeing. About 89 respondents were neutral as concerns this notion.
respondents representing 22.0 percent of the total
respondents were neutral as concerns this notion.

c. Capacity and Throughput

Fig 48: We constantly track the resources used up by


our cloud and IoT products and take steps to
minimize resource loss
Fig 46: Our cloud and IoT platforms are designed to
make optimal use of existing resources

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


385
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 4 Fig 51

In accord to the data computed in Fig.48 and 49, the In accord to the data computed in Fig.50 and 51, the
number of people who responded to the notion number of people who responded to the notion
stating that as CIOs We constantly track the stating that as CIOs We undertake capacity planning
resources used up by our cloud and IoT products and much before the implementation of cloud and IoT
take steps to minimize resource loss, was 405. About based technologies, was 405. About 105 of these
104 of these respondents representing 25.7 percent of respondents representing 25.9 percent of the total
the total respondents strongly agreed with this notion, respondents strongly agreed with this notion, while
while 203 respondents accounting for a whopping 203 respondents accounting for a whopping 50.1
50.1 percent of the total respondents did agree but percent of the total respondents did agree but not
not strongly with this notion. However, a few strongly with this notion. However, a few number of
number of respondents, not more than 12 in number respondents, not more than 10 in number did
did disagree with the notion, with only four more disagree with the notion, with only a single more
respondent strongly disagreeing. About 82 respondent strongly disagreeing. About 86
respondents representing 20.2 percent of the total respondents representing 21.2 percent of the total
respondents were neutral as concerns this notion. respondents were neutral as concerns this notion.

Fig 50: We undertake capacity planning much before Fig 52: We are constantly innovating means to reduce
the implementation of cloud and IoT based resource use and improve productivity
technologies

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


386
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 55

Fig 53 In accord to the data computed in Fig.54 and 55, the


number of people who responded to the notion
In accord to the data computed in Fig.52 and 53, the stating that as CIOs We try to eliminate errors, in
number of people who responded to the notion order to avoid resource loss, was 405. About 113 of
stating that as CIOs We try to eliminate errors, in these respondents representing 27.9 percent of the
order to avoid resource loss, was 405. About 118 of total respondents strongly agreed with this notion,
these respondents representing 29.1 percent of the while 187 respondents accounting for a whopping
total respondents strongly agreed with this notion, 46.2 percent of the total respondents did agree but
while 189 respondents accounting for a whopping not strongly with this notion. However, a few
46.7 percent of the total respondents did agree but number of respondents, not more than 14 in number
not strongly with this notion. However, a few did disagree with the notion, with only three more
number of respondents, not more than 14 in number respondent strongly disagreeing. About 88
did disagree with the notion, with only a single more respondents representing 21.7 percent of the total
respondent strongly disagreeing. About 83 respondents were neutral as concerns this notion.
respondents representing 20.5 percent of the total
respondents were neutral as concerns this notion.

Fig 56: We allocate only required resources to cloud


or IoT platforms to avoid losses
Fig 54: We try to eliminate errors, in order to avoid
resource loss

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


387
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 57 Fig 59

In congruence to the data computed in Fig.56 and 57, In congruence to the data computed in Fig.58 and 59,
the number of people who responded to the notion the number of people who responded to the notion
stating that as CIOs We allocate only required stating that as CIOs our cloud and IoT based
resources to cloud or IoT platforms to avoid losses, applications can process maximum tasks in a given
was 405. About 99 of these respondents representing unit of time, was 405. About 105 of these respondents
24.4 percent of the total respondents strongly agreed representing 25.9 percent of the total respondents
with this notion, while 205 respondents accounting strongly agreed with this notion, while 192
for a whopping 50.6 percent of the total respondents respondents accounting for a whopping 47.4 percent
did agree but not strongly with this notion. However, of the total respondents did agree but not strongly
a few number of respondents, not more than 13 in with this notion. However, a few number of
number did disagree with the notion, with only one respondents, not more than 13 in number did
more respondent strongly disagreeing. About 87 disagree with the notion, with only two more
respondents representing 21.5 percent of the total respondent strongly disagreeing. About 93
respondents were neutral as concerns this notion. respondents representing 23.0 percent of the total
respondents were neutral as concerns this notion.

Fig 58: Our cloud and IoT based applications can


process maximum tasks in a given unit of time Fig 60: We make sure we respond to the requests of
every user within minimum possible time

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


388
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 61 Fig 63

In congruence to the data computed in Fig.60 and 61, In congruence to the data computed in Fig.62 and 63,
the number of people who responded to the notion the number of people who responded to the notion
stating that as CIOs We make sure we respond to the stating that as CIOs We do our best to achieve
requests of every user within minimum possible time, maximum transactions per second, was 405. About
was 405. About 107 of these respondents representing 112 of these respondents representing 27.7 percent of
26.4 percent of the total respondents strongly agreed the total respondents strongly agreed with this notion,
with this notion, while 199 respondents accounting while 193 respondents accounting for a whopping
for a whopping 49.1 percent of the total respondents 47.7 percent of the total respondents did agree but
did agree but not strongly with this notion. However, not strongly with this notion. However, a few
a few number of respondents, not more than 11 in number of respondents, not more than 11 in number
number did disagree with the notion, with only five did disagree with the notion, with only two more
more respondent strongly disagreeing. About 83 respondent strongly disagreeing. About 87
respondents representing 20.5 percent of the total respondents representing 21.5 percent of the total
respondents were neutral as concerns this notion. respondents were neutral as concerns this notion.

Fig 62: We do our best to achieve maximum


transactions per second
Fig 64: Our throughput has been consistently
improving over the years

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


389
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 64b
Fig 66
In congruence to the data computed in Fig.64 and
64b, the number of people who responded to the In congruence to the data computed in Fig.65 and 66,
notion stating that as CIOs our throughput has been the number of people who responded to the notion
consistently improving over the years, was 405. stating that as CIOs We are constantly on the lookout
About 107 of these respondents representing 26.4 for new mechanisms to improve our throughput, was
percent of the total respondents strongly agreed with 405. About 109 of these respondents representing
this notion, while 197 respondents accounting for a 26.9 percent of the total respondents strongly agreed
whopping 48.6 percent of the total respondents did with this notion, while 198 respondents accounting
agree but not strongly with this notion. However, a for a whopping 48.9 percent of the total respondents
few number of respondents, not more than 13 in did agree but not strongly with this notion. However,
number did disagree with the notion, with only three a few number of respondents, not more than 12 in
more respondent strongly disagreeing. About 85 number did disagree with the notion, with only a
respondents representing 21.0 percent of the total single more respondent strongly disagreeing. About
respondents were neutral as concerns this notion. 85 respondents representing 21.0 percent of the total
respondents were neutral as concerns this notion.

d. Lower Costs

Fig 65: We are constantly on the lookout for new


mechanisms to improve our throughput

Fig 67: Our cloud and IoT platforms offer integrated


data storage solutions, hereby helping reduce data
storage costs

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


390
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 68 Fig 70

In congruence to the data computed in Fig.67 and 68, In congruence to the data computed in Fig.69 and 70,
the number of people who responded to the notion the number of people who responded to the notion
stating that as CIOs our cloud and IoT platforms offer stating that As cloud based programs can be accessed
integrated data storage solutions, hereby helping anywhere, it saves logistics and data transfer costs,
reduce data storage costs, was 405. About 99 of these was 405. About 109 of these respondents representing
respondents representing 24.4 percent of the total 26.9 percent of the total respondents strongly agreed
respondents strongly agreed with this notion, while with this notion, while 195 respondents accounting
204 respondents accounting for a whopping 50.4 for a whopping 48.1 percent of the total respondents
percent of the total respondents did agree but not did agree but not strongly with this notion. However,
strongly with this notion. However, a few number of a few number of respondents, not more than 12 in
respondents, not more than 14 in number did number did disagree with the notion, with about 5
disagree with the notion, with about 3 more more respondent strongly disagreeing. About 84
respondent strongly disagreeing. About 85 respondents representing 20.7 percent of the total
respondents representing 21.0 percent of the total respondents were neutral as concerns this notion.
respondents were neutral as concerns this notion.

Fig 71: As our cloud and IoT technologies can be


Fig 69: As cloud based programs can be accessed expanded for large number of users, the cost of
anywhere, it saves logistics and data transfer costs additional infrastructure is saved

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


391
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

Fig 72 Fig 74

In congruence to the data computed in Fig.71 and 72, In congruence to the data computed in Fig.73 and
the number of people who responded to the notion 764, the number of people who responded to the
stating As our cloud and IoT technologies can be notion stating that Cloud and IoT based technologies
expanded for large number of users, the cost of in reduction of server maintenance, power and other
additional infrastructure is saved, was 405. About 115 costs, was 405. About 106 of these respondents
of these respondents representing 28.4 percent of the representing 26.2 percent of the total respondents
total respondents strongly agreed with this notion, strongly agreed with this notion, while 189
while 191 respondents accounting for a whopping respondents accounting for a whopping 46.7 percent
47.2 percent of the total respondents did agree but of the total respondents did agree but not strongly
not strongly with this notion. However, a few with this notion. However, a few number of
number of respondents, not more than 19 in number respondents, not more than 16 in number did
did disagree with the notion, with about 4 more disagree with the notion, with about 6 more
respondent strongly disagreeing. About 76 respondent strongly disagreeing. About 88
respondents representing 18.8 percent of the total respondents representing 21.7 percent of the total
respondents were neutral as concerns this notion. respondents were neutral as concerns this notion.

Fig 73: Cloud and IoT based technologies in reduction Fig 75: Cloud based data backup saves the cost of
of server maintenance, power and other costs retrieving lost data during system crashes or natural
calamities

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


392
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

confront. Two similarity issues were perceived –


similarity with existing framework, what's more, a
similarity with other Cloud framework. The
examination set up excess as a factor. The discoveries
proposed that any expenses of the current foundation
must be considered before an official choice to
receive Cloud processing can be made in light of the
fact that some of the time utilizing the current
foundation could be the more productive choice.

The results shows that the new role of the CIO has
resulted in successful implementation of cloud and
IoT based technologies. The study also states that the
Fig 76
most of the companies cloud and IoT platforms has
less chance of failure and they are set up in such a
In congruence to the data computed in Fig.75 and 76,
way that they have adequate methods for recovery.
the number of people who responded to the notion
To have the best availability and reliability most of
stating that Cloud based data backup saves the cost of
the cloud infrastructures are using effective tools to
retrieving lost data during system crashes or natural
address the issues which can give higher confidence
calamities, was 405. About 117 of these respondents
rate to the cloud adopters. The confirmation from IT
representing 28.9 percent of the total respondents
and business leaders states that they have taken the
strongly agreed with this notion, while 188
necessary measure to backup and store data to be
respondents accounting for a whopping 46.4 percent
easily retrieved during emergencies. The results
of the total respondents did agree but not strongly
shows that the availability of the cloud is high with
with this notion. However, a few number of
minimal downtime and server failures. The
respondents, not more than 14 in number did
infrastructure is designed with enough redundancy so
disagree with the notion and about 83 respondents
that resiliency is achieved in case of failures. The
representing 20.5 percent of the total respondents
automated backup mechanisms are helping the
were neutral as concerns this notion.
organizations from any data loss.

V. CONCLUSION AND RECOMMENDATIONS


The analysis on the scalability and security aspects
states that the cloud and IoT platforms are capable to
Cloud services could give a business relative favorable scale and it is built with the required capacity so that
position over its rivals and cost advantage, through the performance doesn't get affected during peak
the advantages they give i.e. enhanced business operation period. The organizations are spending
agility and diminished capital use. The multifaceted considerable effort to bring in improvement to the
nature of the Cloud was observed to be a noteworthy cloud and IoT platforms to improve the scalability.
challenge in the adoption procedure, however, it This can also give an idea that there is huge amount
ought to be noticed that Cloud arrangements are of research or innovations are going on in this area.
getting less mind boggling with the improvement of The organizations are trying to explore options to
the innovation. Similarity was distinguished as improve the scalability by innovative methods. The
another test that a business is probably going to platforms are designed with proper business goals so

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


393
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

that any change in the business environment can be Moreover, the result analysis unveiled that the
effectively addressed by the cloud infrastructures. overall experience and/or exposure, and the industry
This also gives a learning to the companies sector of one’s present organization plus his/her
implementing and adopting cloud to make sure that background education status goes a long way in the
they are doing the required business forecasting and successful implementation of Cloud-IoT based
planning for designing the cloud and IoT technologies. Also, it became obviously necessary
infrastructures. A poor planning in this area can that cloud-based data back-up saves the cost of
negatively affect the future business of any retrieving lost data during system crashes or any
organization. So the organizations has to give utmost natural calamity, as well as helps in the reduction of
importance to match up with the environmental server maintenance, power and other costs. It is
changes in the business. The organizations are important that as cloud & IoT technologies are being
concerned about the security of their infrastructures expanded to be accessible by large number of users,
and they are doing the best possible measures to the cost of all additional infrastructures is being
safeguard their infrastructure. The organizations are saved. The results also reveals that as cloud-based
constantly monitoring their cloud and IoT platforms programs are being accessed anywhere, it saves
to identify any potential security threats that would logistics and data transfer costs much more
impact their business. From this study it gives a importantly. It proved that cloud & IoT platforms
reference that there is huge scope for research in offer integrated data storage solutions, thereby
scalability and security aspects of the cloud and IoT. helping to reduce data storage costs in the
Assessing the scalability and security of cloud-IoT atmosphere of its successful implementation.
technologies, including its capacity, throughput and
additional lower costs, it is concluded that the In the same vein, as business leaders willing to
targeted successful implementation of cloud-IoT successfully implement and integrate cloud
based technologies will be very much easier to technologies, it is necessary to achieve maximum
accomplish by improving cloud throughputs transaction per seconds and respond to the requests
frequently by looking out for newer operating of each user within the shortest possible timeframe,
mechanisms. thus enhancing good customer satisfaction and
service. It is also quite important to allocate only the
The analysis of the capacity and throughput required resources to cloud-IoT based platforms to
parameters states that the organizations are showing avoid critical losses. As business and IT leaders, it
more interest in optimizing the existing resources. becomes essential to constantly innovate, thus
The organizations implementing the cloud and IoT improving productivity while minimising the use of
has to prepare a roadmap of the capacity well in scarce resources and eliminating unnecessary errors.
advance before the implementation of cloud and IoT. It is also essential to indulge in capacity planning and
A proper planning would help the organizations to tracking of the available resources. Furthermore, it
optimise the resources when the volume of will do the company good to update data-safety
transactions increases in the infrastructure. An mechanisms frequently, and checking any breach in
architects and designers also need to consider the network security or security threats, as a way of
chances of error rates and do the required resource enhancing optimal usage of resources, i.e, network
planning to avoid any losses. The results of this study security and data security. In addition, cloud
states that there is huge scope for research in the area technologies can multi-task in terms of application
of digital optimization. requests without slowing down, and it is built to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


394
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

accommodate changing business objectives [3]. Pocatilu, P., Alecu, F., &Vetrici, M. (2010).
(flexibility). Measuring the efficiency of cloud computing
for e-learning systems. Wseas transactions on
The cloud-IoT platforms should be able to withstand
computers, 9(1), 42-51.
server failures, endure power outages, network
[4]. Deka Ganesh Chandra & Dutta Borah Malaya,
failures and any form of environmental damage one
(2012), Role of Cloud Computing in Education,
again. These and many more will surely enhance the
ICCEET, Pg: 832-836.
reliability, availability, scalability, security, capacity,
[5]. Dong, B., Zheng, Q., Qiao, M., Shu, J., & Yang,
throughput and cut operating costs of cloud-Iot based
J. (2009). BlueSky cloud framework: an e-
technologies in the quest of its successful
learning framework embracing cloud
implementation and integration. The execution was
computing. In Cloud Computing (pp. 577-582).
another issue that was recognized as a factor. The
Springer Berlin Heidelberg.
discoveries recommended it needs thought in light of
[6]. Faisal A. Alshuwaier, Abdullah A. Alshwaier,
the fact that receiving Cloud services could influence
Ali M. Areshey (2011), Applications of Cloud
the system execution and accessibility of the services
Computing in Education.
is outside of the business control. Likewise, security
[7]. Feng, D. G., Zhang, M., Zhang, Y., & Xu, Z.
was distinguished as the best mechanical test. The
(2011). Study on cloud computing security.
discoveries contradicted that by recommending that
Journal of Software, 22(1), 71-83.
substantial organizations have their very own
[8]. Fox, A., Griffith, R., Joseph, A., Katz, R.,
foundation that needs thought and it would be less
Konwinski, A., Lee, G., ... &Stoica, I. (2009).
demanding for little organizations to receive cloud
Above the clouds: A Berkeley view of cloud
services as a result of the moderately little set-up
computing. Dept. Electrical Eng. and Comput.
expense. Considering the constraints to the
Sciences, University of California, Berkeley,
exploration of this examination a proposal for further
Rep. UCB/EECS, 28, 13.
research can be made. With the end goal to address
[9]. Hailu, A. (2012). Factors influencing cloud-
these issues, a further research could be made
computing technology adoption in developing
utilizing a quantitative technique that could look at a
countries (Doctoral dissertation, Capella
bigger example of the populace.
University).
[10]. Chen, Y., & Kunz, T. (2016, April).
VI. REFERENCES
Performance evaluation of IoT protocols under
a constrained wireless access network. In
[1]. Iosup, A., Ostermann, S., Yigitbasi, M. N.,
Selected Topics in Mobile & Wireless
Prodan, R., Fahringer, T., &Epema, D. (2011).
Networking (MoWNeT), 2016 International
Performance analysis of cloud computing
Conference on (pp. 1-7). IEEE.
services for many-tasks scientific computing.
[11]. Collinchau. (2013, August 19). 7 performance
IEEE Transactions on Parallel and Distributed
metrics for measuring cloud ROI within HP
systems, 22(6), 931-945.
Cloud Management. Micro Focus. Retrieved 19
[2]. The cloud service industry’s 10 most critical
October 2018 from
metrics. Guiding Metrics. Retrieved 19
https://community.softwaregrp.com/t5/IT-
February 2018 from
Operations-Management-ITOM/7-
https://guidingmetrics.com/content/cloud-
performance-metrics-for-measuring-cloud-
services-industrys-10-most-critical-metrics

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


395
Jipson George Thoomkuzhy, Dr. Mohammed Nazeh. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 366-396

ROI-within-HP-Cloud/ba-
p/300699#.Wopi1INua01
[12]. Frey, S., Reich, C., &Lüthje, C. (2013,
September). Key performance indicators for
cloud computing SLAs. In The Fifth
International Conference on Emerging
Network Intelligence, EMERGING (pp. 60-64).

Cite this article as :

Jipson George Thoomkuzhy, Dr. Mohammed Nazeh,


"Factors Affecting the Successful Implementation of
Cloud and IoT - A Study on IT and Business Leaders ",
International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 366-396, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT195185
Journal URL : http://ijsrcseit.com/CSEIT195185

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


396
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195177
Development of Remote Operated Agri Vehicle
Juno Isabel Susinthra M *1, K. Gayadri 2
*1Assistant Professor, Department of Computer Science, Bon Secours College for Women, Bharathidasan
University, Thanjavur, Tamil Nadu, India
2 M.Sc, Department of Computer Science, Bon Secours College for Women, Bharathidasan University,
Thanjavur, Tamil Nadu, India

ABSTRACT
An agricultural mechanism or agri mechanism could be a mechanism deployed for agricultural functions. The
most space of application of robots in agriculture is seeding robots square measure designed to exchange human
labour. The agricultural business is behind alternative complementary industries in victimization robots as a
result of the kind of jobs concerned in agriculture are not any and dig. This project strives to develop a
mechanism capable of activity operations like automatic seeding and dig. It additionally provides manual
management once needed and keeps tabs on the wetness with the assistance of wetness sensors. Then main part
here is that the microcontroller that supervises the whole method.
Keywords : Agri Vehicle, GPS based Remote Controlled Robot, Zigbee, Raspberry Pi, Global Positioning
System, UART, LM35
I. INTRODUCTION environmental factors which in turns helps to
monitor the system. Monitoring environmental
Agriculture is considered as the basis of life for the factors is not enough and complete solution to
human species as it is the main source of food grains improve the yield of the crops. There are number of
and other raw materials. It plays vital role in the other factors that affect the productivity to great
growth of country’s economy. It also provides large extent. These factors include attack of insects and
ample employment opportunities to the people. pests which can be controlled by spraying the crop
Growth in agricultural sector is necessary for the with proper insecticide and pesticides. Secondly,
development of economic condition of the country. attack of wild animals and birds when the crop grows
Unfortunately, many farmers still use the traditional up. There is also possibility of thefts when crop is at
methods of farming which results in low yielding of the stage of harvesting. Even after harvesting, farmers
crops and fruits. But wherever automation had been also face problems in storage of harvested crop. So, in
implemented and human beings had been replaced order to provide solutions to all such problems, it is
by automatic machineries, the yield has been necessary to develop integrated system which will
improved. Hence there is need to implement modern take care of all factors affecting the productivity in
science and technology in the agriculture sector for every stages like; cultivation, harvesting and post
increasing the yield. Most of the papers signifies the harvesting storage. This paper therefore proposes a
use of wireless sensor network which collects the system which is useful in monitoring the field data as
data from different types of sensors and then send it well as controlling the field operations which
to main server using wireless protocol. The collected provides the flexibility. The paper aims at making
data provides the information about different agriculture smart using automation and IoT

CSEIT195177 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 397-403 ]
397
Juno Isabel Susinthra M, K. Gayadri Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 397-403

technologies. The highlighting features of this paper sensing and control irrigation system using
includes smart GPS based remote controlled robot to distributed wireless sensor network aiming for
perform tasks like; weeding, spraying, moisture variable rate irrigation, real time in field sensing,
sensing, bird and animal scaring, keeping vigilance, controlling of a site-specific precision linear move
etc. Secondly, it includes smart irrigation with smart irrigation system to maximize the productivity with
control based on real time field data. Thirdly, smart minimal use of water was developed by Y. Kim. The
warehouse management which includes temperature system described details about the design and
maintenance, humidity maintenance and theft instrumentation of variable rate irrigation, wireless
detection in the warehouse. Controlling of all these sensor network and real time in field sensing and
operations will be through any remote smart device control by using appropriate software. The whole
or computer connected to Internet and the operations system was developed using five in field sensor
will be performed by interfacing sensors, Wi-Fi or stations which collects the data and send it to the
ZigBee modules, camera and actuators with micro- base station using global positioning system (GPS)
controller and raspberry pi. where necessary action was taken for controlling
irrigation according to the database available with the
II. METHODS AND MATERIAL system. The system provides a promising low cost
wireless solution as well as remote controlling for
The newer scenario of decreasing water tables, drying precision irrigation. In the studies related to wireless
up of rivers and tanks, unpredictable environment sensor network, researchers measured soil related
present an urgent need of proper utilization of water. parameters such as temperature and humidity.
To cope up with this use of temperature and moisture Sensors were placed below the soil which
sensor at suitable locations for monitoring of crops is communicates with relay nodes by the use of
implemented in. An algorithm developed with effective communication protocol providing very low
threshold values of temperature and soil moisture can duty cycle and hence increasing the life time of soil
be programmed into a microcontroller-based gateway monitoring system. The system was developed using
to control water quantity. The system can be microcontroller, universal asynchronous receiver
powered by photovoltaic panels and can have a transmitter (UART) interface and sensors while the
duplex communication link based on a cellular- transmission was done by hourly sampling and
Internet interface that allows data inspection and buffering the data, transmit it and then checking the
irrigation scheduling to be programmed through a status messages. The drawbacks of the system were its
web page. The technological development in cost and deployment of sensor under the soil which
Wireless Sensor Networks made it possible to use in causes attenuation of radio frequency (RF) signals.
monitoring and control of greenhouse parameter in
precision agriculture. After the research in the III. RESULTS AND DISCUSSION
agricultural field, researchers found that the yield of
agriculture is decreasing day by day. However, use of
technology in the field of agriculture plays important
role in increasing the production as well as in
reducing the extra man power efforts. Some of the
research attempts are done for betterment of farmers
which provides the systems that use technologies Fig 1. System Overview
helpful for increasing the agricultural yield. A remote

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


398
Juno Isabel Susinthra M, K. Gayadri Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 397-403

In the present system, every node is integration with with AVR microcontroller. Motion detector will
different sensors and devices and they are detect the motion in the room when security mode
interconnected to one central server via wireless will be ON and on detection of motion, it will send
communication modules. The server sends and the alert signal to user via Raspberry pi and thus
receives information from user end using internet providing theft detection.
connectivity. There are two modes of operation of
the system; auto mode and manual mode. In auto
mode system takes its own decisions and controls the
installed devices whereas in manual mode user can
control the operations of system using android app or
PC commands.

IV. ARCHITECTURE OF THE SYSTEM

Node 1:
Node1 is GPS based mobile robot which can be
controlled remotely using computer as well as it can
be programmed so as to navigate autonomously Fig 3.
within the boundary of field using the Co-ordinates Node 2
given by GPS module. Temperature sensor and Humidity sensor senses the
temperature and humidity respectively and if the
value crosses the threshold then room heater or
cooling fan will be switched ON/OFF automatically
providing temperature and humidity maintenance.
Node2 will also controls water pump depending upon
the soil moisture data sent by node3.
Node 3:
Node3 is a smart irrigation node with features like;
Smart control of water pump based on real time field
data i.e. automatically turning on/off the pump after
Fig 2. attaining the required soil moisture level in auto
Node 1 mode, switching water pump on/off remotely via
The Remote controlled robot have various sensors mobile or computer in manual mode, and continuous
and devices like camera, obstacle sensor, siren, cutter, monitoring of soil moisture.
sprayer and using them it will perform tasks like; Hardware used:
Keeping vigilance, Bird and animal scaring, Weeding, a) AVR Microcontroller At mega 16/32:
and Spraying The microcontroller used is, Low-power AVR® 8-bit
Node 2: Microcontroller, having 8K Bytes of In-System Self-
Node2 will be the warehouse. It consists of motion programmable Flash program memory,
detector, light sensor, humidity sensor, temperature Programmable Serial USART, 8-channel, 10-bit ADC,
sensor, room heater, cooling fan altogether interfaced 23 Programmable I/O Lines.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


399
Juno Isabel Susinthra M, K. Gayadri Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 397-403

b) ZigBee Module: The ultra-sonic sensor operates on the principle of


ZigBee is used for achieving wireless communication sound waves and their reflection property. It has two
between Node1 and Node2. The range for Zigbee is parts; ultra-sonic transmitter and ultra-sonic receiver.
roughly 50 meters and it can be increased using high Transmitter transmits the 40 KHz sound wave and
power modules or by using network of modules. It receiver receives the reflected 40 KHz wave and on
operates on 2.4 GHz frequency. Its power its reception, it sends the electrical signal to the
consumption is very low and it is less expensive as microcontroller. The speed of sound in air is already
compared to other wireless modules like Wi-Fi or known. Hence from time required to receive back the
Bluetooth. It is usually used to establish wireless local transmitted sound wave, the distance of obstacle is
area networks. calculated. Here, it is used for obstacle detection in
c) Temperature Sensor LM35: case of mobile robot and as a motion detector in ware
The LM35 is precision IC temperature sensor. Output house for preventing thefts. The ultra-sonic sensor
voltage of LM35 is directly proportional to the enables the robot to detect and avoid obstacles and
Centigrade/Celsius of temperature. The LM35 does also to measure the distance from the obstacle. The
not need external calibration or trimming to provide range of operation of ultra-sonic sensor is 10 cm to 30
accurate temperature range. It is very low cost sensor. cm.
It has low output impedance and linear output. The f) Raspberry Pi:
operating temperature range for LM35 is −55˚ to The Raspberry Pi is small pocket size computer used
+150˚C. With rise in temperature, the output voltage to do small computing and networking operations. It
of the sensor increases linearly and the value of is the main element in the field of internet of things.
voltage is given to the microcontroller which is It provides access to the internet and hence the
multiplied by the conversion factor in order to give connection of automation system with remote
the value of actual temperature. location controlling device becomes possible.
d) Moisture sensor: Raspberry Pi is available in various versions. Here,
Soil moisture sensor measures the water content in model Pi 2 model B is used and it has quad-core ARM
soil. It uses the property of the electrical resistance of Cortex-A53 CPU of 900 MHz, and RAM of 1GB. it
the soil. The relationship among the measured also has: 40 GPIO pins, Full HDMI port, 4 USB ports,
property and soil moisture is calibrated and it may Ethernet port, 3.5mm audio jack, video Camera
vary depending on environmental factors such as interface (CSI), the Display interface (DSI), and
temperature, soil type, or electric conductivity. Here, Micro SD card slot.
it is used to sense the moisture in field and transfer it Software used:
to microcontroller in order to take controlling action a) AVR Studio Version 4:
of switching water pump ON/OFF. Humidity sensor: It is used to write, build, compile and debug the
The DHT11 is a basic, low-cost digital temperature embedded c program codes which are needed to b
and humidity sensor. It gives out digital value and
hence there is no need to use conversion algorithm at
ADC of the microcontroller and hence we can give
its output directly to data pin instead of ADC. It has a
capacitive sensor for measuring humidity. The only
real shortcoming of this sensor is that one can only
get new data from it only after every 2 seconds.
e) Obstacle sensor (Ultra-Sonic):

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


400
Juno Isabel Susinthra M, K. Gayadri Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 397-403

d) SinaProg:
SinaProg is a Hex downloader application with AVR
Dude and Fuse Bit Calculator. This is used to
download code/program and to set fuse bits of all
AVR based microcontrollers.
e) Raspbian Operating System:
Raspbian operating system is the free and open
source operating system which Debian based and
optimized for Raspberry Pi. It provides the basic set
of programs and utilities for operating Raspberry Pi.
Fig 4. It comes with around 35,000 packages which are pre-
In node3, moisture sensor transmits the data using compiled software that are bundled in a nice format
HT12E Encoder IC and a RF transmitter. The for hustle free installation on Raspberry Pi.It has
transmitted data is received by node2 and there it is good community of developers which runs the
processed by microcontroller in order to control the discussion forms and provides solutions to many
operation of water pump. relevant problems. However, Raspbian OS is still
under consistent development with an main focus on
Burned in the microcontroller in order to perform improving the performance and the stability of as
desired operations. This software directly provides many Debian packages as possible.
.hex file which can be easily burned into the
microcontroller. V. EXPERIMENTATION AND RESULTS

b) Proteus 8 Simulator:
Proteus 8 is one of the best simulation software for
various circuit designs of microcontroller. It has
almost all microcontrollers and electronic
components readily available in it and hence it is
widely used simulator. It can be used to test programs
and embedded designs for electronics before actual
hardware testing. The simulation of programming of
microcontroller can also be done in Proteus.
Simulation avoids the risk of damaging hardware due
to wrong design.
c) Dip Trace: Fig 5. Experimental setup for Node1
Dip race is EDA/CAD software for creating schematic
As shown in figure 5, experimental setup for node1
diagrams and printed circuit boards. The developers
consists of mobile robot with central server, GPS
provide multi-lingual interface and tutorials
module, camera and other sensors. All sensors are
(currently available in English and 21 other
successfully interfaced with microcontroller and the
languages). Dip Trace has 4 modules: Schematic
microcontroller is interfaced with the raspberry
Capture Editor, PCB Layout Editor with built-in
pi.GPS and cameras are also connected to raspberry
shape-based auto router and 3D Preview & Export,
pi. Test results shows that the robot can be controlled
Component Editor, and Pattern Editor.
remotely using wireless transmission of PC

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


401
Juno Isabel Susinthra M, K. Gayadri Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 397-403

commands to R-Pi.R-Pi forwards the commands to As shown in above figure, node3 consists of a
microcontroller and microcontroller gives signals to moisture sensor connected to HT12E. Moisture
motor driver in order to drive the Robot.GPS module sensor transmits the data using HT12E Encoder IC
provides the co-ordinates for the location of the robot and a RF transmitter to the Node2 where it is
processed by microcontroller and accordingly water
pump is switched ON/OFF.

VI. CONCLUSION

The sensors and microcontrollers of all three Nodes


are successfully interfaced with raspberry pi and
wireless communication is achieved between various
Nodes. All observations and experimental tests proves
Fig 6. Experimental setup for Node2 that project is a complete solution to field activities,
irrigation problems, and storage problems using
As shown in above figure, node2 consists of motion remote controlled robot, smart irrigation system and
detector, temperature sensor, humidity senor, cooling a smart warehouse management system respectively.
fan, water pump, etc. connected to the Implementation of such a system in the field can
microcontroller board. The sensors gives input to the definitely help to improve the yield of the crops and
controller and according to that microcontroller overall production.
controls the devices in auto mode and also sends the
value of sensors to R-Pi and R-Pi forwards it to user’s VII. ACKNOWLEDGMENT
smart device using internet.Test results shows that
when temperature level increases above preset I am sincerely thankful to all my teachers for their
threshold level then cooling fan is started guidance for my seminar. Without their help it was
automatically in auto mode. The water pump also tough job for me to accomplish this task. I am
gets turned ON if moisture level goes below fixed especially very thankful to my guide Dr. R.S. Kawaka
threshold value. In manual mode, microcontroller for his consistent guidance, encouragement and
receives the controlling signals from R-Pi through motivation throughout the period of this work. I also
ZigBee and accordingly takes the control action. want to thank our Head of the Department (E&TC)
Dr. M. B. Mali for providing me all necessary
facilities.

VIII. REFERENCES

[1] S. R. Nandurkar, V. R. Thool, R. C. Thool,


“Design and Development of Precision
Agriculture System Using Wireless Sensor
Network”, IEEE International Conference on
Automation, Control, Energy and Systems
(ACES), 2014
Fig 7. Experimental setup for Node3

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


402
Juno Isabel Susinthra M, K. Gayadri Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 397-403

[2] JoaquínGutiérrez, Juan Francisco Villa-Medina, Cite this article as :


Alejandra Nieto-Garibay, and Miguel Ángel
Porta-Gándara, “Automated Irrigation System Juno Isabel Susinthra M, K. Gayadri, "Development
Using a Wireless Sensor Network and GPRS of Remote Operated Agri Vehicle", International
Module”,IEEE TRANSACTIONS ON Journal of Scientific Research in Computer Science,
INSTRUMENTATION AND MEASUREMENT, Engineering and Information Technology
0018-9456,2013 (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
[3] Dr. V .Vidya Devi,G. Meena Kumari, “Real- 397-403, January-February 2019. Available at doi :
Time Automation and Monitoring System for https://doi.org/10.32628/CSEIT195177
Modernized Agriculture” ,International Journal Journal URL : http://ijsrcseit.com/CSEIT195177
of Review and Research in Applied Sciences and
Engineering (IJRRASE) Vol3 No.1. PP 7-12,
2013
[4] Y. Kim, R. Evans and W. Iversen, “Remote
Sensing and Control of an Irrigation System
Using a Distributed Wireless Sensor Network”,
IEEE Transactions on Instrumentation and
Measurement, pp. 1379–1387, 2008.
[5] Q. Wang, A. Terzis and A. Szalay, “A Novel Soil
Measuring Wireless Sensor Network”, IEEE
Transactions on Instrumentation and
Measurement, pp. 412–415, 2010
[6] Yoo, S.; Kim, J.; Kim, T.; Ahn, S.; Sung, J.; Kim,
D. A2S: Automated agriculture system based on
WSN. In ISCE 2007. IEEE International
Symposium on Consumer Electronics, 2007,
Irving,TX, USA, 2007
[7] Arampatzis, T.; Lygeros, J.; Manesis, S. A survey
of applications of wireless sensors and Wireless
Sensor Networks. In 2005 IEEE International
Symposium on Intelligent Control & 13th
Mediterranean Conference on Control and
Automation. Limassol, Cyprus, 2005, 1-2, 719-
724

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


403
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195178
Adblock Usage in Web Advertisement
V. Santhi *1, S. Abirami 2
*1Assistant Professor, Department of Computer Applications, Bon Secours College for Women, Bharathidasan
University, Thanjavur, Tamil Nadu, India
2 M.Sc, Department of Computer Science, Bon Secours College for Women, Bharathidasan University,
Thanjavur, Tamil Nadu, India
ABSTRACT

Web users are increasingly turning to ad blockers to avoid ads, which are often perceived as annoying or an
invasion of privacy. While there has been significant research into the factors driving ad blocker adoption and
the detrimental effect to ad publishers on the Web, the resulting effects of ad blocker usage on Web users’
browsing experience is not well understood. To approach this problem, we conduct a retrospective natural field
experiment using Firefox browser usage data, with the goal of estimating the effect of ad blocking on user
engagement with the Web. We focus on new users who installed an ad blocker after a baseline observation
period, to avoid comparing different populations. Their subsequent browser activity is compared against that of
a control group, whose members do not use ad blockers, over a corresponding observation period, controlling
for prior baseline usage. In order to estimate causal effects, we employ propensity score matching on a number
of other features recorded during the baseline period. In the group that installed an ad blocker, we find
significant increases in both active time spent in the browser (+28% over control) and the number of pages
viewed (+15% over control), while seeing no change in the number of searches. Additionally, by reapplying the
same methodology to other popular Firefox browser extensions, we show that these effects are specific to ad
blockers. We conclude that ad blocking has a positive impact on user engagement with the Web, suggesting
that any costs of using ad blockers to users’ browsing experience are largely drowned out by the utility that
they offer.
Keywords: AdBlockPlus, Unblock Origin, Trivial Algorithm.

I. INTRODUCTION turn leads to reduced effectiveness from the point of


view of ad publishers. As a result, Web users are
For an average user, a typical day on the Web increasingly turning to ad blocking to mitigate the
involves exposure to ads. Indeed, advertising has negative effects of online ads. A recent study
become the primary revenue model for many popular estimates that over 600M devices worldwide were
websites, most notably search engines, media outlets using ad blocking by the end of 2016, of which over
and streaming services. There has been substantial half were mobile; this represents a 30% increase since
research into user perception of online ads and the 2015. During the early years of development of the
steps taken to avoid them. Ads are often seen as internet, advertisements displayed on websites were
annoying, or lead to a negative Web browsing not considered as invasive. Usually, they were
experience, and the prevalence of behavioral or presented in the form of static or dynamic banners
retargeted ads raises concerns about privacy. This in which included graphic designs. As technology

CSEIT195178 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 404-408 ]
404
V. Santhi, S. Abirami Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 404-408

developed and IT solutions became more available, Adaptive filtering of advertisements on web pages
the manner in which advertisements were shown on We presented a method to dynamically create custom
websites also developed. filters to avoid downloading unwanted URLs with
minimal interaction with the user, using a weighted-
II. LITERATURE SURVEY majority type algorithm. Both the standalone
extension (AdBlockLearner) and the test driver
Native Advertising and its Effects on Online (AdBlockLearnerTest) are available in [2], including
Advertising Native Advertising and its Effects on source code and documentation. They are compatible
Online Advertising According to Financial Times with most versions of Mozilla Firefox.
Limited (2013), the digital revolution has made many
business models obsolete and advertising companies Adblocking and Counter-Blocking: A Slice of the
see this as a welcome change. Online social Arms Race Adblocking tools like Adblock Plus
networking and internal structural changes in the continue to rise in popularity, potentially threatening
media market have changed how online advertising the dynamics of advertising revenue streams. In
is done. This change has many concurrent forms and response, a number of publishers have ramped up
implications, and the most apparent of them might be efforts to develop and deploy mechanisms for
the change in the various strategies and tactics detecting and/or counter-blocking adblockers (which
companies are adapting in response to said changes; we refer to as anti-adblockers), effectively escalating
mainly advertising online in the form of native the online advertising arms race. In this paper, we
advertising; while one company is inhibiting media develop a scalable approach for identifying third-
profitability in its current form, at the same time party services shared across multiple websites and use
furthering the its corporate sponsors' agenda to it to provide a first characterization of antiadblocking
promote and implement native advertising-or "non- across the Alexa Top-5K websites. We map websites
intrusive" advertising-as the norm of online that perform anti-adblocking as well as the entities
advertising. Abstract According to Financial Times that provide anti-adblocking scripts. We study the
Limited (2013), the digital revolution has made many modus operandi of these scripts and their impact on
business models obsolete and advertising companies popular adblockers. We find that at least 6.7% of
see this as a welcome change. Online social websites in the Alexa Top-5K use anti-adblocking
networking and internal structural changes in the scripts, acquired from 12 distinct entities – some of
media market have changed how online advertising which have a direct interest in nourishing the online
is done. This change has many concurrent forms and advertising industry.
implications, and the most apparent of them might be
the change in the various strategies and tactics Green Mining : Energy Consumption of
companies are adapting in response to said changes; Advertisement Blocking Methods
mainly advertising online in the form of native Extending battery life on mobile devices has become
advertising; while one company is inhibiting media an important topic recently due to the increasing
profitability in its current form, at the same time frequency of smartphone adoption. A primary
furthering the its corporate sponsors' agenda to component of smart phone energy consumption is
promote and implement native advertising-or "non- the apps that run on these devices. Many apps have
intrusive" advertising-as the norm of online embedded advertising and web browser apps will
advertising. show ads that are embedded on webpages. Other
researchers have found that advertising libraries and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


405
V. Santhi, S. Abirami Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 404-408

advertisements tend to increase power usage. But is determinants concerning users blocking
the converse true? If we use advertisement-blocking advertisements and IT tools. The paper refers to this
software will we consume less energy, or will the and systematizes knowledge in the scope of types of
overhead of ad-blocking consume more energy? This online advertisements and methods for blocking
study seeks to determine the effects of advertisements them using an adblock, and it identifies reasons and
on energy consumption, and the effects of attempts to main categories of reasons for users blocking
block the advertisements. We compared different advertisements.
methods of blocking advertisements on an Android
mobile phone platform and compared the power PROPOSED METHODOLOGY
efficiency of these methods. We found many cases
where ad-blocking software or methods resulted in The research presented in the paper was confronted
increased power use. with results of an analysis of application of ad blocks.
The obtained results will facilitate conducting further,
Blocking online advertising-A state of the art. more thorough research. Considerations included in
the paper can constitute a set of recommendations for
Compared with online advertising industry, there is publishers displaying advertisements on websites and
an even faster increase of ad blocker usage, which they can be useful for drawing conclusions and
influence badly on publishers’ and advertisers’ preparing guidelines for projects supporting
business. Thus more and more companies initialize sustainable development in the scope of online
their counter-ad blocking strategies, in which advertising.
customers choose to either disable their ad blockers
or leave without seeing the content. There are also
companies which abandon their counter-ad blocking
strategies after conducting them for a while due to
insufficient understanding of users’ ad blocking
behavior. In this study, we employed a quasi-
experiment framework and collected a large-size data
with the cooperation with Forbes Media. We aim to
identify factors influencing ad blocker usage.
Furthermore, we will model the interaction effects
among user profile, online behavior patterns, device
features on ad blocker usage propensity. Our study
contributes the literature of understanding ad blocker Connecting Network and Data Exchanging: -
usage by evaluating those principles using big amount
of real-world data. Computer network that interconnects computers or
systems within a large-scale network among internet
III. METHODS AND MATERIAL networks, it connects individual computer terminals,
mobile devices, and computer networks to the
EXISTING PROCESS Internet, enabling users to access Internet services,
Research concerning users blocking advertisements such as email and the software application. Network
constitutes a new research area both in the scope of computer devices that originate, route and terminate
analysis of collected data regarding that topic, the data are called network nodes. Nodes can include

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


406
V. Santhi, S. Abirami Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 404-408

hosts such as personal computers, phones, servers as Nodes Protected from Ad Block:-
well as networking hardware. Two such devices are Using metropolis sampler makes it more security for
said to be networked together when one device is Internet activity to be traced back to the user during
able to exchange data information with the other the visits of Web sites, online posts, and other
device, whether or not they have a direct connection communication forms. It intended to protect the
to each other. personal privacy of users, as well as their freedom
and ability to conduct confidential communication
Ad Block occurred (i.e. Ad Block Software):- by keeping their Internet activities from being
monitored. Assuming the networks is fully
Malicious software, is any software used to disrupt trustworthy, the system node must provide the
computer operation, gather sensitive information, or means to properly authenticate to connect other
gain access to private computer systems. Malware is nodes. Other nodes may impersonate trusted
defined by its malicious intent, acting against the computers, thus requiring system authentication. The
requirements of the computer user, and does not nodes and user may be trusted but within an
include software that causes unintentional harm due untrustworthy environment.
to some deficiency. The term badware is sometimes
used, and applied to both true (malicious) malware IV. RESULTS AND DISCUSSION
and unintentionally harmful software. During the
data transferring or downloading Computer users CONCLUSIONS
often download malware (i.e. malicious software) to The paper shows the reasons for using adblock
their computer by unknowingly visiting a malicious programs and presents the current most popular
webpage hosting a drive-by download attack, programs of that type as well as their characteristics.
clicking on a malicious link included in email, Apart from aforementioned, a survey study was
opening an attachment which includes an exploit, conducted involving a group of Polish internet users,
and destroy the data. which shows that they actively and consciously block
the display of advertisements in their web browsers.
Detecting, Blocking using Check Sum and Metropolis In this study we show how to adblock software
Sampler:- technologies are applied in Polish environment. We
discover a strong desire to block advertisements and
Anti-malware software is computer software used to well known ways to do this. Also in Poland there are
prevent, detect and remove the virus software. But many adapted solutions to improve adblock accuracy
the malware (i.e. malicious software) occurred during and coverage.
download the software or clicking the malware link
in spite of clicking the e-mail unknowingly it affects FUTURE ENHANCEMENT
entire system. Check sum algorithm was originally Future directions of research will concern the
developed to detect malware based on the priority of problem of intrusiveness of advertisements and ways
other kinds of malware during the data processing. for decreasing it. As it was mentioned in the paper,
And encounter based algorithm is used to block the users are not against advertisements being
malware (i.e. malicious software). Metropolis Sampler displayed—they oppose the way in which they
algorithm is to block and encounter the malware. To appear on their screen, e.g. by suddenly covering the
prevent the nodes and against to filtered the thread. viewed content. Further research will be also aimed
And it is recover the data in a large scale systems. at a quantitative analysis regarding elements blocked

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


407
V. Santhi, S. Abirami Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 404-408

on websites, such as unwanted advertisements,


scripts tracking user's activity and malicious activities
(malware).

V. REFERENCES

[1]. Adblock Plus, https://adblockplus.org, last


accessed 2017/07/07
[2]. AdBlock, https://getadblock.com, last accessed
2017/07/07
[3]. Ajdari, D., Hoofnagle, C., Stocksdale, T., Good,
N.: Web Privacy Tools and Their Effect on
Tracking and User Experience on the Internet
(2013)
[4]. AlleBlock, https://tarmas.pl/alleblock, last
accessed 2017/07/07
[5]. Allowing acceptable ads in Adblock Plus,
https://adblockplus.org/en/acceptable-ads, last
accessed 2017/07/07
[6]. CertyficateIT, https://www.certyficate.it, last
accessed 2017/07/07
[7]. EasyList, https://easylist.to, last accessed
2017/07/07
[8]. Eyeo GmbH, https://eyeo.com/en/press#stats,
last accessed 2017/07/07
[9]. Google, Microsoft and Amazon pay to get
around ad blocking tool,
https://www.ft.com/content /80a8ce54-a61d-
11e4-9bd3-00144feab7de, last accessed
2017/07/07

Cite this article as :

V. Santhi, S. Abirami, "Adblock Usage in Web


Advertisement", International Journal of Scientific Research
in Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume 5
Issue 1, pp. 404-408, January-February 2019.
Available at doi : https://doi.org/10.32628/CSEIT195178
Journal URL : http://ijsrcseit.com/CSEIT195178

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


408
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195190
Smart-Home Automation Using IoT-Based Sensing and
Monitoring Platform
R. Sugantha Lakshmi1, P. Karthika2, A. Rajalakshmi3, M. Sathya4
1 Assistant Professor, Department of Computer Science and Engineering, Kings College of Engineering,
Thanjavur, Tamil Nadu, India
2Department of Computer Science and Engineering, Kings College of Engineering, Thanjavur, Tamil Nadu,
India
ABSTRACT

To comfort the living conditions within of a human his home, she can use home monitoring and automation
are utilized. The standards of human’s comfort in homes can be categorized into several types. Among these
categories, the most significant ones are the thermal comfort, which is related to temperature and humidity,
followed by the visual comfort, related to colors and light, and hygienic comfort, associated with air quality.We
are proposing a system to monitoring the paremeters range which will increase the comfortness of the
human.Additionally,we incorporate the smart home technique which controls the home appliances by
intelligent automatic execution of commands after analyzing the collected data. Automation can be
accomplished by using the Internet of Things (IoT). This gives the inhabitant accesses to certain data in the
house and the ability to control some parameters remotely. We propose the complete design of an IoT based
sensing and monitoring system for smart home automation. That uses the EmonCMS platform for collecting
and Visualizing monitored data and remote controlling of home appliances and devices. The selected platform
is very flexible and user-friendly. The sensing of different variables inside the house is conducted using the
NodeMCU-ESP8266 microcontroller board, which allows real-time data sensing, processing and
uploading/downloading to/from the EmonCMS cloud server.
Keywords : IOT, CMS, Humidity, Air Quality, Temperature

I. INTRODUCTION load, but smart home technology is combination of


both and much more. Smart home technology is
Smart Home is collaboration of technology and currently being implemented for entire house in
services through a network for better quality living. particularly kitchen and living room. Basically, smart
A smart home allows the entire home to be home facilitates users with security, comfortable
automated and therefore provide ease and living and energy management features as well as
convenience to everyday activities in the home. This added benefits for disabled individuals.
technology is used to make all electronic devices to
act ‘smart’. In the near future almost all the electronic This technology might sound new but it just uses the
devices will take advantage of this technology existing technologies. A smart device is a common
through home networks and the internet. Many appliance with a much more complex computer
people think this technology as pure networking. installed to give it more functionality. These
Others think this technology will reduce their work functions are the ones which makes it so different.

CSEIT195190 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 409-413 ]
409
R. Sugantha Lakshmi et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 409-413

Cable broadband, DSL, Bluetooth and wireless ZigBee modules due to their low power and cost
technologies provide a way to have a home efficiency.
networked for devices to communicate with each
other as well as internet. These technologies either D. GSM Based Home Automation System
wired or wireless provide foundation on which smart A smart home automation system implemented by
home will operate. using Global System for Mobile communication
(GSM). The hardware architecture of the system
II. LITERATURE SURVEY consists of GSM modem, PIC16F887 microcontroller
and smartphone. The system used a GSM modem to
A. Bluetooth based home automation system control electric appliances through SMS request.
Home automation systems using smartphone, PIC16F887 microcontroller interfaced with a GSM
Arduino board and Bluetooth technology are secured modem and it is used to read and decode the received
and low cost. A Bluetooth based home automation SMS to execute the specific command.
system proposed by R.Piyare and M.Tazil. The
hardware architecture of this system consists of the F. EnOcean based home automation system
Arduino BT board and a cell phone, the The EnOcean is newly developing energy harvesting
communication between Arduino BT board and cell technology used in transportation, building and home
phone is wirelessly using Bluetooth technology. The automation systems. EnOcean’s technology is
Arduino BT board has a range of 10 to 100 meters, 3 productively in logistics as well as in the industry due
Mbps data rate and 2.4 GHz bandwidth. to energy efficiency and easily installing device
anywhere for users ease which significantly save the
B. Voice recognition based home automation installation cost up to 40%. Moreover EnOcean’s
Voice recognition based home automation system devices utilize 315 MHz band and it provides
proposed and implemented by a researcher. The convenient ways for home automation system.
hardware architecture of this system consists of
Arduino UNO and smartphone. The wireless III. EXISTING PROCESS
communication between the smartphone and the
Arduino UNO is done through Bluetooth technology. Smart home technology has the potential of bringing
Android OS has a built-in voice recognizing feature benefits to modern households and their inhabitants.
which is used to develop a smartphone application Yet, ever since its early development, it has been
which has ability to control the home appliances struggling to reach mass consumer adoption. Privacy
from user voice command. and security, trust issues, reliability, and price are just
some of the challenges smart home technology is
C. ZigBee Based Wireless Home Automation System facing. In addition, literature suggests that there is an
ZigBee based wireless home automation system has evident gap between the functionalities offered by
also been studied, it consists of three main modules, smart devices and users’ needs. Investigating these
handheld microphone module, central controller potential adoption challenges in some more detail, we
module and appliance controller module. Handheld conducted an interview study with existing smart
microphone module use ZigBee protocol and central home technology users. Results show that privacy
controller module are based on PC. In this system, and security are still the most prominent hindering
Microsoft speech API is used as a voice recognition factors, and that the often insufficient
application, wireless network is established using RF interoperability of devices becomes an ever-growing

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


410
R. Sugantha Lakshmi et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 409-413

concern. Also, smart home devices are consistently humidity, luminosity, and air quality (i.e. CO2 and
perceived as complex and expensive, and lack dust levels). Usually, the term “Big Data” is often
perceived value and trustworthiness. associated with IoT and refers to the large number
data collected from a large number of sensors and
IV. PROPOSED METHODOLOGY devices to be processed, which is the case here. In
fact, depending on the number of sensors and the
As mentioned earlier, the monitored parameters in measurement-sampling rate, a large number of data
homes are selected to ensure a smart, efficient and might be generated and transmitted through the
comfortable living conditions. Therefore, one hub is internet to the cloud server, which should be able to
made to measure temperature, humidity, light handle such large amount of data communication and
intensity, proximity, and CO2 levels. This hub can be archiving.
placed in each room of the house, and preferably at a
high level above the ground and below the ceiling. B. Data Transmission
Collected data can be processed and used to control Transmitting data between devices and controllers is
smart devices such as smart LEDs and A/Cs. The done using one or more communication technologies.
schematics of the proposed design. It comprises These include Bluetooth, ZigBee, Wi-fi, Ethernet or
mainly sensors, processor, user interface, and data GSM. Bluetooth and ZigBee are commonly used for
communication and transmission links. in-home data transmission and control and provide
the link between the sensors and the central
processing unit. These data communication
technologies are popular because of their low power
consumption, and simplicity to implement.
Furthermore, IoT requires the use of either Wi-fi or
Ethernet to connect to the internet. While Ethernet
is by far faster than Wi-fi connection, the high data
rate is not essential for home automation applications.
Additionally, Wi-fi has the advantage of mobility,
making it more widespread in most systems. Wi-fi
technology consumes more energy than alternatives
such as Bluetooth or ZigBee. The power consumption
can be reduced by lowering the frequency of data
V. IMPLEMENTATION RESULTS uploading.

This section reviews the typical configuration of C. Data Processing (Microcontroller)


home automation system using an IoT platform. The data collected in a home automation system is
Illustrates such a platform which consists mainly of usually processed and managed by a microcontroller
data sensing and acquisition, processing, transmission, such as Arduino, Raspberry Pi, and NodeMCU.
and display. Raspberry Pi is a small single microcontroller
computer. With a higher RAM of 256MB or 512Mb,
A. Data Sensing and Acquisition depending on the model, it can handle more complex
To ensure a good comfort level, the typically tasks than other controllers and used mostly as a
monitored parameters in houses are temperature, central processing unit for multiple devices. Most

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


411
R. Sugantha Lakshmi et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 409-413

new models of Raspberry Pi have USB and Ethernet VI. CONCLUSION


ports, making it easy to upload data to the internet.
On the other hand, Arduino is a single-board The proposed design of the smart solar home is very
microcontroller that can be simply programmed to flexible and can be easily expanded and applied to
execute commands. Arduino comes in a variety of larger buildings by increasing the number of sensors,
models with onboard flash memory ranging from measured parameters, and control devices. More
32kB to 512kB, and a typically a RAM of 2kB. functionality and smartness could be also added to
Evidently, this controller is less powerful than the the existing system for making the house automation
Raspberry Pi. However, most Arduino models are system grow, adapt, and evolve by itself using
cheaper, easier to handle, and powerful enough to advanced artificial intelligence.
deal with home automation tasks. Another option is
to use the Node MCU. It is an Arduino based VII. FUTURE ENHANCEMENT
microcontroller but with the addition of the ESP8266
Wi-Fi chipset. This microcontroller has a memory of There are a variety of enhancements that could be
128kB and 4MB storage. It is mostly used for a single made to this system to achieve greater accuracy in
IoT application, or to eliminate the need for a central sensing and detection.
processing unit. Since each part of the system can a) There are a lot of other sensors that can be used to
upload data to the server individually, this also increase the security and control of the home like
lowers the complexity of the coding and the pressure sensor that can be put outside the home to
connection chain. The biggest advantage of the Node detect that someone will enter the home.
MCU over the other alternatives is the significantly b) Changing the way of the automated notifications
low price for a controller that can connect to the by using the GSM module to make this system more
internet directly using Wi-fi, without the need for professional.
any additional peripherals of modules. An issue is c) A smart garage that can measure the length of the
that the Node MCU board has only one analog input, car and choose which block to put the car into it and
which limits its applications to as single data it will navigate the car through the garage to make
monitoring system. However, this drawback can be the parking easy for the homeowner in his garage.
compensated for by using the ASD115, which is an
analog to digital converter that has four analog input VIII. REFERENCES
pins and has a higher conversion resolution of 16-bits.
[1]. Hayet Lamine and Hafedh Abid, "Remote
D. Data Display & User Interface control of domestic equipment from an
The interaction between the user and the system can Android application based on Raspberry Pi
be done in several ways. One option is to use an card", IEEE transaction 15th international
application. There are many simple means to create a conference on Sciences and Techniques of
mobile or a web based app to display data, even with Automatic control & computer engineering -
a limited knowledge of programming. Another STA'2014, Hammamet, Tunisia, December 21-
control option is through mobile GSM where the user 23, 2014.
can send commands in codes by SMS to the [2]. Vaishnavi S. Gunge and Pratibha S. Yalagi,
microcontroller. This control method requires a "Smart Home Automation: A Literature
special GSM module added to the circuit. This Review", National Seminar on Recent Trends in
method can also be applied using emails. Data Mining- RTDM 2016.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


412
R. Sugantha Lakshmi et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 409-413

[3]. M. B. a. W.Fincher, "Standards of Human International Conference on Renewable


Comfort: Relative and Absolute," The Energies and Vehicular Technology, 2012.
University of Texas at Austin, Austin, 2009.
[4]. Archana N. Shewale and Jyoti P. Bari,
"Renewable Energy Based Home Automation Cite this article as :
System Using ZigBee", IJCTEE Volume 5, Issue
R. Sugantha Lakshmi, P. Karthika, A. Rajalakshmi, M.
3, June 2015.
Sathya, "Smart-Home Automation Using IoT-Based
[5]. Ali Mohammed Al-Kuwari, Cesar Ortega-
Sensing and Monitoring Platform", International
Sanchez, Atif Sharif and Vidyasagar Potdar,
Journal of Scientific Research in Computer Science,
"User-Friendly Smart Home Infrastructure:
Engineering and Information Technology
BeeHouse", IEEE 5th 5th International
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
Conference on Digital Ecosystems and
409-413, January-February 2019. Available at doi :
Technologies, May31-June3 2011, Daejeon,
https://doi.org/10.32628/CSEIT195190
Korea.
Journal URL : http://ijsrcseit.com/CSEIT195190
[6]. Lamir Shkurti, Xhevahir Bajrami, Ercan
Canhasi, Besim Limani, Samedin Krrabaj and
Astrit Hulaj, "Development of Ambient
Environmental Monitoring System Through
Wireless Sensor Network Using NodeMCU and
WSN Monitoring", 6th MEDITERRANEAN
CONFERENCE ON EMBEDDED
COMPUTING (MECO), JUNE 11-15 2017,
BAR, MONTENEGRO.
[7]. International Electrotechnical Commission,
Internet of Things: Wireless Sensor Networks,
Geneva, Switzerland: International
Electrotechnical Commission, 2014. [8]
Grégory Nain, Francois Fouquet, Brice Morin,
Olivier Barais, JeanMarc Jézéquel, "Integrating
IoT and IoS with a Component-Based
approach", 36th EUROMICRO Conference on
Software Engineering and Advanced
Applications, 2010
[8]. S. Jaffe, "DESIGN OF INEXPENSIVE AND
EASY TO USE DIY INTERNET OF THINGS
PLATFORM," Faculty of California Polytechnic
State University, 2016.
[9]. F. Touati, M. Al-Hitmi, and H. Bouchech,
"Towards understanding the effects of climatic
and environmental factors on solar PV
performance in arid desert regions (Qatar) for
various PV technologies," 2012 First

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


413
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195198
Anticipated Security Model for Session Transfer and Services
Using OTP
J. Jose Merlin1, A. Prathipa2, G. Ramyadevi3, P. Radhika4
Department of Computer Science and Engineering, St. Joseph’s College of Engineering and Technology,
1-3

Thanjavur, Tamil Nadu, India


4 Assistant Professor, Department of Computer Science and Engineering, St. Joseph’s College of Engineering and
Technology, Thanjavur, Tamil Nadu, India
ABSTRACT

Internet security is a branch of computer sciences often involving browser security, network security,
applications and operating systems to keep the internet as a secure channel to exchange information by
reducing the risk and attacks. There are a number of studies that have been conducted in this field resulting in
the development of various security models to achieve internet security. However, periodic security reports and
previous studies prove that the most secure systems are not immune from risk and much effort is needed to
improve internet security. This paper proposed a simple security model to improve internet applications
security and services protections, specified access control, cryptographic, cookies and session managements,
defense programing practices, care for security from early stage on development life cycle, use hardware
authentication techniques in access control, then propose cryptographic approach by mix MD5 with Based64,
consider session and cookies types and ways to keep it secure. Additionally, these practices discussed the most
important web security vulnerability and access control weakness and how to overcome such weaknesses,
proposed an approach to measure, analyze and evaluate security project according to software quality standard
ISO 25010 by using Liker scale, finally ended by case study. The effort of this paper represents a set of
techniques and tips that should be applied within each web application development process to maintain its
security.
Keywords : Combinationarial Optimization Identification, Public Key Infrastructure, Partial Forward Secrecy.

I. INTRODUCTION as efficiency, simplicity, and cost effectiveness. On


the other hand, these enterprises face a new
Web applications emerged as one of the main challenge on how to keep these applications and
technologies and progressed in the last decade. They services secure. They are spending many resources to
play a main role in our daily lives on areas such as manage and handle data storage in plain text in malt
health-care, online services, businesses (public and locations throughout the enterprise. Recently, the
private sectors), E-banking, and e-commerce. Web security of web applications and services derived
applications are expected to increase the trading attention from both the industry field and research
volume on the market exceeding $ 1 trillion in the community. There is a strong need for standard web
next several years. Many public or private enterprises security models for several reasons including: the
developed their applications and services based on integration of the heterogeneous and distributed
the internet to take advantage of its features as well systems, availability of high volumes of sensitive

CSEIT195198 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 414-418 ]
414
J. Jose Merlin et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 414-418

information and data maintained by corporation's application development process to maintain its
servers and government agencies, safety from easily security.
distributed malicious software, and computer crimes. Security Issues for Web-Based Applications: Issues
On the other hand, software security has become a And Solutions For The Safe Transfer Of Clinical
quality attribute used to measure the quality of Trials Data Over The Internet When building any
software according to standards such as ISO software web-based application, the issue of security is always
quality standard 25010. a concern. When the application is designed to allow
the sharing of clinical trials data across the Internet,
II. LITERATURE SURVEY security becomes one of the major topics. This paper
describes what approach iBiomatics, a SAS company,
Proposed Security Model For Web Based has taken in the construction of their biometrics
Applications And Services Portal. The paper presents a basic overview of
Internet security is a branch of computer sciences security for an Internet based application, describes
often involving browser security, network security, what security features were built into the application,
applications and operating systems to keep the what software tools were used to implement various
internet as a secure channel to exchange information aspects of the security model, and how the actual
by reducing the risk and attacks. There are a number hardware configuration can be a critical piece of the
of studies that have been conducted in this field overall security of a web based application.
resulting in the development of various security Securing Frame Communication in Browsers
models to achieve internet security. However, Many web sites embed third-party content in frames,
periodic security reports and previous studies prove relying on the browser’s security policy to protect
that the most secure systems are not immune from them from malicious content. Frames, however, are
risk and much effort is needed to improve internet often insufficient isolation primitives because most
security. This paper proposed a simple security model browsers let framed content manipulate other frames
to improve internet applications security and services through navigation. We evaluate existing frame
protections, specified access control, cryptographic, navigation policies and advocate a stricter policy,
cookies and session managements, defense which we deploy in the open source browsers. In
programming practices, care for security from early addition to preventing undesirable interactions, the
stage on development life cycle, use hardware browser’s strict isolation policy also hinders
authentication techniques in access control, then communication between cooperating frames. We
propose cryptographic approach by mix MD5 with analyze two techniques for inter-frame
Based64, consider session and cookies types and ways communication. The first method, fragment
to keep it secure. Additionally, these practices identifier messaging, provides confidentiality without
discussed the most important web security authentication, which we repair using concepts from
vulnerability and access control weakness and how to a well-known network protocol. The second method,
overcome such weaknesses, proposed an approach to post Message, provides authentication, but we
measure, analyze and evaluate security project discover an attack that breaches confidentiality. We
according to software quality standard ISO 25010 by modify the post Message API to provide
using Likert scale, finally ended by case study. The confidentiality and see our modifications
effort of this paper represents a set of techniques and standardized and adopted in browser
tips that should be applied within each web implementations.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


415
J. Jose Merlin et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 414-418

Protecting Browsers From Dns Rebinding Attacks the secret key or password shared between two users.
DNS rebinding attacks subvert the same-origin policy The main goal of password-based authenticated key
of browsers, converting them into open network exchange protocols is to restrict the adversary to this
proxies. Using DNS rebinding, an attacker can case only. The existing system consider that
circumvent organizational and personal firewalls, password-based authenticated key exchange in the
send spam email, and defraud pay-per-click three-party scenario, in which the users trying to
advertisers. We evaluate the cost effectiveness of establish a secret do not share a password between
mounting DNS rebinding attacks, finding that an themselves but only with a trusted server. In this
attacker requires less than $100 to hijack 100,000 IP existing system the key exchange between three
addresses. We analyze defenses to DNS rebinding authenticated users so, that the privacy of the system
attacks, including improvements to the classic “DNS is failed to overcome. Since we want to trust as little
pinning,” and recommend changes to browser plug- as possible the third party, we develop a new notion
ins, firewalls, and Web servers. Our defenses have called key privacy which roughly means that, even
been adopted by plug-in vendors and by a number of though the server’s help is required to establish a
open-source firewall implementations. session key between two users in the system, the
server should not be able to gain any information on
Assessment of Access Control Systems the value of that session key.
Adequate security of information and information
systems is a fundamental management responsibility. IV. PROPOSED METHODOLOGY
Nearly all applications that deal with financial,
privacy, safety, or defense include some form of A secure 2-party password-based key exchange is a
access control. Access control is concerned with 2PAKE protocol where the parties use their password
determining the allowed activities of legitimate users, in order to derive a common session key that will be
mediating every attempt by a user to access a used to build secure channels. Several practical
resource in the system. In some systems, complete schemes can be used in the instantiation of the 2-
access is granted after successful authentication of the party password-based key exchange of our generic
user, but most systems require more sophisticated construction. In these proposed system, instead of
and complex control. In addition to the exchange the password the owner going to send the
authentication mechanism (such as a password), link to get that particular data and in user side by
access control is concerned with how authorizations using that link they can retrieve that data easily.
are structured. In some cases, authorization may Once the user make logout means they cannot again
mirror the structure of the organization, while in enter into the system. The link can be used only once.
others it may be based on the sensitivity level of
various documents and the clearance level of the user V. IMPLEMENTATION RESULTS
accessing those documents. This publication explains
some of the commonly used access control services Search engine recognized
available in information technology systems. Search engine indexing collects, parses, and
stores data to facilitate fast and accurate information
III. EXISTING PROCESS retrieval. Index design incorporates interdisciplinary
concepts from linguistics, cognitive psychology,
Password-based authenticated key exchange is mathematics, informatics, and computer science. An
protocols which are designed to be secure even when alternate name for the process in the context

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


416
J. Jose Merlin et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 414-418

of search engines designed to find web pages on the industry-specific vertical search engines. SEO differs
Internet is web indexing. from local search engine optimization in that the
Popular engines focus on the full-text indexing of latter is focused on optimizing a business' online
online, natural language documents. Media presence so that its web pages will be displayed by
types such as video and audio and graphics are also search engines when a user enters a local search for
searchable. its products or services. The former instead is more
focused on national or international searches.
Meta search engines reuse the indices of other
Authentication
services and do not store a local index, whereas
An authentication module is a plug-in that collects
cache-based search engines permanently store the
user information such as a user ID and password, and
index along with the corpus. Unlike full-text indices,
compares the information against entries in a
partial-text services restrict the depth indexed to
database. If a user provides information that meets
reduce index size. Larger services typically perform
the authentication criteria, the user is validated and,
indexing at a predetermined time interval due to the
assuming the appropriate policy configuration,
required time and processing costs, while agent-
granted access to the requested resource. If the user
based search engines index in real time.
provides information that does not meet the
authentication criteria, the user is not validated and
denied access to the requested resource. OpenSSO
Enterprise is deployed with a number of
authentication modules.

Session Handling
A web session is a sequence of network HTTP request
and response transactions associated to the same user.
Modern and complex web applications require the
retaining of information or status about each user for
the duration of multiple requests.

User preference
Optimized search results
Understanding user preference settings. Completing
Search engine optimization (SEO) is the process of
the user preferences pages enables the system
affecting the online visibility of a website or a web
to use the field values that you specify as the default
page in a web search engine's unpaid results—often data that appears on various inquiry pages. ... This
referred to as "natural", "organic", or "earned" results.
enables default values to be set by user ID; the
In general, the earlier (or higher ranked on the
user can override the defaults when running an
search results page), and more frequently a website
inquiry.
appears in the search results list, the more visitors it
will receive from the search engine's users; these Temporary User Restriction
visitors can then be converted into To ensure a consistent user experience, we monitor
customers.[1] SEO may target different kinds of the level of requests to the my hr toolkit system from
search, including image search, video user accounts. If an abnormally high frequency of
search, academic search,[2] news search, and requests is detected, an escalating blocking process is
deployed

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


417
J. Jose Merlin et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 414-418

VI. CONCLUSION offers improvements in quality and video


Thus the process a secure 2-party password-based key conferencing applications in the near future.
exchange is a 2PAKE protocol where the parties use
their password in order to derive a common session VIII. REFERENCES
key that will be used to build secure channels. [1]. Michel Abdalla, Pierre-Alain Fouque, and David
Several practical schemes can be used in the Pointcheval, "Password-Based Authenticated Key
instantiation of the 2-party password-based key Exchange in the Three-Party Setting", IEEE
Proceedings on InformationSecurity, Volume 153,
exchange of our generic construction. In these
number 1, March 2006 , pp. 27-39.
proposed system, instead of exchange the password
[2]. Whitefield Diffie, Martin E. Hellman, "New directions
the owner going to send the link to get that particular
in cryptography", IEEE Transactions on Information
data and in user side by using that link they can Theory, pp. 644-654,1976.
retrieve that data easily. Once the user makes logout [3]. Behrouz A. Forouzan, Cryptography and network
means they cannot again enter into the system. The security, Tata McGraw-Hill, 2007.
link can be used only once. [4]. Hyun-SeokKim ,Jin-Young Choi, "Enhanced
passwordbased simple three-party key exchange
VII. FUTURE ENHANCEMENT protocol", Elsevier, Computers and Electrical
The introduction of VoIP has enabled many Engineering, 35, pp.107–114,2009.
[5]. Yuanhui Lin, MengboHou, Qiuliang Xu, "Strongly
organizations to supplement or replace existing
password based three party authenticated key
circuit switched telephone networks, providing a
exchange protocol", Ninth International conference on
new value added services. Numerous technological
Computational Intelligence and security, IEEE, pp.
problems are faced on the issue of quality of service. 555- 558,2013.
The packet loss is a major problem. This problem [6]. Rongxing Lu, Zhenfu Cao, "Simple three-party key
cannot be avoided but it can be minimized. The exchange protocol", Elsevier, computers & security 26,
packet repetition with static playout algorithm pp. 94-97,2007.
performs moderately than other replacement [7]. Chao Lv , MaodeMab, Hui Li, JianfengMaa, Yaoyu
techniques. Based on time delay, loss rate and MOS, Zhang, "An novel three-party authenticated key
the better results are obtained. Hence, the receiver exchange protocol using one-time key", Elsevier,
Journal of Network and Computer Applications (36),
based repetition provides high quality VoIP signal in
pp. 498–503,2013.
the streaming audio. The processing of damaged
[8]. Alfred Menezes, BerkantUstaoglu, "On reusing
packets has been established as a challengeable topic
ephemeral keys in Diffie Hellman key agreement
for further research. This work can be extended by protocol",International Journal of Applied
incorporating other statistical based methods. Voice Cryptography,ACM,pp. 154-158, 2010.
over IP provides unique and inexpensive services to
the internet people. Internet is a single network Cite this article as : J. Jose Merlin, A. Prathipa, G.
shared by the world. Audio conferencing is one of Ramyadevi, P. Radhika, "Anticipated Security Model for
the applications used nowadays in the internet Session Transfer and Services Using OTP", International
community. This is purely based on multiplexing and Journal of Scientific Research in Computer Science,
multicasting scheme. It promises to deliver cost Engineering and Information Technology (IJSRCSEIT),
ISSN : 2456-3307, Volume 5 Issue 1, pp. 414-418, January-
savings to users and service providers. Hence the
February 2019. Available at doi :
Voice M-M based interleaving provides high quality
https://doi.org/10.32628/CSEIT195198
VoIP signal in the streaming audio. This work can be
Journal URL : http://ijsrcseit.com/CSEIT195198
extended by incorporating video based methods. It

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


418
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951106
Managing university defined Study material by using Online
Portal
Neha Bitodkar #, Prof. Aniket Bhoyar*, Suraj Chavhan#, Priyanka Wahane#
#
Department of Information Technology, Rashtrasant Tukadoji Maharaj Nagpur University,
Nagpur, Maharashtra, India
*J D College of Engineering and Management, Nagpur, Maharashtra, India
ABSTRACT

E-learning system is very much popular in today`s digital era due to which one can easily get access to the
required information or knowledge. Due to online learning there are several solutions to a single problem is
available as the internet has grown in the largest manner. Even the online learning is changing the load of
traditional classrooms. E-learning provides an integrated environment where access to particular information
can be found at one place in various forms. The proposed technique is for creating an E-learning system with an
integrated environment for students of higher education. So that, they are able to get access to an appropriate
solutions for the particular question asked in an university examination. The system will provide free access to
the study material for students simply by logging in. This will be creating an integrated library for each subject
and create model answer section for questions related to the subject. This will help student to learn subject
easily, the question and their answer that can be asked in the university examination.
Keywords: Raspberry pi, USB camera, motor driver, motor, LCD display, road side speed sign controlling

I. INTRODUCTION question papers, model answer paper, video lecture,


E-books etc.
In today`s digital era online education is not a big
deal. There are several online courses started by govt This system is designed using technologies such as
for the online learning of students. Due to internet a Bootstrap, CSS and PHP. The database used is Mysql
large no. of information can be available of particular to store the login ID of the users and their data. This
subject. Where as integrated environment if designed technologies will make an integrated environment
there are less chances of confusion about the subject. that will contain information in several formats.
According to the syllabus designed by the university
examination, and the question paper set, model II. PROBLEM STATEMENT
answers section is designed which will make it easy
for the students to get extraordinary knowledge of To create a database that will manage study material,
the syllabus. the system is created of sections such as question
paper, model answer, E-books, semester-wise subject
This system will make students available with differentiation. Basically, this will provide the
complete information about the subject. The syllabus students all the study material of higher education at
will be covered completely in the question answer one place. So that it will be easier for them to study
section. The study material includes syllabus,

CSEIT1951106 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 419-421 ]
419
Neha Bitodkar et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 419-421

without any confusion as there are several solutions Author : Yogesh Deshpande , Samit Bhattacharya,
are available online for a single solution. Pradeep Yammiyavar
Description: This paper focuses on analyzing the
III. LITERATURE REVIEW ability of children`s for adopting e-learning
environment in comparison with traditional
Title 1 : Implementation of E-learning on full stack classrooms. Online learning can be an additional tool
concept for the school environment to get more knowledge
Author: Nitish Kumar, Chaitali Alhat, Rupali about the particular subject.
Bachhav
Description : The designed system provides materials IV. PROPOSED SYSTEM
in different formats. It can be in the form of videos,
pdf etc. Their purpose is to provide an open free The proposed system will be a beneficiary product for
course which provides access to the material after the the students of university and will get a complete
registration process is completed. This system focuses access to all the material related to their syllabus
on interactive learning between the students and the designed by university.
proffesors.
Title 2: Open educational resource : opportunities The system is designed using Bootstrap, CSS and PHP.
and challenges for Indian higher education. Mysql database is used for storing the login ID as well
Author : Dr. Indrajeet Dutta as the data of the user. The system will allow students
Description : This paper gives information about the for complete study of subjects due to the question
opportunities and challenges for Indian higher answer section is available. Question papers will be
education related with open educational resources. available on the subjects along with their model
This paper gives the detailed information about the answer.
several online courses started by the govt. such as
NMECIT, NKN, SHAKSHAT, NPTEL, EKLAVYA, The designed system contains information in
OSCAR E-GRID etc. This papers says that the different formats. It will also contain information in
opportunities can be increased as students can hear the form of video lectures and e-books. The system is
the lectures from international proffesors. branch wise and semester wise separated according to
Title 3 : Role of consortia for effective use of e- the university.
resources in Higher education : A practical approach Data Flow Diagram:
in Indian liabraries.
Author: Sunil Goria
Description : This paper says that in the last decade
the liabraries of higher education and the research
institutes are using the e-resources. In consortium
mode, one can be able to access the journals in Indian
libraries using e-resources. This paper focuses on
increasing use of e-resources.
Title 4: A behavioural approach to modelling Indian
children`s ability of adopting to e-learning
environment.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


420
Neha Bitodkar et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 419-421

V. CONCLUSION Cite this article as :

In this work, after login the study material will be Neha Bitodkar, Prof. Aniket Bhoyar, Suraj Chavhan,
available for the students. This is a beneficiary Priyanka Wahane, "Managing university defined
product for the students. University syllabus will be Study material by using Online Portal", International
available on this website along with related material Journal of Scientific Research in Computer Science,
such as question answer set, video lectures, model Engineering and Information Technology
answer, e-books, notes.It will help the students for (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
complete learning of subjects designed by university. 419-421, January-February 2019.
Due to E-learning one can access this information Journal URL : http://ijsrcseit.com/CSEIT1951106
form anywhere at anytime.

VI. REFERENCES

[1]. Nitish Kumar, Chaitali Alhat, Rupali Bachhav


"Implementation of E-learning system on full
stack concept", Mat Journals year 2018
[2]. Dr. Indrajeet DUTTA," Open educational
resources (oer): opportunities and challenges
for indian higher education", Turkish Online
Journal of Distance Education -TOJDE April
2016
[3]. Sunil Goria, "Role of consortia for effective use
of e-resources in higher education: A practical
approach in Indian libraries."International
Journal of information Dissemiation &
Technology, 2012
[4]. Yogesh Deshpande , Samit Bhattacharya,
Pradeep Yammiyavar, "A behavioral approach
to modeling Indian children’s ability of
adopting to e-learning environment" 2012 4th
International Conference on Intelligent Human
Computer Interaction(IHCI)

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


421
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951107

A Review on Car Collision Avoidance System based on


Interpretation of Vehicle Event Data Recorder
Prof. Tejal Irkhede, 2Rohit Kashayp, 2Nikhil Kumar, 2Vivek Ranjan, 2Manish Kumar
1

1Assistant Professor, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,


Maharashtra, India
2BE Students, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,

Maharashtra, India

ABSTRACT

According to the World Health Organization (WHO), in excess of a million people on the planet pass on
consistently by virtue of vehicle accidents. Notwithstanding care cause, this issue is up 'til now growing due to
rider's poor practices, for instance, tank driving, speed driving, riding without sufficient rest, riding with no top
protection, etc. The Event Data Recorder (EDR) or vehicle 'black box' is a standout amongst the most valuable
instruments accessible to crash Reconstructionist. For vehicles outfitted with this innovation, and when the
recorders fill in as planned, the EDR gives crash agents significant data that in numerous examples could not be
gathered. In this paper, we present the issues related to accident detections and some existing systems for
prevention of accidents.
Keywords : Car Accidents, Road Safety, Public Health, Black Box, Event Data Recorder

I. INTRODUCTION intensifying circumstance. In the event that the


pattern proceeds with, the absolute number of street
Fatalities and wounds coming about because of traffic passings in India would increment by 100%
streetcar crashes are a noteworthy and developing somewhere in the range of 2013 and 2027 [1].
general medical issue in India. Consistently about Without expanded endeavors and new activities, the
2,650 individuals get executed and 9,000 get harmed all-out number of street traffic passings in India is
because of car crashes. In 2013, the most recent year probably going to cross the sign of 250,000 by 2025.
for which data is accessible, 137,423 individuals The primary point of this investigation is to examine
passed on and 469,900 individuals got harmed the streetcar crashes in India at the national, state,
because of street accidents in India [1]. Car crashes and metropolitan city level. The center is
have now earned India a questionable qualification; distinguished the significant street wellbeing issues
with about 140,000 passings every year, the nation and examine countermeasures that would possibly
has overwhelmed China to top the world in street address the particular street security issues. The
fatalities. India is the main nation on the planet essential wellspring of data for the examination is
which faces in excess of 15 fatalities and 53 wounds Accidental Deaths and Suicides in India, 1970 to 2013
each hour as a result of street crashes. While in many distributed by the National Crime Records Bureau,
created and creating nations including China, the Ministry of Home Affairs, Government of India, New
circumstance is commonly improving, India faces an Delhi.

CSEIT1951107 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 422-428 ]
422
Prof. Tejal Irkhede et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 422-428

The investigation demonstrates that amid the most The mishap examination includes the investigation of
recent ten years, street unplanned fatalities in India accident scenes and the documentation of all
have expanded at the rate of 5% every year while the fundamental and accessible data of every segment
number of inhabitants in the nation has expanded (for example a human, vehicle, and street condition).
just at the rate of 1.4% every year. Because of this, Mishap reproduction is characterized by Baker and
casualty chance, street inadvertent passings per Fricke as[11]"… the endeavors to decide from
100,000 individuals, has expanded from 7.9 in 2003 to whatever data is accessible, how the mishap
11.2 in 2013. Casualty hazard in India isn't just happened". Mishap reproduction approach works in
fourfold than that in a portion of the created nations, reverse from the proof of the accident examination
for example, the United Kingdom and Sweden yet in and the remaining parts of the collide with
addition as yet expanding quickly. It is likewise investigating the situation of previously (pre-crash),
discovered that the conveyance of street coincidental amid (crash) and after the accident (post-crash). The
passings and wounds shifts as per age, sexual successive investigation of final products to the
orientation, month and time. Among individuals of underlying state of the events can build up "how" and
all age gatherings, individuals of monetarily dynamic "why" a specific sort of accident happens. Arithmetic
age gathering of 30-59 years is the most powerless. and Newtonian material science are connected in this
Nonetheless, on the off chance that we analyze sexual investigation. It tends to be expressed that crash
orientation insightful fatalities and accidents, we recreation returns to explore the contributory
found that the guys represented 85.2% all things elements and additionally causes behind the accident
considered and 82.1% of all wounds in 2013. event dependent on major and minor physical pieces
of information abandoned at the accident scene.
Besides, street accidents are generally higher in
May-June and December-January, which The strategies of accident reproduction, direction,
demonstrates that extraordinary climate, impacts the and harm based investigation by utilizing material
event of street accidents. Accidents remain generally science streamline the assurance of numerous vital
steady and high amid 9 AM - 9 PM and variable yet parameters of accident events. In addition, to acquire
low amid mid-night and early hours of the day. In a dependable end, itemized data incorporating the
any case, this does not infer that daytime driving is framework segments should be completely explored.
more hazardous than evening time driving. The The data important for reproduction begins with the
investigation additionally attempts to discover the accident scene [12]. The responses to the inquiries of
reason shrewd conveyance of street accidents. 'why', 'what', 'when' and 'how' should lead the
recreation procedure to develop the genuine situation
There are a few variables in charge of accidents yet of the pre-crash, crash, and post-crash. Capturing of
the drivers' blame is observed to be the most essential imperative pieces of information and recording of the
one; drivers' blame represented 78% of all-out accident scene assumes an essential job in the
accidents in 2013. The recognizable proof of elements reproduction. Damage data from tenant restorative
influencing street crashes got from the accident reports can be checked with the direction of the
examination and remaking has not been directed by inhabitants found inside the included vehicles at the
and by in the Asian nations. The objective of this scene. In this way, a "receptive outlook" insightful
examination was to start this street wellbeing practice frame of mind is extremely urgent to scan for all the
in Thailand by tending to the auspicious requirement point-by-point data from the scene.
for a top to bottom investigation of street accidents.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


423
Prof. Tejal Irkhede et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 422-428

II. LITERATURE REVIEW occurrence recognition calculation dictated by


nonparametric relapse in [3].
Presently a-days, there is intense interest for cars,
as a result of this traffic control ends up boisterous The structure in [4] spare the message just as report
and it prompts street accidents. If there should arise with convergence utilizing metadata library. An
an occurrence of mishap, long reaction time to go to incident location strategy in street [5], make
the injured individual may prompts increment utilization of CCTV which regularly observe
number of death. In writing a few paper examine development of auto follow. Anyway this system
mishap recognition and vehicle following. In [2], an screens the primary focused on guests move that are
inventive remote disclosure using MEMS fitted with a great deal of circle openings inside it is
accelerometer and GPS following system is utilized execution. Because of the escape clauses inside going
for identification of accident.The structure can before is viable a couple of shiny new computerized
recognize kind of setback (immediate and nonlinear episode identification just as credit detailing
fall) from accelerometer sign using limit figuring, approaches wound up advanced. A large portion of
remove in the wake of crushing of GPS ground these gadgets is generally significantly centered
speedand cruiser. After incident is recognized, short proposed for some wheeler vehicles. This particular
ready data (alert back rub and position of setback) is computerized recognition just as credit announcing
sent by methods for GSM framework. The system is approach began working with GPS UNIT just as GSM
attempted in genuine 850 applications using bicycles. innovative expertise. Anyway these sort of wound up
The test results exhibit that it can recognize direct specifically built expected for cars. The real
fall, non-straight fall and ostensible ride with no mechanized episode location approach which
bogus alarm, remote discovery utilizing MEMS frequently utilizes air case sensor just as
accelerometer and GPS following for unplanned accelerometer to help analyze a mishap wound up set
checking of vehicle. The confinement of structure forward by M. Chuan-Zhietal [6]. Later gadgets that
proposed in [2] is mishap recognition for just bikes, have been advanced from this begun working with
and detailing just utilizing GPS with no black box. GPS unit innovation expected for occurrence
recognition alongside planned for deciding the
The paper [3] titled "The 3-arrange AcuTrac, setting in regards to episode.
Motorcycle Tracking System",Elite security supplies
outline workis constrained to bike on the grounds The framework arranged by Jerath just as Jung Lee
that a bikes setbacks could have a few cases which utilizes GPS unit innovation to help watch out for
can't absolutely use for four wheeler mishap speed with the auto just as microcontroller to help
discovery. The constraints of the above structure inspect speed with the auto planned for sequential
ismaking utilization of capacity of the GPS insignificant seconds [7][8]. Speedy braking will
framework just as a region sensor to influence an maybe result in a bogus helpful while speed
extremely decent spread technique to identify enormous contrast could be the basically angle
incidents. Various gadgets are accessible for viewed as expected for identifying episode. The
occurrence recognition by researchers. The real greater part of these strategies wound up made
episode location strategies wound up by deciding explicitly expected for a few wheelers [9].
realtime focused on vehicles and examination to help
expectation focused on vehicle. This particular The greater part of these gadgets isn't utilized in a
discovery approach is known while Traffic- few wheelers while they is likely not effective. In

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


424
Prof. Tejal Irkhede et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 422-428

this manner gadgets wound up intended for both the were included to ensure airbags deployed when they
cars.This is a stage for crisis salvage which will work were supposed to.
ideally so as to lessen the brilliant time of entry of
rescuers if there should arise an occurrence of street Over the years, as electronics got cheaper, smaller
accidents, when each microsecond checks. Our paper and smarter, event data recorders became capable of
plans to introduce an innovation consequently doing more than simply monitoring airbags.
identifying the mishap and an equipment GPS beacon Automakers realized the devices could be used to
dependent on GSM/GPS innovation illuminating at provide information about the seriousness of an
the event of mishap with adequate subtleties like accident, and if a car was being operated properly
accurate area and time at which mishap occurred and when a crash occurred. Based on a separate NHTSA
alongside that the voice was recorded for two or regulation passed in 2012, if a vehicle today does
three minutes when the mishap occurs.This paper have an event data recorder, it must track 15 specific
build up a correspondence between the control data points, including speed, steering, braking,
station and the unit introduced in vehicles. Vehicles acceleration, seatbelt use, and, in the event of a crash,
will have GPS/GSM empowered following modules force of impact and whether airbags deployed.
and will be followed progressively utilizing cell
systems. The product installed in the microcontroller Depending on the automaker and car model, an event
will control the different activities of the gadget by data recorder may capture many more functions,
observing waveform from the vibration sensor. In the though car companies are not required to disclose
event of mishap the gadget will send an alarm exactly what those are. The language many use to
message alongside area data from GPS module to explain black boxes in owner's manuals also is
control station utilizing GSM arrange. It is a purposely general to cover technology updates and to
thorough and viable answer for the poor salvage save space.
reaction in the event of mishap. The mishap detailing
can naturally discover an auto collision; scan for the Put everything the devices do in an owner's guide
spot and after that send the fundamental data to the and "instead of one paragraph, you'd have potentially
salvage office covering topographical directions and another 20 or 30 pages. That really wouldn't be
the time and conditions in which a car crash realistic," says Richard Ruth, a black box equipment
occurred. At the server end, a control capacity will trainer, expert witness and consultant who worked at
remove important data and store it in a database, to Ford Motor Co. for 33 years, including a stint
which mishap data from models will be surveyed evaluating event data recorders and other safety
progressively. Our framework joins propelled equipment. "It's not going to change whether or not
equipment plan and refined control innovation into a you're going to buy the car."
conservative, dependable bundle.
Most event data recorders are programmed to record
III. DECODING CAR'S BLACK BOX data in a continuous loop, writing over information
again and again until a vehicle is in a front-end
A. How Black Boxes Work collision or other crash. When an accident occurs,
Event data recorders are not actually black boxes but the device automatically saves up to 5 seconds of data
tiny microcomputer chip sets. In most vehicles, they from immediately before, during and after an
are part of the airbag control module, and originally incident.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


425
Prof. Tejal Irkhede et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 422-428

Today, practically every major automaker selling cars police and other law enforcement agencies that are
in the United States builds event data recorders into investigating an accident, insurance companies
new vehicles. The exceptions are Volkswagen (which looking into a claim, lawyers representing parties in
auto industry watchers say is preparing for the car-crash lawsuits and accident reconstruction
NHTSA regulation to kick in), Ferrari and Maserati. consultants working for any of the above.
Traffic accident analysis consultant Harris Technical
Services maintains a list of car makes and models In states with no black box laws on the books, "state
from 1994-2014 with event data recorders. troopers could get the data without a subpoena if
there was a fatality," says Tom Kowalick, a self-taught
The NHTSA rule, which the agency has been black box expert who chairs an event data recorder
working on for years, was supposed to take effect standards working group that's part of the Institute of
September 1, 2014. However, auto industry insiders Electrical and Electronics Engineers. Kowalick also
say the agency is still reviewing more than 1,000 wrote some of the black box information on the
comments it received about the proposed regulation, NHTSA Web site. "If they want to grab it, there's
making that implementation date unrealistic. A nobody saying they can't."
NHTSA representative declined to comment on the
delay. To rectify that situation, 15 states have passed EDR
B. Getting Black Box Data regulation over the past decade. Under the theory
Black box data is difficult and expensive to get to, and that car owners have privacy rights, many of the state
interpreting it takes special training. Extracting the laws require automakers to notify new-car buyers
data after an accident involves using a data-retrieval that vehicles contain black boxes, such as in the
tool kit that consists of hardware, software and a owner's manual. State laws also spell out the
cable that plugs into a car's on board diagnostics port. conditions under which police or other parties can
That is the same port mechanics use to identify obtain EDR information without an owner's consent,
engine problems and insurance companies tap as the such as with a court order; for dispatching emergency
basis for use-based insurance policies. Crash data personnel; diagnosing, servicing or repairing the
retrieval tool kits are not cheap, running $2,000- vehicle; or probable cause in an accident. The
$10,000 and up, not including training costs. National Council of State Legislatures maintains an
updated list of state EDR laws.
It follows that since drivers own their cars or trucks,
they own data the vehicles generate, including black Black boxes have become a battleground in states
box data. But because it's so difficult and costly to such as California, where earlier this year, insurance
extract, it's virtually impossible for average car companies and automakers lined up on opposite sides
owners to do it on their own — assuming that they of a black box data protection bill that would have
even want to. required automakers to let car owners block or opt
out of recording vehicle information. The bill didn't
Who else can access the information is a point of make it out of the state Senate Transportation
contention. Automakers would like the right to Committee after heavyweights including the Alliance
access the information for numerous reasons of Automobile Manufacturers opposed it.
including safety, to make sure systems work the way
they should and to check for defects. Other parties Earlier in 2014, two U.S. senators introduced a
that want a black box's car crash data can include bipartisan bill that would provide some of the same

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


426
Prof. Tejal Irkhede et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 422-428

protections on a national level. The Driver Privacy turned off. No one can hack into something without
Act explicitly states that a black box's data can't be Internet access." Even if someone could break in,
retrieved by anyone other than vehicle owners specifically in an effort to get data from an event data
without their consent and protects any personally recorder, the devices have access codes that need
identifiable information. By April 2014, the bill had authorized commands to work, he says.
collected 23 co-sponsors and been approved by the
Senate Commerce Committee. As of July 2014, In addition, car companies use threat modeling and
however, no further action had been taken. simulated attacks to test security and to help design
controls that protect data, says Wade Newton,
C. Black Boxes, Privacy and Security communications director for the Alliance of
Meanwhile, electronic privacy advocates worry about Automobile Manufacturers, an auto industry trade
a related car-data security issue: that a car's diagnostic group that represents 12 large car companies. "From
port, through which black box data streams, isn't bumper to bumper, automakers use proven security
secure enough to withstand hacking, and therefore techniques to help prevent unauthorized access to
poses a danger. Security experts and "white-hat" software," he says.
hackers already are testing how to break into the
ports to show how vulnerable they are. They are IV. CONCLUSION
publicly sharing the results, as in a video called "How
to Hack a Car." The analysis shows that the distribution of road
accidental deaths and injuries in India varies
Kowalick, a longtime black box data privacy advocate, according to age, gender, month and time. It is found
started a company to sell a diagnostic port lock that that the economically active age group is the most
he invented. The $30 AutoCyb lock, which he vulnerable population group. In general, males face
markets on his company Web site, is inserted into the higher fatality and accident risk than their female
diagnostics port to turn off access and prevent counterparts. Moreover, road accidents are relatively
unwelcome parties from getting to the data or higher in May-June and December-January which
interfering with car systems. "Every car in America shows that extreme weather influences the
can be hacked," he says. "The diagnostic link occurrence of road accidents. Accidents are relatively
connector is unsecure. All you have to do is set up constant and high during 9 AM to 9 PM and variable
access to the vehicle and have the right tool." but low during mid-night and early hours of the day.
There are several factors responsible for accidents but
However, Ruth, the EDR consultant and former Ford drivers’ fault is the most important factor; drivers’
executive, maintains a physical lock couldn't stop a fault accounted for 78% of total accidents, 76.5% of
black box expert or mechanic from bypassing the total injuries and 73.7% of total fatalities in 2013.
diagnostics port and obtaining the data another way. Automatic vehicle accident detection and reporting
He also dismisses the notion that hackers would be System using black box is is the need of the hour. The
interested in information. framework should be placed in moving vehicle to
detect accident and report to In Case of Emergency
"What would the incentive be?" he asks. "I think it's (ICE). The black box will records the voice of victim
an over-reaction. For all practical purposes, the after the accident occurs which will be used for
owner controls physical access. There's no Internet further investigation.
port on the car that's live, especially when the car's

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


427
Prof. Tejal Irkhede et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 422-428

Lee Conference on Communications (APCC),


V. REFERENCES 2012 18th Asia-Pacific pp.no: 978-1-4673-4726-6.
[10] Jog, S.R., Sutaone, M.S,Badawe, V.V.,
[1] Accidental Deaths & Suicides in India, 1970 to "Ruggedisation methodologies for GPS based
2013 published by the National Crime Records Vehicle Tracking System," (ICECT), 2011 3rd
Bureau, Ministry of Home Affairs, Government International Conference on , vol.2, no.,
of India, New Delhi. pp.214,218, 8-10 April 2011.
[2] Wireless Black Box using MEMS Accelerometer [11] Baker, J., and Fricke, L., Process of Traffic
and GPS Tracking for Accidental Monitoring of Accident Reconstruction. Traffic Accident
Vehicles, 978-1-4577-2177- 9/12$25 (c) 2012 Reconstruction (L. Fricke, ed.), Evanston, IL,
IEEE. Northwestern University Traffic Institute. (2010).
[3] “The 3-stage acuTrac Motorcycle Tracking [12] Van Kirk, D. J. Vehicular Accident
System”,Elite security supplies, Investigation and Reconstruction. CRC press.
http://www.gpsfast.com. USA. pp.15-19. (2001).
[4] "Traffic-incident detection-algorithm basedon
nonparametric regression”,S. M. Tang and H. J. Cite this article as :
Gao, IEEE Transactions on Intelligent
Transportation Systems, vol. 6, 2005, pp. 38-42. Prof. Tejal Irkhede, Rohit Kashayp, Nikhil Kumar, Vivek

[5] Yong-Kul Ki ,Jin-Woo Kim and Doo-Kwon Baik," Ranjan, Manish Kumar, "A Review on Car Collision
Avoidance System based on Interpretation of Vehicle
A Traffic Accident Detection Model using
Event Data Recorder ", International Journal of Scientific
Metadata Registry ",Conference on Software
Research in Computer Science, Engineering and
Engineering Research, Management and
Information Technology (IJSRCSEIT), ISSN : 2456-3307,
Applications, 2006 ppno: 0-7695-2656-X. Volume 5 Issue 1, pp. 422-428, January-February 2019.
[6] “Providing Accident Detection in Vehicular Journal URL : http://ijsrcseit.com/CSEIT1951107
Networks Through OBD-II Devices and Android-
based Smartphones”Juan Carlos Cano,Jorge
Zaldivar, Carlos T. Calafate, Pietro Manzoni in
5th IEEE Workshop On User Mobility and
Vehicular Networks 2011 pp no : 978-1-61284-
928-7.
[7] "Method of Freeway Incident Detection Using
wireless Positioning,"H. Ru-fu,L. Chuan-zhi, Y.E.
Hong-wu, in Proceedings of the IEEE
International Conference on Automation and
Logistics, 2008, pp. 2801 -2804.
[8] "GPS-free terrain-based vehicle tracking on road
networks," Jerath, K.; Brennan, S.N., American
Control Conference (ACC), 2012, pp.307, 311,
27-29 June 2012
[9] "An accident detection system on highway
through CCTV with calogero-mosersystem,Jung

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


428
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951108
A Review on Online Medical Drug Tracking
Swati Hedau1, Ankita Shegaonkar1, Amruta Aknurwar1, Devyani Vaidya1, Mehul Sahu1, Prof. Jogi John2
1 BE Students, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,
Maharashtra, India
2Assistant Professor, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,
Maharashtra, India

ABSTRACT

Recently, there is an enormous increment in online looking for changed wares like fabric, adornments and the
closeout of medicines have been expanded by means of the online/web. The online following of medicine is one
of the innovation headways that is going to make a colossal interest in the up and coming days. Despite the fact
that it is helpful to get them, there is a high danger of self-medicine and abuse of drugs, particularly that go
under calendar H and X. In this article, we have examined in insight concerning online pharmacy, its focal
points, and drawbacks contrast among online and neighborhood pharmacy, and certain measures to keep away
from the abuse of e-pharmacy.
Keywords : Drug Tracking, Online Pharmacies, Laws of Pharmacy, Prescription

I. INTRODUCTION history and afterward compose a remedy. This kind


of online pharmacy is all the more speaking to
Web or online drug stores move pharmaceuticals, shoppers
including solution and non-physician, recommended
meds. The online closeout of drugs began in the late since it offers both doctor administrations and drug
1990s and has broadened so extensively that the US specialist administrations to shoppers [3]. At last, the
Food and Drug Administration (FDA) has executed a third sort of online pharmacy is delegated a
whole segment on its site devoted to "BeSafeRx: "maverick" pharmacy or online drug shops which
Know Your Online Pharmacy" [1]. enables shoppers to buy remedy drugs without a
doctor's counsel and a substantial solution 3. Such
The online pharmacy industry can be arranged into pharmacy is viewed as unlawful by most law
three noteworthy sorts. The first and most genuine implementation organizations.
sort is the conventional online pharmacy. These
pharmacies just apportion medicine to a purchaser The World Health Organization faces this issue
who has first gotten a solution from a specialist and with regards to fake medicines, which are
after that submits it to the online pharmacy [2]. The characterized as 'a worldwide general wellbeing
second kind of online pharmacy is commonly known emergency', expressing that 'medicines bought over
as online meeting drug stores, remote counseling the Internet from destinations that cover their
drug stores, or recommending drug stores where physical location are fake in over half of cases [4].
doctors survey a patient's self-announced therapeutic

CSEIT1951108 | Received : 10 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 429-432 ]
429
Swati Hedau et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 429-432

There are numerous obvious dangers of purchasing (1) Legitimate Internet pharmacy sites giving
medicine over the Internet. These incorporate simple fantastic pharmacy administrations as indicated by
access to unlawful or unapproved drugs, medicine confirmation principles, and
offered without a solution, and drugs promoted with
phony wellbeing claims [5]. Additionally, it reroutes (2) Illegitimate online drug stores that are not
the conventional protections i.e., noteworthy confirmed and may not consent to national or global
cooperation with medicinal services experts (doctor expert benchmarks and directions.
as well as drug specialist) and subsequently putting
shoppers in danger [6]. Additionally, the greatest II. LITERATURE REVIEW
hazard that online drug stores present is to the
patient's wellbeing through fake and unacceptable A. Laws of Pharmacy
prescriptions [7]. According to Supreme Court statement, except in
emergency cases, doctor cannot consult through
Despite various potential dangers of buying meds phone. As far as India is concerned, there are no
online, there are a few evident advantages. Getting dedicated e-health and telemedicine laws in India.
medicines through the Internet is basic and There are no well-defined dedicated laws for online
advantageous as going forward and backward from a pharmacies. The laws applicable are:
traditional pharmacy had turned out to be extremely • Drug and Cosmetics Act 1940,
troublesome. Drugs can be requested whenever of • Drugs and Cosmetic Rules 1945,
day or on the other hand night and medicine could • Pharmacy Act 1948,
be dispatched by an online pharmacy medium-term • Indian Medical Act 1956 and
to his front entryway, thus sparing time and stress. • Laws related to e-commerce are defined under
What's more, utilizing online drug stores manages the Information Technology Act, 2000.
customers with more protection than the customary
specialist or pharmacy visit. Online drug stores are
especially engaging for those with ailments they wish B. Role Of Prescription
to keep private, as they don't require any immediate Online pharmacies will promote drug abuse, drug
human cooperation. misuse, self-medication, etc. Self-medication is a
rampant practice in India, and online sale of drugs
In past, the best-specialized technique for doctors would only encourage it.
in choices identifying with the drug for drug • Any medication taken without the supervision
specialists to disseminate medicines was written by of doctors may be dangerous and even
hand remedy. Presently it is being supplanted by potentially life-threatening.
electronic solution organize usually known as e- • A prescription issued by a doctor cannot be re-
pharmacy/online pharmacy. Online drug stores are used randomly. There is a danger that
organizations that move pharmaceutical scheduled drugs can be re-ordered and misused
arrangements that incorporate solution just drugs, on by the consumer.
the Internet by means of online requesting and mail • The D&C Act 1940, and the D&C Rules 1945
conveyance. Online drug stores can be arranged has guidelines on the sale of Schedule H and
fundamentally as: Schedule X drugs. These can be sold only on
prescription, and there are specific rules,
including for labeling and bar coding.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


430
Swati Hedau et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 429-432

Schedule X drugs include narcotics and pain relievers. Patients can take steps to ensure that
psychotropic substances. Chances of drug abuse and they use prescription medications appropriately by:
addiction are higher with the Schedule X category. • Never change a dosing regimen without first
Schedule X drugs require meticulous storage and discussing it with the doctor
dispensing records. The prescription has to be in • Never using another person’s prescription,
duplicate and one copy of which is to be retained by and never giving his or her prescription
the licensed pharmacist for 2 years. medications to others.
• Storing prescription stimulants, sedatives, and
C. Comparison Of Local Pharmacy And Online opioids safely patients should properly
Pharmacy discard unused or expired medications
Price: The prices for medications from an online Schedule H and schedule X drugs cannot be sold
pharmacy are less because there is no particular without prescription because it is easy for local
production unit, there is no physical store for the pharmacy stores to verify the authencity of a
medications storage and dispensing, and the company prescription, as it contains details like registration
has a large network of retailers to draw from, or a number of medical practitioner and name of the
combination of factors. hospital and there is no such checking by online
Convenience: Prescription can be submitted at any stores.
time, there is no need to rush to the pharmacy during • According to the report of high-level
business hours and do not have to wait in any long government the physicians should digitally
lines. However, there are also disadvantages. If it is sign committee since doctor’s prescription is
not the same website every time, one may not be mandatory for purchasing drugs, online
aware of any potential drug interactions, it can be pharmacies will need to mandatorily display
difficult to find someone to answer any specific the portal link on their homepage for
questions about medications. authencity verification by patients or
Local pharmacies are small independently owned customers and the prescriptions.
business rather than a pharmacy within one of the • In certain countries like United States,
large chains and they have staff during regular England, Sweden, Denmark and Finland have
business hours, and many also have an online accurate, regular, and systematic plans to use
counterpart as well. The large chains do use locals for electronic prescription system, and health
a large portion of their workforce. ministry of these countries was responsible
Pharmacist is responsible for dispensing the right for coordinating and leading the electronic
medicines and even counseling a patient about side- health.
effects and dosage. In the online space, where the • It is suggested to use experiences and
medicines are delivered at the patient’s home, there is programs of the leading countries to design
no possibility of an interaction between a patient and and develop the electronic prescription
a pharmacist. systems.
D. Measures To Be Taken To Avoid The Misuse Of
Online Pharmacy III. CONCLUSION
Before preventing the misuse of online pharmacy,
patients should stop the nonmedical use of Despite the fact that online pharmacy is more
prescription drugs like stimulants, sedatives or opiod subject than nearby pharmacy particularly in remote
regions, it ought to be planned to build up standards

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


431
Swati Hedau et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 429-432

and direction like transfer of filtered solutions. <http://www.fda.gov/NewsEvents/Newsroom/P


Patients ought to know on utilization of doctor ressAnnouncements/2004/ucm108232.htm>.
prescribed prescriptions. It is great to embrace the
plans and projects of driving nations to structure and Cite this article as :
build up the e-pharmacy framework.
Swati Hedau, Ankita Shegaonkar, Amruta Aknurwar,
IV. REFERENCES Devyani Vaidya, Mehul Sahu, Prof. Jogi John, "A
Review on Online Medical Drug Tracking",
[1] FDA. Food and Drugs Administration. Be- International Journal of Scientific Research in
SafeRx: Know Your Online Pharmacy. Computer Science, Engineering and Information
Available at Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
<http://www.fda.gov/Drugs/ResourcesForYou/ 5 Issue 1, pp. 429-432, January-February 2019.
Consumers/BuyingUsingMedicineSafely/Buyin Journal URL : http://ijsrcseit.com/CSEIT1951108
gMedicinesOvertheInternet/BeSafeRx-
KnowYourOnlinePharmacy/default.htm>.
[2] Fung, C.H., H.E. Woo & S.M. Asch (2014)
Mayo Clin Proc. 79: 188-94.
[3] Castronova, J.R. (2006) J Leg Med. 27: 207-24.
[4] The World Health Organization (2006)
Counterfeit medicines. Fact sheet revised 14
November 2006. Available at
<ttp://www.who.int/search/
semantic/en/#search=medicines%20purchased
%20over%20the%20Internet&sort=score%20de
sc&fq=%7B!noshow%3Dtrue%7Dlanguages%3
Aen>.
[5] Johnston, L.D., P.M. O’Malley, J.G. Bachman &
J.E. Schulenberg (2004) Monitoring the Future
national results on adolescent drug use:
Overview of key findings, 2003. Bethesda, MD:
National Institute on Drug Abuse. Available at
<http://www.monitoringthefuture.org/pubs/mo
nographs/overview2003.pdf>.
[6] Henney, J.E., J.E. Shuren, S.L. Nightingale &
T.J. McGinnis (1999) Ann. Intern. Med. 131:
861-2.
[7] US Food and Drug Administration (2004)
Recent FDA/U.S. Customs Import Blitz Exams
Continue to Reveal Potentially Dangerous
Illegally Imported Drug Shipments. January 27,
2004. Available at

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


432
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951110
A Survey on Secure User Data Transaction using Attribute-Based
Encryption Scheme over Cloud Data
Navin Sethiya1, Hrishikesh Patel1, Akshay Harshe1, Alekh Gaigole1, Harshvardhan Donadkar1, Prof. Priya
Karemore2
1UG Students, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,
Maharashtra, India.
2Assistant Professor, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,
Maharashtra, India.

ABSTRACT

Cloud computing is an innovation which utilizes the Internet to store, oversee, and register the information, as
opposed to a nearby server or a PC. Clients can access the cloud condition wherever they need on
compensation for each utilization way. Transferring of delicate information on un-believed servers in the cloud
is a testing assignment. Established encryption procedures are utilized when the re-appropriating of
information in the cloud to give access control and secrecy. Access control instruments are not doable in the
cloud as a result of the absence of adaptability, versatility, and fine-grained access control. As an option, So,
Attribute-Based Encryption (ABE) procedures are acquainted in the cloud with increment the security and
adaptability. Attribute-based encryption is an open key cryptographic strategy in which the private key and
figure content are based on the attributes. ABE involves IBE, KP-ABE, and CP-ABE. In Identity-Based
Encryption (IBE) the mystery key and the attributes are based on the character of the client. In Ciphertext-
Policy Attribute-Based Encryption (CP-ABE) a mystery key is gotten from the arrangement of attributes and
the access strategy is defined over the universe of attributes inside the framework. In Key Policy – Attribute-
Based Encryption (KP – ABE) the figure content is related to the arrangement of attributes and the mystery key
is based on the access structure. This paper is a study of all ABE strategies and specialty an evaluation table
about ABE keys job in cloud condition.
Keywords : Attribute-Based Encryption, Attributes Revocation, Fine-Grained Access Control, Keywords Search,
Mobile Cloud Storage.

I. INTRODUCTION different basic attributes like On-request access,


pervasive access, multitenancy, versatility, estimated
Cloud computing is a model for empowering utilization, strength. In On-request access, the client
pervasive, helpful, on-request organize access to a can utilize the computing assets in pay per use way.
common pool of configurable computing assets (Eg: Omnipresent access speaks to the capacity to access a
Network, Servers, Storage, applications and wide scope of gadgets, transport conventions,
administrations) that can be quickly provisioned and interfaces and different security innovations in cloud
discharged with negligible administration exertion or situations. Multitenancy empowers distinctive buyers
specialist organization interaction[1]. It comprises of to share the different cloud storage administrations.

CSEIT1951110 | Received : 10 Feb 2019 | Accepted : 21 Feb 2019 | January-February -2019 [ 5 (1) : 433-437 ]
433
Navin Sethiya et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 433-437

Versatility is the capacity of a cloud to direct scale IT 2. Integrity: It will keep unapproved clients from
assets as required by the cloud purchaser or cloud changing the information which is put away or
supplier. Estimated usage addresses the limitations of transmitted in the cloud.
a cloud stage to screen the utilization of its IT assets,
in a general sense by cloud buyers. Flexibility alludes 3. Availability: It enables the information to be
to excess IT assets inside a similar cloud or over accessible and usable amid an explicit timeframe.
various clouds. Three regular cloud conveyance
models are generally utilized and set up: 4. Scalability: Capability to deal with the clients
Infrastructure-as-a-Service (IaaS), Platform-as-a- included the system powerfully with no disturbance
Service (PaaS), Software-as-a-Service (SaaS). IaaS of administration.
conveyance demonstrates involved framework driven
IT assets that can be accessed and oversaw through 5. Client revocation: If the client leaves the system
cloud benefit based interfaces and instruments. PaaS the plan should disavow his access rights from the
relies upon the current condition that makes a lot of system straightforwardly. Something else, the client
apparatuses used to help the whole applications. SaaS can't utilize the information put away and the access
gives reusable cloud benefit generally accessible to is repudiated.
the scope of cloud buyers. Essentially, three sorts of
cloud arrangement models are the open cloud, 6. User revocation: Users can decode the encoded
private cloud, and third-party cloud. The open cloud information with their very own attributes since the
is claimed by an outsider cloud which is accessible by attributes are identified with polynomial arbitrary
everybody in the cloud condition. Private cloud is encryption work.
possessed by a solitary association where limits are
defined. The third party cloud is a mix of open and 7. Access Tree: Based on the attributes, the access tree
private cloud models. is produced. The access tree is included hubs. The leaf
hubs are spoken to as attributes and the middle of the
Cloud computing is a sort of distributive computing road hubs are spoken to utilizing doors likes AND,
that speaks to the usage models for remotely OR, and so on.
provisioning versatile and estimated assets [12]. At II. LITERATURE REVIEW
the point when the business information is moved to
the cloud, at that point the security issue will ascend, By expecting the believed cloud service provider
as information security ends up imparted to the cloud gives security additionally to the expansive measure
specialist organization. Fundamental wordings of delicate and significant information put away. ABE
utilized in the cloud will be: cloud supplier gives the calculations can be utilized for securing the
cloud-based IT assets. Cloud buyer utilizes the cloud- confidentiality of the put away information and
based IT assets. Cloud examiner assesses the security furthermore ABE gives the access control instrument
controls, protection effects, and execution. Central to information on the cloud. In cloud condition,
security terms important to cloud computing are: information confidentiality is imperative to shield
against insider assault, crash assault and refusal of
1. Confidentiality: It enables approved gatherings to administration assault. This area gives the current
access the information or assets that is it confines the attribute based encryption components in cloud
unapproved access to information in travel and condition.
storage.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


434
Navin Sethiya et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 433-437

ABE was presented by Sahai and Waters [2] in 2005. intended for one to numerous correspondences.
It is an open key based one to numerous encryption Decoding should be possible just when the attribute
that enables client to scramble and decode the set fulfills the client's access structure. It bolsters
information based on client attributes. The mystery access control conspire.
key and figure content are subject to the client
attributes. The decoding of ciphertext is conceivable Disadvantages:
just if the arrangement of attributes of client key Information proprietor can't choose who can decode
matches with the attributes of ciphertext. Decoding the encoded information. It can just pick graphic
should be possible just when the quantity of attributes for the information. It isn't appropriate for
coordinating keys is equivalent to the referenced broadcasting applications in light of the fact that the
limit level. ABE calculation comprise of four stages: information proprietor needs to confide in the key
Setup, key age, encryption and unscrambling. Crash backer.
obstruction is a pivotal highlights of ABE. A rival that
holds numerous keys can access the information if 3. Expressive Key Policy ABE (EKP-ABE)
the individual key matches. It is the all-inclusive variant of KP-ABE, the non-
monotonic access structures [5] are utilized. Non-
Drawbacks: monotonic access structure utilizes repudiated doors,
For encryption, the Data proprietor needs to utilize for example, NOT in the access structure. It is
each approved client's open key so it expands the increasingly adaptable to limit the unapproved use of
calculation overhead. This strategy is confined in information.
light of the fact that it utilizes access of monotonic
attributes to control client's access to the framework. Disadvantage:
The overhead is expanded on account of negative
1. Identity Based Encryption (IBE) attributes which isn't pertinent to the encoded
It was proposed by Shamir in 1984[3].In Identity information.
Based Encryption (IBE) the mystery key and the
attributes are based on the identity of the client. For 4. Ciphertext Policy ABE (CP-ABE)
instance, email id be the identity of the client. It is a turn around the form of KP-ABE [6]. In CP-
Believed third party will create the private key. ABE, the ciphertext is related with an access
structure and client's private key is based on set of
Drawback: attributes. A client can unscramble the ciphertext just
In the event that the believed third party gets traded whenever set of attributes related with clients private
off, there is no security for the framework. The client key fulfills the access policy related with the
protection can't be saved utilizing IBE. ciphertext. CP-ABE is more anchored even the
believed the third party is imperiled.
2. Key Policy Attribute Based Encryption (KP-ABE)
It was proposed by Goyal et al. [4]. Each client is Disadvantage:
assigned with some access tree over the arrangement
of attributes. Ciphertexts are based on the CP-ABE doesn't fulfilled undertaking needs of the
arrangement of attributes and the private key is based access control component. Decoding keys underpin
on the monotonic access structure that controls client attributes that are sorted out intelligently as a
which ciphertexts a client can decode. This is solitary set.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


435
Navin Sethiya et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 433-437

5. Hierarchical Attribute Based Encryption (HABE) The FHABE conspire is turned out to be secure under
The HABE was proposed by Wang et al. [7]. This the standard suspicion.
model comprises of a root ace that relates to the
confided in third party, different space ace in which 8. Ciphertext policy Weighted ABE (CPWABE)
top dimension area aces compares to various venture This plan assigns [10] load to the attributes based on
clients and various clients that relates to all work the inclination defined in the access control structure.
force in an undertaking. This plan client's Clients will have appointed the load to their
hierarchical age of keys. It utilizes randomized attributes. The information proprietor will encode
polynomial time calculation. It underpins anchored the information based on the attributes. The private
access control, adaptability and completely key is based on the weighted access structure and the
designation. It can share the ensured information for decryption [11] is conceivable just when the
clients in a cloud in an endeavor situation. It bolsters ciphertext and the weighted attribute matches with
intermediary reencryption conspire. the weighted access structure. It takes care of the key
Disadvantages: escrow issue by utilizing weighted attributes.
It is extremely hard to actualize continuously Drawback:
condition. A similar attribute will be utilized by Cost of computation is increased.
numerous area aces.
III. CONCLUSIONS
6. Multiple Authority ABE (MAABE)
This strategy utilizes numerous experts and the client This paper abridges different attribute-based
attributes are dispersed to different specialists. encryption systems for information security in cloud
Suppose [8] it comprises of N multiple attribute condition. Security will be the essential factor to be
experts and one focal specialist. The information can tended to. The storage cost of the figure writings,
be decoded by the client under attribute set an and time of encryption and unscrambling to be
unscrambling keys for an attribute set Au. It enables diminished. Based on the qualities like fine-grained
any free experts to disseminate private keys and to access control, adaptability and adaptability in cloud
deal with flawed or lost specialists. computing we can infer that FH-WABE performs
well superior to the next attribute encryption
Disadvantages: strategies.
Trouble in muti-expert plan is it expects specialist's to
keep up disjoint arrangements of attributes. IV. REFERENCES

7. File Hierarchy Attribute Based Encryption (FH- [1]. Muhammad Yasir Shabir, Asif Iqbal, Zahid
ABE) Mahmood and Ataullah Ghafoor, Analysis of
This strategy utilizes record progressive system based Classical Encryption Techniques in Cloud
encryption strategies. The documents are scrambled Computing, IEEE Tsinghua Science &
based on the layered access structure. Both figure Technology, Vol.21, February 2016.
content storage and time expenses of encryption are [2]. A.Sahai and B.Waters, Fuzzy identity based
saved [9]. encryption, in Proc. Advances in Cryptology –
Eurocrypt, 2005, pp.457-473.
Disadvantage: [3]. Adi Shamir. Identity-based cryptosystems and
signature schemes, in Proc. of CRYPTO 84 on

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


436
Navin Sethiya et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 433-437

Advances in cryptology, pages 47–53. Attribute based data sharing scheme revisited
SpringerVerlag New York, Inc., 1985. in cloud computing, IEEE Transactions on
[4]. V.Goyal, O.Pandey, A.Sahai and B.Waters, information forensics and security, Vol.11, No.8,
Attribute based encryption for fine grained August 2016.
access control of encrypted data, in proc. ACM [13]. G.K.Sandhia, G.K.Bhaskar, secure data sharing
conf. computer and communications, 2006. using attribute based encryption in cloud
[5]. J.Bethencourt, A.Sahai, and B.Waters, computing, Journal of Chemical &
Ciphertext policy attribute based encryption, in Pharmaceutical Sciences, Vol.9, December 2016.
proc. IEEE symposium security and privacy,
2007. Cite this article as :
[6]. K.Meena,S.Vinodhini, T.Pallavi & R.Vasugi,”
Surveillance Based Gcm Home Security System Navin Sethiya, Hrishikesh Patel, Akshay Harshe, Alekh
Gaigole, Harshvardhan Donadkar, Prof. Priya Karemore, "A
Using Object Motion Detection”,International
Survey on Secure User Data Transaction using Attribute -
Innovative Research Journal of Engineering Based Encryption Scheme over Cloud Data", International
and Technology,pp.44-47,2016. Journal of Scientific Research in Computer Science,
[7]. S.Rifki, Y.Park, and S.Moon, A fully secure Engineering and Information Technology (IJSRCSEIT),
ISSN : 2456-3307, Volume 5 Issue 1, pp. 433-437, January-
ciphertext policy attribute based encryption
February 2019.
with a tree-based access structure, Journal of Journal URL : http://ijsrcseit.com/CSEIT1951110
Information Science and Engineering, Vo.31,
pp.247-265,2015.
[8]. G.Wang, Q.Liu and J.Wu, Hierarchical
attribute based encryption for fine grained
access control in cloud storage services, in Proc.
ACM conf. computer and communication
security, 2010.
[9]. K.Yang, X.Jia, K.Ren and B.Zhang, Dac-Macs:
Effective data access control for multiauthority
cloud storage systems, in Proc. Of IEEE
Infocom, 2013.
[10]. Shulan Wang, Junwei Zhou, Joseph K.Liu,
Jianping Yu, Jianyong Chen and Weixin Xie,
An efficient file hierarchy attribute based
encryption scheme in cloud computing, IEEE
Transactions on information forensics and
security, Vol. 11, No.6, June 2016.
[11]. Ximeng Liu, Jiafeng Ma, Jinbo Xiong, Qi Li, Jun
Ma, Ciphertext – policy weighted attribute
based encryption for fine grained access control,
International conference on Intelligent
networkings and collaborative systems, 2013.
[12]. Shulan Wang, Kaitai Liang, Joseph K.Liu,
Jianyong Chen, Jianping Yu, Weixin Xie,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


437
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195164
Accessing Network Using Secure Android Applications
R. Gomathyjayam1, R. Thilakavathi2
1 Professor, Department of Computer Applications, Bon Secours College for Women, Thanjavur, Tamil Nadu,
India
2M.Sc., Department of Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India

ABSTRACT

Security plays a vital role in today’s mobile world. There are security issues like sniffing of data while accessing
information through open channel. Proper security measures can help to deal with the common security
threats faced by mobile phone users such as data protection, privacy, application and personal information
security. Cryptographic techniques play an important role in protecting communication links and data, since
access to data can be limited to those who hold the proper key. This paper discusses a method to securely access
information in a network by an android mobile application using AES cryptographic technique. The paper
describes a new key sharing algorithm, based on the symmetric key management, for faster and efficient
encryption of data that is suitable for use in a mobile device.
Keywords : Cognitive Radio Networks, Primary User, Secondary Users.

I. INTRODUCTION layer that directly deals with the transmission media


network support applications such as access to the
The telecommunications network allows computers World Wide Web, shared use of application and
to exchange data. The connections between nodes are storage servers, printers, and fax machines, and use of
established using either cable media or wireless email and instant messaging applications
media. Network computer devices that originate,
Route and terminate the data are called network II. LITERATURE SURVEY
nodes. Nodes can include hosts such as personal
computers, phones, servers as well as networking Reputation and Trust Management
hardware. Two such devices are said to be networked The reputation based frameworks are the nodes
together when one device is able to exchange maintain reputations of other nodes and evaluate
information with the other device, whether or not their trustworthiness are deployed to provide
they have a direct connection to each other. Scalable, Diverse and for countering different types
Computer networks differ in the transmission media of misbehavior resulting from malicious selfish nodes
used to carry their signals, the communications in the networks. The author (Jaydip et al.,2009) is
protocols to organize network traffic, the network’s proposed that Reputation systems are described along
size, topology and organizational intent. In most with their unique characteristics and working
cases, communications protocols are layered on (i.e. principles. Reputation and trust are very useful tools
work using) other more specific or more general that are used to facilitate decision making Reputation
communications protocols, except for the physical is the opinion of one entity about another it signifies

CSEIT195164 | Received : 08 Feb 2019 | Accepted : 21 Feb 2019 | January-February -2019 [ 5 (1) : 438-442 ]
438
R. Gomathyjayam, R. Thilakavathi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 438-442

the trustworthiness of an entity. Trust is the detection and energy based detection algorithms.
expectation of one entity about the action of another. Coherent sensor is an optimal linear detector for
Reputation and trust have been adapted to wireless known primary signals in presence of white Gaussian
communication Network it can solve many problems noise. However, detector implementation requires.
in the Networks. Demodulation of received signals for achieving the
optimal gains.
Combinationarial Optimization dentification
Algorithm (Coi) Agent-Based Trust Calculation
The COI (Combinationarial Optimization The author (yenumula B.Reddy et al., 2011) is
Identification Algorithm) to defend against such proposed agent based approach algorithm. The
attacks .Cooperative Spectrum sensing has been Cooperation in wireless sensor networks to detect the
shown good performance in improving the accuracy malicious node without any Infrastructure is a recent
of primary user detection .zhenquiri et al., 2013 trend in research. The current models need more
proposed a modified Design an attack model,called storage, Computation, security tools and
cooperative attack in which an attacker injects self- communication requirements. The fail wireless
consistent false data to multiple sensors sensor Network due to limitation of resources. The
simultaneously. A theorem that the center node of a proposed agent based approach eliminates the
cognitive radio network may face uncertainties under computations in the sensor nodes with appropriate
a cooperative attack, especially in the case when a trust factor. The proposed approach uses an agent
large portion of sensors are compromised. A modified based collaborative concept ensure the trust in the
COI algorithm to deal with cooperative attacks. Our successive node in the Path. The proposed agent
algorithm is a good scheme to complement IRIS for based framework uses reputation of neighboring node
cooperative attacks, and can be flexibly adjusted to as part of trust calculation in its successive node.
fulfill the detection delay requirement. Intensively
evaluate our algorithm through simulation, with the Sophisticated Detection Methods
results validating its performance. The original COI is The Spectrum scarcity is becoming a major issue for
an approach for identifying multiple instances of bad service providers interested in either deploying new
data in power system state estimation. The essential services or enhancing the capacity for existing
idea is to construct a partial decision tree using the applications. The author (Amir Ghasemi et al.,2007)is
branch-and-bound method to obtain a feasible proposed that Recent Measurements suggest that
solution with the minimum number of bad data. many portions of the licensed (primary) spectrum
remain unused for significant periods of time.
Clustered Based Algorithms Sensing-based access incurs a very low infrastructure
The spectrum sensing methods can be distributed in cost and is backward compatible with the legacy
three categories i.e. transmitter, receiver. (Dr. Talat primary systems.
Altaf et al.,2013) is propose cluster based algorithms
is based spectrum sensing and interference- III. EXISTING PROCESS
temperature based detection algorithms. From these,
transmitter detection based methods are a preferred The T-CSS (Traditional collaborative spectrum
way of sensing for presence/absence of spectral holes. sensing) protocol delays in data transmission and
These methods can be implemented through various unaware of secondary user selection for licensed
techniques including coherent detection, feature bandwidth and interference occur in unlicensed

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


439
R. Gomathyjayam, R. Thilakavathi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 438-442

secondary network, it lack in secondary power, not V. IMPLEMENTATION RESULTS


able to detect weak primary signals, in order to COGNITIVE RADIO NETWORK
protect primary receivers from interference. By Cognitive techniques have been used in wireless
proposing an energy efficient CSS protocol, namely networks to circumvent the limitations imposed by
energy efficient collaborative spectrum sensing EE- conventional WSNs. Cognitive radio (CR) is a
CSS protocol is used to transmit the data efficiently. candidate for the next generation of wireless
EE-CSS attempts to reduce the number of transmitted communications system. The cognitive technique is
reports from HSUs, based on the observation that the process of knowing through perception, planning,
HSUs agree on the spectrum usage more often than reasoning, acting, and continuously updating and
they disagree. CRN is to utilize the unused licensed upgrading with a history of learning. If cognitive
spectrum opportunistically. The SUs should protect radio can be integrated with wireless sensors. CR has
the accessing right of the PUs whenever necessary. the ability to know the unutilized spectrum in a
license and unlicensed spectrum band, and utilize the
IV. PROPOSED METHODOLOGY unused spectrum opportunistically. The incumbents
or primary users (PU) have the right to use the
The architecture consists of the following system spectrum anytime, whereas secondary users (SU) can
entities. utilize the spectrum only when the PU is not using it.
I.primary user: A user who has higher priority or
legacy rights on the usage of a specific part of the
spectrum. Primary user provides path or channel to
the secondary users
ii.secondary user: secondary user base
station(SUBS)transmit the data and reception is done
by honest secondary users(HSUs)secondary user

Figure 2. Cognitive radio Network

COLLABORATIVE SPECTRUM SENSING


TECHNIQUE

In this spectrum is used to detect spectrum band for


transformation and reception. The matched filter
detection technique requires a demodulation of the
PU's information signal, such as the modulation type
and order, pulse shaping, packet format, operating
Figure 1. Proposed System Architecture frequency, bandwidth, etc. CR Network sensing
iii. Honest secondary user: receive information from the PU's pilots, preambles,
The person who is receiving the data from the synchronization words or spreading codes etc in
secondary user base station is said to be honest figure 4.4. The advantage of the matched filter
secondary user. method is that it takes a short time and requires
fewer samples of the received signal. Sensing reports

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


440
R. Gomathyjayam, R. Thilakavathi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 438-442

provided by SUs for a given licensed band may differ


due to differences in channel fading gains, locations
of SUs and primary network transmitters, number of
signal energy quantization levels used at the sensing
SU, and sensing errors.

Figure 4. Cooperative Trust Management and avoid


malicious behavior

PRIORITY BASED SPECTRUM TRANSFORMATION

In Cognitive Radio network the users are classified


into Licensed Primary Users and Unlicensed
Secondary Users and there is no dedicated channel to
send data, sensors need to negotiate with the
neighbors and select a channel for data
communication in CR-WSNs in figure 4.6. This is a
Figure 3. Collaborative spectrum sensing technique very challenging issue, because there is no
cooperation between the PUs and SUs. PUs may
COOPERATIVE TRUST MANAGEMENT AND arrive on the channel any ime. If the PU claims the
AVOID MALICIOUS BEHAVIOR channel, the SUs have to leave the channel
immediately. CRN is implemented for short range
CR Network sensors may encounter incorrect wireless applications such as wireless sensor networks
judgments because radio-wave propagation through (WSNs) such wireless and Bluetooth, where the
the wireless channels has adverse factors, such as transmission distance is usually small (e.g., tens of
multi-path fading, shadowing, and building meters the steady-state average total number of
penetration. In addition, CR wireless sensors are sensing reports transmitted for each band And
hardware constraints and cannot sense multiple assume that the packets transmitted from the FC and
channels simultaneously. It has a malicious behavior SUs are of equal length in both EE-CSS and T-CSS.
to intermediate the signal spectrum. TRMSs record
the accuracy of previous sensing reports sent by SUs
and compute a trust value for each SU which is taken
as the trustworthiness of its future sensing reports.
And encounter the reports from SUs may be required
to mitigate against the effects of malicious behavior
of MSUs in figure 4.5. Therefore, CR wireless sensors
cooperate and share their sensing information with
each other to improve the sensing performance and Figure 5. Priority based spectrum Transformation
accuracy.
VI. CONCLUSION

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


441
R. Gomathyjayam, R. Thilakavathi Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 438-442

Cognitive Radio (CR) is an adaptive, intelligent radio VIII. REFERENCES


and network technology that can automatically
detect available channels in a wireless spectrum and [1]. A. Ghasemi and E. Sousa, "Collaborative spectrum
change transmission parameters enabling more sensing for opportunistic access in fading
communications to run concurrently and also environments," in Proc. IEEE DySPAN, 2005, pp.
improve radio operating behavior. Cognitive radio 131-136.
uses a number of technologies including Adaptive [2]. A. Ghasemi and E. Sousa, "Opportunistic spectrum
Radio (where the communications system monitors access in fading channelsthrough collaborative
and modifies its own performance) and Software sensing," J. Commun., vol. 2, no. 2, pp. 71-82, Mar.
2007.
Defined Radio (SDR) where traditional hardware
[3]. S.Haykin, "Cognitive radio: Brain-empowered
components including mixers, modulators and
wireless communications," IEEE J. Sel. Areas
amplifies have been replaced with intelligent
Commun., vol. 23, no. 2, pp. 201-220, Feb. 2005.
software. A spectrum sensing scheme, was proposed
[4]. J. Mitola and G. Maguire, "Cognitive radio: Making
to improve the utilization efficiency of the radio
software radios more personal," IEEE Pers.
spectrum by increasing detection reliability and
Commun., vol. 6, no. 4, pp. 13-18, Aug. 1999.
decreasing sensing time. The proposed scheme
[5]. E. Noon and H. Li, "Defending against hit-and-run
presented spectrum sensing in effective manner So attackers in collaborative spectrum sensing of
the priority based and security-based spectrum cognitive radio networks: A point system," in Proc.
sensing is produced. This system also implemented in IEEE VTC Spring, 2010, pp. 1-5.
hardware successfully. [6]. G. Staple and K. Werbach, "The end of spectrum
scarcity," IEEE Spectr., vol. 41, no. 3, pp. 48-52,
VII. FUTURE ENHANCEMENT Mar. 2004.
[7]. H. Yu, Z. Shen, C. Miao, C. Leung, and D. Niyato,
Wireless technology is proliferating rapidly, and the "A survey of trust and reputation management
vision of pervasive wireless computing and systems in wireless communications," Proc. IEEE,
communications offers the promise of many societal vol. 98, no. 10, pp. 1755-1772, Oct. 2010.
and individual benefits. While consumer devices [8]. T. Yucek and H. Arslan, "A survey of spectrum
such as cell phones, PDAs and laptops receive a lot of sensing algorithms for cognitive radio applications,"
attention, the impact of wireless technology is much IEEE Commun. Surveys Tuts., vol. 11, no. 1, pp.
broader, e.g., through sensor networks for safety 116-130, 2009.
[9]. W. Wang, H. Li, Y. Sun, and Z. Han, "Catchit:
applications and home automation, smart grid control,
Detect malicious nodes in collaborative spectrum
medical wearable and embedded wireless devices,
sensing," in Proc. IEEE GLOBECOM, 2009, pp. 1-6.
and entertainment systems. This explosion of wireless
Cite this article as : R. Gomathyjayam, R. Thilakavathi, "Accessing
applications creates an ever-increasing demand for Network Using Secure Android Applications", International Journal
more radio spectrum. However, most easily usable of Scientific Research in Computer Science, Engineering and
Information Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
spectrum bands have been allocated, although many
5 Issue 1, pp. 438-442, January-February 2019. Available at doi :
studies have shown that these bands are significantly https://doi.org/10.32628/CSEIT195164
underutilized. These considerations have motivated Journal URL : http://ijsrcseit.com/CSEIT195164

the search for breakthrough radio technologies that


can scale to meet future demands both in terms of
spectrum efficiency and application performance.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


442
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195135

Web Site Ranking Feedback System


Shaik Muzammil*, Sai Kiran Yerramaneni
Department of Computer Science and Engineering, Vasireddy venkatadri Institute of Technology, Guntur,
Andhra Pradesh, India

ABSTRACT

We all search on google for something and get the results in the form of different websites with some
description. We generally click on first or second website links if results are not found we go down on google
page. The website ranking is given by search engines by different criteria. Last website won’t be seen by none
of the people in most cases and first website will be having a great market compared to last website. So we need
to help the last website to be moved up in results and help in generating revenue and have good rank on
searching by giving feedback. This system will provide the difference between the first website and last website
on the google results and will provide the feedback to the last website like content, links, images used by first
website which helps the last website to be used in his webpage. This system is user friendly which is built on
HTML as front end and Python flask as back end and used python package beautiful soup to parse HTML data
and to automate browser behaviour with python. This system is done on web mining which has three
categories firstly web content mining in which we scan the web pages and get to know the links, text, images
used. Secondly web usage mining in which reports are generated after analysis which contain the details of text,
images, links. Finally the web structure mining states that structural summary of website.
Keywords : Beautiful Soup Package, Web Crawling, Feedback, Python- Flask, Searching Results, Web Mining,
Website

I. INTRODUCTION perform this task, search engines use different page


rank algorithms. Some ranking algorithms use
World Wide Web is a major source to retrieve content mining and some use web structure mining.
information. It has very large amount of data. For the But all the existing algorithms have some limitations.
user, it is a difficult task to find a particular page or a The system gives feedback to the results which was
set of pages from this massive amount of data. It is clicked by user here the first the user gives query and
highly impossible to remember the URLs of all these it was searched on google automatically and user
web pages. To solve this difficulty, search engine acts have to choose the web links to show the difference
as an intermediator between the users and WWW. between it and provide the feedback to last website
to improve his ranking. In typical Search engine
The search engines are able to extract the environment Web pages having matched keywords
information from the WWW for the users’ given are fetched from the predefined database and ranked.
query. They are used to collect the web pages and Wide range of relevant Web pages having unmatched
display them to the user. The order of the URLs keywords are omitted in this procedure. Neglected
depends upon the rank given to the pages. To Web pages having relevant information are covered

CSEIT195135 | Received : 01 Feb 2019 | Accepted : 20 Feb 2019 | January-February -2019 [ 5 (1) : 443-447 ]
443
Shaik Muzammil, Sai Kiran Yerramaneni Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 443-447

by the proposed method. Relevant Web pages are provided to webpage to increase its rank with respect
fetched and displayed irrespective of keyword to first page.
matching.
A. Implementation
Web Content Mining (WCM) concerns with retrieval
of information from WWW in a structured form and
thereby indexes the information for quick retrieval.
It is the process to extract useful information from
web documents contents. The documents consist of
images, audio, video, text files, structured records etc.
Figure 2 : Search bar to search queries
WCM technique is used on web documents and the
results page obtained from a search engine.
Here we can search for anything we get the weblinks
what the google has described in the google search.
II. METHODS AND MATERIAL

Web site ranking feedback system provides the user


to increase their rank in the search results and user
have to search the query as follows.

Figure 3. List of weblinks

Here we are get the list of weblinks which user


searched and here user can select the weblinks why
their link was least when compare to others.

The weblinks will be crawled and we will find the no


of links, images, page load time and found the
content of web page and keywords and find the social
networks present in the web page and some other
factors which help the web page to improve his
ranking and if the new one want to design a web
Figure 1 : workflows of system page he can use this system how the content should
be and how the factors to be taken in his web page to
User query is searched using google search engine have a better ranking in search results.
and extract the top n results from search query and
user selects the web links by checking the weblinks
and there system will count the following factor like
links, images, load time, content etc ,. Then
comparison will take place here the feedback is

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


444
Shaik Muzammil, Sai Kiran Yerramaneni Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 443-447

Rank factors plays a major role in ranking of webpage


here most of the users don’t know what is it? First of
all we need to scrap the webpage and then find page
load time and count the images and count the web
links for short urls and find the content it should be
minimum of 1800 words in a webpage and there
must be social connect in webpage which helps in
increase the page rank.
Figure 4 : Factors of different webpages

III. RESULTS AND DISCUSSION


Here the factors are calculated for every user checked
weblinks which helps the user why the least webpage
are having the rank with first page.

Algorithm

1. Input: user query


2. Display: Top n google results
3. user: choose the weblinks
4. calculate the rank factors
5. Display: Difference of webpages Figure 5. plotting the factors with webpages
6. Display: Feedback to least factor of
webpage
Here the graph is drawn with factors with the
Here same as google search the user enters query in different webpages which makes the user can
search bar and it display the top n results and user identify the drawbacks of the least webpage.
choose the weblinks to find why the rank is
difference and later it find the rank factors and We can easily identify the difference by using this
difference of webpages are made using the results of graph why the web page having least rank and how
rank factor and feedback is given to webpage whose to improve his ranking by using the first web page
rank factors are low. factors and he can have the good rating over the
webpage and have the better income.
Algorithm :
calculate the rank factors IV. CONCLUSION
1. Take the first link
2. scrap the webpage Web search engines are generally designed to provide
3. count the links services to all users without taking consideration of
4. count the images the user interests. Personalized web search includes
5. count the load time an individual user's interests when determining
6. repeat step 1 to second link relevant results to return. The main focus of project is
7. Difference of webpages using factors to increase the rank of last webpage of the google
result with the help of feedback given to the user

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


445
Shaik Muzammil, Sai Kiran Yerramaneni Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 443-447

checked query. Which helps the others who are International Conference on Advances in
designing the web page what to put in the webpage Computer Engineering, IEEE, 2010
and how the content resides on webpage. For the [8]. Frank McSherry "A Uniform Approach to
future work, consider the SEO tools and machine Accelerated PageRank Computation", ACM,
learning techniques for ranking the web page ranking 2005
not by factors.

V. REFERENCES Cite this article as :

Shaik Muzammil, Sai Kiran Yerramaneni, "Web Site


[1]. Gyanendra Kumar, Neelam Duhan, A. K.
Ranking Feedback System", International Journal of
Sharma "Page Ranking Based on Number of
Scientific Research in Computer Science, Engineering
Visits of Links of Web Page" 2011 2nd
and Information Technology (IJSRCSEIT), ISSN :
International Conference on Computer and
2456-3307, Volume 5 Issue 1, pp. 443-447, January-
Communication Technology (ICCCT-2011)
February 2019.
[2]. S. Geetharani, Dr M. Soranamageswari
Available at doi :
"Locationbased Ranking Method (LBRM) for
https://doi.org/10.32628/CSEIT195135
Ranking Search Results in Search Engines" at
Journal URL : http://ijsrcseit.com/CSEIT195135
2016 10th International Conference on
Intelligent Systems and Control (ISCO)
[3]. Nagappan V.K, Dr. P. Elango "Agent Based
Weighted Page Ranking Algorithm for Web
Content Information Retrieval"
[4]. Nandnee jain, Upendra Dwivedi "Ranking Web
Pages Based on User Interaction Time" at 2016
10th International Conference on Intelligent
Systems and Control (ISCO)
[5]. Tuhena Sen, Dev Kumar Chaudhary, Tanupriya
Choudhury "Modified Page Rank Algorithm:
Efficient Version of Simple Page Rank with
Time, Navigation and Synonym Factor" at 2015
International Conference on Advances in
Computing, Communications and Informatics
(ICACCI)
[6]. Ricardo Baeza-Yates and Emilio Davis ,"Web
page ranking using link attributes" , In
proceedings of 13 international World Wide
Web conference on Alternate track papers &
posters, PP.328-329, 2004.
[7]. Harmunish Taneja and Richa Gupta "Web
Information Retrieval using Query
Independent Page Rank Algorithm",

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


446
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951120

IoT Based Heart Attack Early Prediction


Shweta Gajbhiye , Bharati Vyas2, Shrushti Shrikhande 2, Anshika Janbandhu2, Komal Nagpure2,Mrunali Agashe2
1

1 Assistant Professor, Department of Computer Science and Engineering, Priyadarshini J. L. College of


Engineering, Nagpur, Maharashtra, India
2BE Students, Department of Computer Science and Engineering, Priyadarshini J. L. College of Engineering,
Nagpur, Maharashtra, India

ABSTRACT

Heart attack is most common disease that engulfs the patient’s precious life. This disease attacks a man so in a
flash that it barely gets treated. Many systems that detect heart attack came into the picture but they do have
many flaws. Some system does not work properly in cold weather and other doesn’t give accuracy. To have a
solution for Heart attack utilization of IoT and sensor system could be beneficial. Heart Attack Early Prediction
System predicts the Heart rate and accordingly sends the notification to the patient’s doctor and acquaintances.
Keywords : Heart Attack, IoT (Internet Of Things), Health Parameters, Heart Disease Prediction, Oximeter,
Heart rate sensor, Android App.

I. INTRODUCTION Monitors the Heart and even status logging and


social Sharing of recorded data.
Heart attack is most frequently occurring disease
that has a consequence of loosing a precious life. To overcome this health issue, an Android
Numerous people are mislaying their life owing to application could be designed and implemented.
heart attack and shortage of medical attention to The patient will be having the hardware device
patient at correct stage. There are many solutions with android application. The Heartbeat sensor will
that detects the heart beat like Fit Bit, Garmin and allow checking heartbeat reading and oxygen level
many more. Fit Bit and Garmin works like a smart and transmit them through Wi-Fi module.
watch. These devices track the users heart rate. Fit
Bit uses build in or connected GPS for many Heart Attack Early Prediction System will sense the
distance and tracking features. The device are Heart rate and oxygen level of the patient and will
categorized based on fitness level : Everyday, display on the Android Application. If the Heart
Active and Performance. Distributed computing rate extends the threshold value. Then the patient
helps the patient for monitoring the Health status will be considered in danger situation.
through albeit cell phones. Their is another Device
that is a laser device that monitors the Heart Beat, what is IoT?
Breathing rate and muscle activity from up to a
Internet of things simply “A network of internet
meter away with a single sensor called Zeev. Multi
connected objects able to collect and exchange
Detecting gadgets also came into the picture that
data”. It is commonly abbreviated as IOT.

CSEIT1951120 | Received : 12 Feb 2019 | Accepted : 24 Feb 2019 | January-February -2019 [ 5 (1) : 448-451 ]
448
Shweta Gajbhiye et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 448-451

IOT is the network of devices such as vehicles and et al. The models are trained and validated against a
home appliances that contain electronics, software, test dataset. Classification Matrix Methods are used
sensors, actuators, and connectivity that allows to evaluate the effectiveness of the models. Neural
these things to connect, interact and exchange data. Network is Difficult to understand. The system has
Embedded with the technology, these devices can 89.9 percent accuracy according to the Research[3].
communicate and interact through internet, and
they can be remotely monitored and controlled. Distributed computing has been incorporated in a
health care framework. The creators have
• Sensor Technology introduced a cloud-based keen healthcare
The capacity to recognize changes in the physical monitoring framework (CIHMS) for giving
status of things is additionally fundamental for restorative input to a patient through the cloud.
chronicle changes in the earth. In such manner, Albeit cell phones are constantly viewed as a
sensors assume an urgent job in overcoming any promising instrument to screen and deal with the
issues between the physical and virtual universes, patient's own health status, these gadgets have some
and empowering things to react to changes in their innate restrictions in calculation or information
physical condition. Sensors gather information concentrated undertakings. ECG monitoring system
from their condition, producing data and bringing contains WI-FI Bluetooth and Zigbee. ECG signs
issues to light about setting. For instance, sensors in were gathered and were transmitted to the IOT
an electronic coat can gather data about changes in cloud utilizing WI-FI. This information were send
outer temperature and the parameters of the coat to the family of the patient [4].
can be balanced in like manner.
Remote health care framework for monitoring
II. REVIEW OF LITERATURE electrocardiographic and temperature information
has been introduced in [5]. The framework
Milan Kumari et al. fathom cardiovascular disease comprises of three modules, in particular, an
dataset utilizing diverse information mining equipment module, Bluetooth module, and show
calculations, for example, Support Vector Machine, module. The equipment module is utilized for
Artificial neural systems (ANNs), Decision Tree, and information securing. The Bluetooth module is
RIPPER classifier. The creators break down the utilized for information transmission. At long last,
execution of these calculations through a few the information is shown by utilizing the
factual investigation factors, for example, presentation module. The obtained clinical
affectability, specificity, exactness, mistake rate. A information is sent to a database server by utilizing
choice emotionally supportive network for GPRS or Wi-Fi.
conclusion of Congenital Heart. Using data mining
and artificial neural networks ,the system predicts The exhibitions of the framework have been tried
Heart disease with nearly 100% accuracy. [1] on various patients and it typical arrhythmia
database.
Niti Guru et al. [2] connected a neural system for
forecast of heart disease, circulatory strain and has been discovered that the introduced framework
sugar. A model Intelligent Heart Disease Prediction is extremely useful for the doctors[5].
System (IHDPS) in view of information mining
procedures is introduced by Sellappan Palaniappan

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


449
Shweta Gajbhiye et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 448-451

A walking framework for monitoring the physical Heart Attack Early Prediction System could contain
restoration patients has been accounted for in [6]. Arduino board and a Wi-Fi module. The Heart Beat
The framework comprises of a multi-sensor-based sensor can begin detecting heart rate readings. If the
monitoring gadget, a cell phone with customer Heart beat will be above 120 bpm then the patient
application, an administration situated engineering will be in the danger situation. If the oxygen level
based server arrangement, and an application. The will be below 80 bpm then again, the patient will be
framework has been tried in a controlled condition considered in the dangerous situation. Likewise,
comprising of some healthy volunteers and some with the utilization of Wi-Fi module it will transmit
congestive heart disappointment patients. The test the information over web. The framework will
outcomes demonstrate that the introduced begin monitoring the heart rate of patient and
framework can identify and screen congestive heart instantly the heart rate goes above or beneath as far
disappointment and it can send criticism to the as possible the framework will send notification on
attendants for patient development [6]. the Android Application.

Ongoing universal health care framework for notification in the Android Application. An android
monitoring ECG motions by utilizing cell phone has application will track the heartbeat of specific
been exhibited in [7]. By utilizing this framework, patient and screen it effectively and give the crisis
the client can screen his ECG flag. The creators message on odds of Heart Attack.
have introduced a calculation for unusual heartbeat
detection and irregular heartbeat check delineate). III. CONCLUSION
The execution of the introduced framework has
been assessed against the MIT-BIH It has It has Heart disease is one of the major cause that leads to
been accounted for that the framework can death and early prediction of heart disease is very
distinguish at a top with a win rate of 97.8 percent important for every one. Garmin is a smart watch
and it is additionally ready to recognize strange and it predicts the heart attack but it does not work
heartbeat condition with a win rate of 78.9 percent. accurately in cold weather. Fit Bit is another
An unavoidable healthcare framework empowering alternative but even it has many flaws i.e. while the
self-administration for constant the introduced tracker is charging, patient cannot move and it is
framework in [8] comprises of patient health expensive too.Heart Attack Early Prediction System
monitoring framework, status logging, and social has many pros Over the existing solutions. It takes
sharing of the recorded data. The framework has two inputs i.e. Pulse rate and oxygen level of the
been executed by a cell phone, a wearable multi- patient to have the accuracy. It transmits the
detecting gadget contains pulse sensor and Notifications to the patients doctor and
temperature sensor, an administration situated acquaintances.
engineering for correspondence, and smaller scale
blogging administrations. The framework has been IV. REFERENCES
tried on 16 patients. The test outcomes demonstrate
that the introduced framework is anything but [1]. Milan Kumari, Sunila Godara, Comparative
difficult to learn and advantageous to use by endless Study of Data Mining Classification Methods
patients. in Cardiovascular Disease Prediction, IJCST
Vol. 2, Issue 2, June 2011.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


450
Shweta Gajbhiye et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 448-451

[2]. Niti Guru, Anil Dahiya, Navin Rajpal, Cite this article as :
Decision Support System for Heart Disease
Diagnosis Using Neural Network, Delhi Shweta Gajbhiye, Bharati Vyas, Shrushti
Business Review, Vol. 8, No. 1, January-June Shrikhande, Anshika Janbandhu, Komal Nagpure,
2007. Mrunali Agashe, "IoT Based Heart Attack Early
[3]. Sellappan Palaniappan, Rafiah Awang, Prediction", International Journal of Scientific
Intelligent Heart Disease Prediction System Research in Computer Science, Engineering and
Using Data Mining Technique, 978-1-4244- Information Technology (IJSRCSEIT), ISSN : 2456-
1968-5/08/25.00 2008 IEEE. 3307, Volume 5 Issue 1, pp. 448-451, January-
[4]. http://www.nlm.nih.gov/medlineplus/magazi February 2019.
ne/issues/winter11 Journal URL : http://ijsrcseit.com/CSEIT1951120
[5]. Bourouis, A., Feham, M., and Bouchachia,
A.(2011), Ubiquitous Mobile Health
Monitoring System for Elderly (UMHMSE),
International Journal of Computer Science
and Information Technology, Vol.2, No. 3,
June, pp. 74-82
[6]. Yuce, M. R.(2010) Implementation of
wireless body area networks for healthcare
systems, Sensor and Actuators A:Physical,
Vol. 162, No. 1, July, pp. 116-129
[7]. Lei Clifton, David A. Clifton, Marco A. F.
Pimentel, Peter J. Watkinson, and Lionel
Tarassenko (2014), Predictive Monitoring of
Mobile Patients by Combining Clinical
Observations with Data From Wearable
Sensors, IEEE Journal of Biomedical and
Health Informatics, Vol. 18, No. 3, May , pp.
722-730
[8]. Parane, K.A., Patil, N.C. ; Poojara, S.R. ;
Kamble, T.S(2014) Cloud based Intelligent
Healthcare Monitoring System, In the
proceedings of International Conference on
Issues and Challenges in Intelligent
Computing Techniques (ICICT), February 7-
8, Ghaziabad, Indian, pp. 697-701

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


451
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951119
Shakti : An Android App for Mapping of Crime against Women
Diksha Ingle, Pritee Dhale, Priyanka Palsapure, Sagar Pazare, Dhawal Chimurkar, Niket Farfad, Prof. Prasanna Lohe
Department of Computer Science, JD College of Engineering and Management, Nagpur, Maharashtra, India

ABSTRACT
Women biggest issue we are facing today. Many apps have been developed to provide safety to women through
Android phones. Many of the times government conduct different awareness events for stopping crime against
women. But that's not enough; there are various kinds of crimes women are facing today. Now it's time to aid
the roots of these crimes. This paper presents Shakti, an Android Application for mapping of crime against
women. This will help the government to identify which crime has the highest frequency in a particular area
and can implement an appropriate campaign for aiding its root cause. In case of an emergency, the victim can
send its location along with a help message just by clicking the volume button 5 times. Also, the victim can
reach-out to NGO for help. Shakti will feel like having family power with you at any time.
Keywords : Android, Women, Crime, Mapping, Location, Help message, NGO.

I. INTRODUCTION that the government runs for helping the victims are
appreciable. How about not letting any women be a
From past years, it has been recorded that the graph victim? For this, a study is needed for every area and
of crime against women has risen. Though the list of the type of crime occurring most frequently.
NGO's, Women Empowerment campaigns, safety
applications, and more steps for the safety of women "Dowry" can be aided with a literate society.
has continuously risen with time, the crime graph is "Domestic Violence" needs a little work on the
still incrementing. There are various government mindset of the family; in-laws. "Rape" and "Sexual
bodies as well as private organizations working for Assault" can be stopped if the females are equipped
women empowerment and helping the victims. But with right safety tools, and self-defense training.
the call of today is to find the root cause of these
crimes and treat it accordingly. Rape, Sexual assault, A. Existing Systems:
Domestic violence, Dowry, these are major of the As a part of a literature survey, we investigated some
crime list. Everyone is talking about empowering applications that offer the same or similar services for
women, blaming society. What is making a society Android and other platforms. The aim is to see how
guilty? Is its people with narrow thinking and evil these applications work and to see how they can be
minds, and these are the causes of illiteracy, lack of improved.
life values, in short- the mindset of people.
The existing women safety apps:
In everyday Newspaper, some areas are in headlines Women Safety: This app is offered by Awesome Apps
for "Rape" and Sexual "Assault cases". Some areas are Inc. Released on 28th Aug 2015. With just a tap of a
in the headline for "Domestic Violence". The projects button, the app sends an e-mail to preconfigured

CSEIT1951119 | Received : 12 Feb 2019 | Accepted : 24 Feb 2019 | January-February -2019 [ 5 (1) : 452-455 ]
452
Diksha Ingle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 452-455

email ID(s) along with your location and a link to Limitation : You to keep this women safety app
Google Map. The app also captures 2 pics, one with always running in the background which can drain
front camera and the other with the back camera, the battery of your phone. It does not show every
and a video or audio clip and uploads it to our server. area of information. It gets crashed in cases.
A link to the pic, audio or video is emailed to a
preconfigured email Id(s). redEye - Women Safety: This is a very useful app for
women safety. It does not need any additional action
Limitation: It is not useful in case of emergency. Also, by the user in any emergency situations. The app
there is an issue in managing phone numbers. Also, itself let your dear ones know that you are in danger
the panic message has no notification, so no use if the through an alert alarm. In this app, the user will
other sees it later. choose the source and the destination. The redEye
app then keeps tracking you. If the app finds that the
Raksha, A Women's Safety App: This app is offered user's path has deviated or the user is idle in the same
by Portal Perfect, released on 27th May 2018. This place for a configured duration, then the app will
app understands the worries and problems of dear trigger the alert. Once you reach your destination
ones, be it your child, mother, daughter, sister or safely you can manually disable the app.
wife. By using this application you can check their
current location. The special feature of this app is Limitation: This app only allows 2500 uses per day by
"Are you in trouble" button. all users. More than 5000 people have downloaded
the app, so only half or less than half users are able to
Limitation: The emergency messages are not able to use the app.
send. It also failed working many times.
Here, we introduce an android app that ensures the
My Safetipin: Personal Safety & Women Safety App. safety of women. It reduces the risk and helps us in
This app gives safety scores to an area and helps you need by identifying the location of a person who is in
to get alerts when you enter in an unsafe area. It danger.
helps in finding the safest route for you. You can also
choose the alternate safe location when you give II. PROPOSED SYSTEM
Google maps the access of this women safety app.
SafetiPin collects primary data on Women Safety in For the need of surveying various areas and
public spaces at night, on a number of parameters. implementing appropriate programs in that area to
• Lighting diminish the roots of crime against women, we have
• Visibility: Are there ‘eyes on the street'. Can developed an android application SHAKTI. In today's
others see you world android phone is a basic thing which everyone
• Diversity: Presence of women and children carries, with SHAKTI in your android phone women
around now will feel secure and protected all the time.
• Crowd: How densely crowded is the area Following are the features of our Android application:
• Public Transport: How close it is The user will rate a type of crime for their particular
• Walk Path: Is there a proper place to walk area. This data will be graphically represented on our
• Security: Are police or private security nearby. website.
• Openness: Can you see all around
• Feeling: How safe do you feel

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


453
Diksha Ingle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 452-455

In cases when the victim is unable to report against (2)Server Side: Data entered by the users (members
crime to the police, the victim can complain to the registered, crime reported, etc.) via the app is
social NGO through this application. uploaded to the web server, the crime details for
The user will save three contacts of three family various areas will be graphically represented on the
members. Whenever the user feels any danger like website and complaint is sent via text message to
crime coming its way, the user can press volume key registered NGO for further processing and
of its Android phone five times and an emergency authenticity. Also, admins get all the privileges
message for help will be delivered to their family through this server.
members along with the location URL through GPS.
Family members will also be notified if the victim's BLOCK DIAGRAM
phone battery is below 20%.
Shakti: An Android app for mapping of crime against
women is implemented into two modules: (1) Client-
side, (2) Server Side.

(1)Client Side: Client-side involves the interaction of


the user. The user can register in the application
through their mobile number. Once registered they
can log in any time. After successful login user first
needs to enter its area and its pin code, select the
frequently happening crime from the drop-down list Fig: 1. Depicts the block diagram of server side
and then rate the crime. In this way, the data about
various areas will be collected from the user. III. IMPLEMENTATION

Next, the users need to enter three family members Shakti android application is being implemented
contact number with an emergency message. In the using the following technologies: (1) Android (2) Java
situation of any crime, the user will have to press the (3) SQL (4) Cloud (5) Web Technologies (6) Database
volume button five times. Saved emergency message (7) GPS
will be forwarded to the saved contact numbers along
with the victim's location through GPS. (1)Android: Android is the most widely used mobile
OS motorized by Linux kernel. This is the first
The user can also ask for help to the NGO directly complimentary platform which is robust and is
through our application. expected to gain much popularity. It was developed
by Google team and allows writing managed code in
In cases if the user’s mobile battery drains below 20%, the Java language. The current system is developed
the app will send an alert message to the dear ones. on the basis of android platform. Android utilizes a
Also user can set a secret code in the app. When the custom virtual machine that was designed to
dear ones want to reach the user, they will send this optimize memory and hardware resources in a smart
secret code as a text message and they will phone. Android does not differentiate between the
automatically get the users mobile location through phone’s core applications and third-party applications.
GPS . Any application that is built will definitely have
equal access to a phone’s capabilities providing users

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


454
Diksha Ingle et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 452-455

with a broad spectrum of applications and services (2) different areas. Thus, the social bodies working for
Java: Java is an object-oriented, class-based, the security of women can implement an aid to those
concurrent, secured and general-purpose computer- areas to halt the crimes.
programming language.
V. ACKNOWLEDGEMENT
After front-end elements are finalized java classes
and packages are used to make them work. First, and foremost we would like to thank God for
the wonderful opportunities and challenges he has
(3) SQL: Structured Query Language or SQL is a given to us. We express our sincere gratitude to our
standard Database language which is used to create, guide Prof. Prasanna Lohe, Head, Department of
maintain and retrieve the relational database. It is Computer Science and Engineering for his valuable
particularly used to work with structured data where contribution in fulfilling the requirement related to
there is a relation associated within the data itself. the project and for providing his valuable guidance,
(4) Cloud: Cloud storage involves stashing data on patience and for encouraging us to do our best.
hardware in a remote physical location, which can be
accessed from any device via the internet. Clients Cite this article as :
send files to a data server maintained by
a cloud provider instead of (or as well as) storing it on Diksha Ingle, Pritee Dhale, Priyanka Palsapure, Sagar
their own hard drives. Pazare, Dhawal Chimurkar, Niket Farfad, Prof.
(5) Web Technologies: The data collected in this Prasanna Lohe, "Shakti : An Android App for
android women safety application we be displayed Mapping of Crime against Women", International
using website. Journal of Scientific Research in Computer Science,
(6) Database: A database is an organized collection of Engineering and Information Technology
data, generally stored and accessed electronically (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
from a computer system where databases are more 452-455, January-February 2019.
complex they are often develop using formal design Journal URL : http://ijsrcseit.com/CSEIT1951119
and modeling techniques.
Android comes with in-built SQLite database. SQLite
supports all the relational database features.
(7) GPS: GPS or Global Positioning System is a
satellite navigation system that furnishes location and
time information in all climate conditions to the user.
GPS is used for navigation in planes, ships, cars and
trucks also. The system gives critical abilities to
military and civilian users around the globe. GPS
provides continuous real time, 3-dimensional
positioning, navigation and timing worldwide.

IV. CONCLUSION

Shakti will help to easily know the major crimes


against women and violence happening under

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


455
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951123

Robust Video Data Hiding in Forbidden Zone


Ruchika S. Bhambore1, Ashwini B. Gurudeo1, Shital B. Borkar1, Pallavi G. Thul1, Chanchala V. Udepurkar1, Prof. Anup
Bhange2
1BE, Department of Computer Technology, KDK College of Engineering, Nagpur, Maharashtra, India

2Assistant Professor, Department of Computer Technology, KDK College of Engineering, Nagpur, Maharashtra,

India
ABSTRACT

Video data hiding is yet a vital research subject because of the structure complexities included. As of late, there
are loads of frameworks are present. The general population designed a substantial thing to ensure the data and
there are bunches of hiding procedures are to be developed for security reason. In any case, that strategies can
be hack by unapproved clients is downside in existing frameworks so that propose the new framework for
example Data hiding behind the video utilizing forbidden zone and selective embedding. We propose another
video data hiding technique that makes utilization of eradication amendment ability of Repeat Accumulate
codes and prevalence of Forbidden Zone Data Hiding. This framework makes utilization of redress capacity of
duplication store codes and favorable position of forbidden zone data hiding is utilized. In this examination the
encryption and decoding method is utilized to give the security key. Without that key, nobody can see the first
data. This system is utilized to shield the database from unapproved and the dangerous powers .It has
substantial deletion ability of data hiding.
Keywords: Data Hiding, Forbidden Zone, Quantization Index Modulation [QIM], Repeat Accumulate Codes,
Selective Embedding

I. INTRODUCTION
Consequently, video data hiding keeps on
Data hiding is the way toward embedding data into establishing a functioning examination region. Data
a host medium. When all is said in done, visual and hiding in video arrangements is performed in two
real media are favored because of their wide nearness noteworthy ways: bitstream-level and data-level. In
and the resilience of human perceptual frameworks bitstream-level, the redundancies inside the present
included. pressure measures are misused. Ordinarily, encoders
have different alternatives amid encoding and this
In spite of the fact that the general structure of the opportunity of choice is appropriate for control with
data hiding process does not rely upon the host media the point of data hiding. In any case, these strategies
type, the strategies change contingent upon the idea exceptionally depend on the structure of the
of such media. For example, picture and video data bitstream; henceforth, they are very delicate, as in
hiding share numerous regular focuses; be that as it much of the time they can't endure any organization
may, video data hiding requires progressively transformation or transcoding, even with no
complex structures because of the extra fleeting noteworthy loss of perceptual quality. Subsequently,
measurement. this sort of data hiding techniques is by and large

CSEIT1951123 | Received : 15 Feb 2019 | Accepted : 25 Feb 2019 | January-February -2019 [ 5 (1) : 456-461 ]
446
Ruchika S. Bhambore. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 456-461

proposed for delicate applications, for example, utilize the BCH code to expand blunder amendment
verification. capacity. The creators perform 3D interleaving so as
to dispose of the nearby burst of mistakes. Also, they
Then again, data-level techniques are increasingly propose a worldly synchronization procedure to
strong to assaults. Subsequently, they are reasonable adapt to fleeting assaults, for example, outline drop,
for a more extensive scope of utilization. addition, and rehash. In this paper, we propose
Notwithstanding their delicacy, the bitstream-based another square based selective embedding type data
techniques are as yet alluring for data hiding hiding structure that exemplifies Forbidden Zone
applications. For example, the excess in square size Data Hiding (FZDH) and RA codes as per an extra
choice of H.264 encoding is misused for hiding data. fleeting synchronization instrument. FZDH is a down
In another methodology, the quantization parameter to earth data hiding strategy, which it appeared to be
and DCT (Discrete Cosine Transform) coefficients are better than the traditional Quantization Index
changed in the bitstream-level. In any case, the vast Modulation
majority of the video data hiding techniques use
uncompressed video data. Sarkar proposes a high (QIM) .RA codes are as of now utilized in picture
volume change area data hiding in MPEG-2 videos. and video data hiding because of their vigor against
They apply QIM to low-recurrence DCT coefficients eradication. Vigor permits taking care of
and adjust the quantization parameter dependent on desynchronization among embedder and decoder
MPEG-2 parameters. that happens because of the distinctions in the chose
coefficients. So as to join outline synchronization
Besides, they differ the embedding rate contingent markers, we segment the squares into two gatherings.
upon the sort of the casing. Therefore, inclusions and One gathering is utilized for edge marker embedding
eradications happen at the decoder, which causes and the other is utilized for message bits.
desynchronization. They use Repeat Accumulate (RA)
codes so as to withstand deletions. Since they adjust By methods for basic standards connected to the
the parameters as per the sort of casing, each edge is edge markers, we present a specific dimension of
handled independently RA codes are as of now power against casing drop, rehash, and creepy crawly
connected in picture data hiding. Versatile square assaults. We use precise RA codes to encode message
choice outcomes in desynchronization and they use bits and edge marker bits.
RA codes to deal with eradication. Additions and
deletions can be additionally taken care of by Each piece is related to a square living in a gathering
convolutional codes. The creators use convolutional of casings. Irregular interleaving is performed
codes at embedder. In any case, the weight is put on spatiotemporally; thus, reliance on nearby attributes
the decoder. Different parallel Viterbi decoders are is diminished. Host flag coefficients utilized for data
utilized to address desynchronization mistakes. In hiding are chosen at four phases. In the first place,
any case, it is seen that such a plan is effective when outline determination is performed. Casings with an
the quantity of chose host flag tests is significantly adequate number of squares are chosen. Next, just
less than the absolute number of host flag tests. 3-D some foreordained low-recurrence DCT coefficients
DWT space is utilized to conceal data. They use LL are allowed to conceal data. At that point, the normal
subband coefficients and don't play out any versatile vitality of the square is relied upon to be more
determination. In this way, they don't utilize blunder prominent than a foreordained limit. In the last stage,
redress codes strong to eradication. Rather, they the vitality of every coefficient is contrasted with

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


457
Ruchika S. Bhambore. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 456-461

another edge. The unselected squares are marked as non-covering, which can be accomplished by
eradication and they are not handled. For each utilizing a base quantizer and moving its recreation
chosen square, there exists a variable number of focuses relying upon m, like Dither Modulation. A
coefficients. These coefficients are utilized to install run of the mill embedding capacity that utilizes a
and unravel a solitary message bit by utilizing the uniform quantizer. we need to consider that the In
multi-dimensional type of FZDH that utilizes cubic the unrivaled field, the hiding advancement, for
cross-section as its base quantizer. example, least critical bit(LSB) extra, is done in the
predominant field, while change space techniques;
II. LITERATURE REVIEW conceal data in another area, for example, wavelet
space. Least noteworthy piece (LSB) is the
The writing review is the most critical advance in the unassuming type of Steganography. LSB depends on
product improvement process. Before building up the infusing data in the littlest huge piece of pixels,
device it is important to decide the time factor, which data to a minor change on the spread picture,
economy n organization quality. When these things r which is not noticeable to the human eye. Since this
fulfilled, ten following stage is to figure out which strategy can be effectively part, it is more fragile to
working framework and language can be utilized for assaults. LSB framework effectively affects the
building up the instrument. When the software arithmetical data of picture like a histogram.
engineers begin assembling the instrument, the
developers need a great deal of outside help. This Safeguards could be the alarm of a shrouded
help can be acquired from senior software engineers, correspondence by simply testing the Histogram of a
from a book or from sites. Before building the picture. A decent answer for reject this deformity
framework the above thought r considered for was LSB indistinguishable. LSB Identical was an
building up the proposed framework. Forbidden incredible advance forward in Steganography
Zone Data Hiding (FZDH) is presented in [8]. The strategies and numerous others get plans from it.
technique relies upon the Forbidden Zone (FZ) idea, Downsides of Existing System are that the strategy
which is characterized as the host flag go where no given in Existing framework is effectively broken.
change is permitted amid data hiding process. The measure of the put-away data is little.

FZDH makes utilization of FZ to alter the power III. PROPOSED SYSTEM


intangibility exchange off the mapping capacity in (2)
expresses that the host flag is adjusted by including While there are various programming improvement
an extra term, which is a scaled adaptation of the models the idea of our task has disentangled the
quantization distinction. In 1-D, this extra term is determination procedure. Programming
scalar, though in N-D have flag is moved along the advancement procedures, for example, the cascade
quantization contrast vector and towards the and V display have been dismissed on account of
remaking purpose of the quantizer. Consequently, their inflexible structures, selecting a lithe
embedding contortion is diminished and ended up programming improvement approach which is
littler than the quantization mistake. increasingly responsive and iterative in its tendency.

So as to satisfy the prerequisite of common Building up a steganography system that is fit for
prohibition, the reproduction purposes of the hiding data in a video le lies at the core of this
quantizers that are recorded by various m ought to be undertaking, and it is alluring that the resultant

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


458
Ruchika S. Bhambore. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 456-461

arrangement is as secure as could be expected under We talk about parts of our underlying examination,
the circumstances. Lithe plan strategies enable us to which included a wide examination of steganography
assess and enhance our executed systems as we (utilizing sound, picture and video compartments).
advance. Moreover, our overview of the writing and The understanding given by this exploration in
our starter explore (see A.2) yielded incredibly uenced the plan choices that pursue.
constrained data on the best way to plan and execute
a video steganography arrangement of this nature. In Steganography and Steganalysis are in a ceaseless
light of this, coordinated advancement would enable fight. At whatever point a decent steganography
us to roll out considerable improvements to the system is built up, another Steganalysis method is
structure and execution anytime something that can likewise created endeavouring to overcome it
only with significant effort be accomplished with Steganalysis is the workmanship and exploration of
other programming designing models. identifying mystery messages shrouded utilizing
steganography. Once there is proof that a message is
concealed, the objective of steganography is crushed
regardless of whether the message was not separated.
In spite of the fact that steganography methods may
clearly straightforward to the human eyes, assaults on
them are yet conceivable. Any embedding procedure
definitely leaves follows in the stego-item and
modifies a portion of its properties, which present
irregular attributes and some debasement as far as
quality. Consequently, Steganalysis can be arranged
into two classes: inactive Steganalysis and dynamic
Steganalysis. Aloof Steganalysis distinguishes the
nearness or nonappearance of shrouded message or
recognizes the embedding calculation utilized. While
dynamic Steganalysis change, extricate or obliterate
the concealed message or concentrate a portion of its
traits, for example, message length.

It is at this crossroads that we might want to bring up In the first step, frame selection is completed and the
that our exploration has been considerably more test a selection of frames are managed block-wise. For
than we could have predicted. Our decision of a each block, only a first bit is hidden. After obtaining
coordinated improvement process has been a 88 DCT of the block, energy check is performed on
demonstrated gift on various events. While most of the coefficients that are predefined in a mask. Selected
this section will report the structure and execution of proficient of variable length are used to hide data bit
our last arrangement, we will likewise feature how m. m is a supporter of message bits or frame
our plan and usage have changed with the emphases. synchronization markers. Message sequence of each
Being in contact at the last phase of usage has been an group is obtained by using RA encryptions for T
unpredictable procedure and a few essential exercises consecutive frames. Each block is assigned to one of
have been learned en route. these groups at the start. After the inverse convert
host frame is obtained. Decoder is the dual of the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


459
Ruchika S. Bhambore. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 456-461

embedded, with the exception that frame selection is E. Coefficients Selection


not performed. Energy of each constant is compared to another
beginning T2. If the energy is above T2, then it is used
Marked frames are detected by using frame during data inserting together with other selected
synchronization markers. Decoder employs the same constants the same block.
system parameters and concludes the marked signal
values that will be fed to data insertion step. Not- F. Block Partitioning
selected blocks are handled as erasures. Erasures and Two split data sets are inserted; message bits (m1) and
decoded message data possibilities (om) are passed to frame synchronization markers (m2). The block
RA decoder for sequential frames as a entire and then locations of m2 are resolute randomly depending on a
the secret data is decrypted. random key. The rest of the blocks are kept for m1.
A. Selective Embedding The same splitting is used for all frames. m2 is
Host signal models, which will be used in data hiding, inserted frame by frame. Continuously the other
are single-minded adaptively. The selection is hand, m1 is single to T sequential frames. Both of
performed at four stages: border variation, frequency them are found as the outcomes of the RA encoder.
band willpower, block selection, and coefficient
selection. Of data. The proposed system architecture G. Erasure Handling
is shown in below figure. Due to adaptive block collection, de-synchronization
occurs between embedded and translator. Because of
attacks or even embedding operation, decoder may
not perfectly determine the selected blocks at the
embedded. In order to overcome this problem,
mistake modification codes strong to erasures, such as
RA codes are used in video data hiding in previous
hard work. RA code is a low complication turbo-like
code. It is collected of repetition code, interleave, and
a convolutional encoder. The source bits (u) are
repetitive R times and accidentally permuted
B. Frame Selection dependent on a key. The interleaved sequence is
A number of numbers of blocks in the whole frame is passed through a convolutional encoder with a
calculated. If the percentage of selected blocks to all transfer function 1/ (1 + D), where D represents a
blocks is above a certain value (T0) the frame is _rst-order stay. In efficient RA code, input is 16
handled. Otherwise, this frame is avoided. placed at the start of the output as shown in Fig. 1.

C. Frequency Band In this paper, we apply logical RA codes to find m1 as


Only certain DCT constants are operated. Middle u1+v1 and m2 as u2+v2. Here, u1 denotes the
frequency band of DCT constants. encrypted message bits and u2 is the encrypted frame
management marker bits. RA code is decrypted using
D. Block Selection sum-invention algorithm. We apply the message-
Energy of the constants in the mask is added. If the passing algorithm given in.
vitality of the block is overhead a assured value (T1)
then the block is managed. Otherwise, it is avoided.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


460
Ruchika S. Bhambore. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 456-461

IV. CONCLUSION V. REFERENCES

In this paper, we proposed another video data [1] Ersin Esen, A. Aydin Alatan,\Robust Video Data
hiding structure that makes utilization of deletion Hiding Using Forbidden Zone Data Hiding And
redressability of RA codes and prevalence of FZDH. Selective Embedding ",IEEE ,VOL. 21, NO. 8,
The strategy is likewise powerful to outline control AUGUST 2011.
assaults through casing synchronization markers. To [2] K.Mohan, S.E.Neelakandan \Secured Robust
begin with, we thought about FZDH and QIM as the Video Data Hiding Using Symmetric Encryption
data hiding strategy for the proposed structure. We Algorithms ", IJIRE ,VOL.6,DECEMBER 2012
saw that FZDH is better than QIM, particularly for [3] R. Ravi Kumar V., Kesav Kumar \Selective
low embedding mutilation levels. The structure was Embedding and Forbidden Zone Data Hiding for
tried with MPEG-2, H.264 pressure, scaling and Strong Video Data Thrashing ",IJETT ,VOL.
casing rate transformation assaults. Ordinary 4,SEPTEMBER 2013
framework parameters are accounted for mistake-free [4] Mr.Sudheer Adepu, Mr.P. Ashok , Dr.C.V.Guru
translating. The outcomes show that the structure Rao \A Security Mechanism for Video Data
can be effectively used in video data hiding hiding " ,IJCTT,VOL.4, August 2013
applications. For example, Tardos fingerprinting, [5] Resoju Omprakash and D. Jyothi \Block Based
which is a randomized development of double unique Adaptive Videodata Hiding Technique", IJMSTH,
mark codes that are ideal against agreement assault, 2012
can be utilized inside the proposed structure with the [6] Mr. Mritha Ramalingam \Stego Machine Video
accompanying settings. The length of the Tardos Steganography using Modified LSB Algorithm ",
unique mark is AC2 0 In 1 ε1, where A will be an World Academy of Science, Engineering and
element of false positive likelihood (ε1), false Technology,2011
negative likelihood, and the greatest size of the [7] W. Bender D. Gruhl,N. Morimoto,A. Lu,
colluder alliance, (Co). We likewise analyzed the \Techniques for data hiding ", IBM SYSTEMS
proposed structure against the standard JOURNAL, VOL.35, NOS 3 and 4, 1996
watermarking technique, JAWS, and a later
quantization-based strategy. The outcomes Cite this article as : Ruchika S. Bhambore, Ashwini B.
demonstrate a huge predominance over JAWS and a Gurudeo, Shital B. Borkar, Pallavi G. Thul, Chanchala
V. Udepurkar, Prof. Anup Bhange, "Robust Video
practically identical execution with. The analyses
Data Hiding in Forbidden Zone", International
likewise revealed insight into conceivable
Journal of Scientific Research in Computer Science,
enhancements for the proposed technique. To begin Engineering and Information Technology
with, the system includes various edges (T0, T1, and (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
T2), which are resolved physically. 456-461, January-February 2019.
Journal URL : http://ijsrcseit.com/CSEIT1951123
The scope of these limits can be examined by
utilizing a preparation set. At that point, a few
heuristics can be derived for the legitimate choice of
these edge esteems. Moreover, joining of the human
visual framework based spatiotemporally adjustment
of data hiding technique parameters as in stays as a
future bearing.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


461
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/IJSRCSEIT

The Internet-of-Things (IoT) Security : A Technological


Perspective and Review
Dr. Yusuf Perwej*1, Firoj Parwej2, Mumdouh Mirghani Mohamed Hassan3, Nikhat Akhtar4
*1Assistant Professor, Department of Information Technology, Al Baha University, Al Baha, Kingdom of Saudi Arabia
(KSA)
2Research Scholar, Department of Computer Science & Engineering, Singhania University, Jhunjhunu, Rajasthan, India
3Assistant Professor, Department of Computer Science, Al Baha University, Al Baha, Kingdom of Saudi Arabia (KSA)
4Research Scholar-Ph.D, Department of Computer Science & Engineering, Babu Banarasi Das University, Lucknow, India

ABSTRACT

Recent years have seen the swift development and deployment of Internet-of-Things (IoT) applications in a
variety of application domains. In this scenario, people worldwide are now ready to delight the benefits of the
Internet of Things (IoT). The IoT is emerging as the third wave in the evolution of the Internet. The 1990s’
Internet wave connected 1.2 billion subscribers while the 2000s’ mobile wave connected another 2.4 billion.
Actually, IoT is expected to consist of more than 84 billion connected devices generating 186 zettabyte of
data by 2025, in the opinion of IDC. It includes major types of networks, such as distributed, ubiquitous, grid,
and vehicular, these have conquered the world of information technology over a decade. IoT is growing fast
across several industry verticals along with increases in the number of interconnected devices and diversify of
IoT applications. In spite of the fact that, IoT technologies are not reaching maturity yet and there are many
challenges to overcome. The Internet of Things combines actual and virtual anywhere and anytime, fascinate
the attention of both constructor and hacker. Necessarily, leaving the devices without human interference for a
long period could lead to theft and IoT incorporates many such things. In this paper, we are briefly discussing
technological perspective of Internet of Things security. Because, the protection was a major concern when just
two devices were coupled. In this context, security is the most significant of them. Today scenario, there are
millions of connected devices and billions of sensors and their numbers are growing. All of them are expected
secure and reliable connectivity. Consequently, companies and organizations adopting IoT technologies require
well-designed security IoT architectures.
Keywords : Internet of Things (IoT), Sybil Attack, IoT Authentication, IoT Trust, IoT Protocols, IoT Security

I. INTRODUCTION connected to common Internet using tiny computing


devices with communication technology. At present,
The size of computer systems has diminished our world includes billions of computing devices and
drastically over the years, from mainframes sensors that are continually sensing, collecting,
encompassing whole rooms, via desktop computers, integrate [2], and analyzing significant amount of our
and down to smart cell phones [1]. At the turn of the personal information. This would permit anything to
century, a new concept appear called the Internet of speak to everything, making everyday life trouble-
Things, envisioning all “things” in the world free for everybody. The massive use case for the

CSEIT195193 | Received : 12 Feb 2019 | Accepted : 25 Feb 2019 | January-February -2019 [ 5 (1) : 462-482 ]
462
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

Internet of Things today is consolidation of data, and exposed to attacks by malicious opponent [1]. An
responding to the collected data in a useful way [3]. invasion may compromise an IoT device and use it as
While connecting all our things to the Internet will a platform for launching invasion on other IoT
permit us to advantage insight into our lives and devices. IoT security introduces technological
environment, we can potentially permit others to defiance at the device, network and platform level.
advantage the same insight if security is not managed Therewith, there is the process challenge of
correctly. A couple of security anxiety on a single organizing the security technologies in an end-to-end
device like as a mobile phone can swiftly turn to 60 manner. One thing is certain, when evaluating
or 70 anxiety when considering multiple IoT devices security necessity is that, then IoT is still very much a
in an interconnected business or home. In light of the work in progress. This paper provides an overview of
significance of what IoT devices have access to, it is Internet of Things (IoT) stage in section 2. We are
an essential to understand their security peril [1]. briefly discussing Internet of Things (IoT) threats in
section 3, and in section 4, we describe Internet of
The Internet of Things (IoT) [4] enables everywhere Things scheme. We categorize the reliance in section
communication between various devices. From 5. In section 6, we discuss public key infrastructures
entering patient details to watch post-surgery, from in Internet of Things. In section, 7 and 8 Internet of
parking vehicles to tracking vehicles, from childcare Things design thought for digital certificates and
to elder care, from smart cards to near field cards, protocols. In the last section discuss guidelines for
sensors are making their impendence felt. Sensors secure the IoT devices.
play an important role in the IoT as well. In spite of,
the functionality and operations of the IoT heavily II. INTERNET OF THINGS (IoT) STAGE
depend on the underlying network connectivity
structure. In the opinion of Gartner [5], it is required The IoT requires five stages, from data collection to
that the number of Internet connected devices will data delivery to the end [1] users on or off demand, as
increase from around 28 billion to 50 billion by 2020. shown in figure 1.
In spite of the IoT features everywhere
communication [6] among all kinds of electronic
devices, it inevitably raises security concerns due to
seamless infiltration and automated integration
among all sorts of applications. The IoT works across
miscellaneous networks and standards. Security and
privacy are considered the most important IoT
challenges [7]. In particular, no network is free from
security threats and vulnerabilities. Each of the IoT
layers is uncovered to various types of threats. In the
opinion of Gartner [8] security and risk, concerns
will continue to be the greatest hindrance to IoT
adoption. The market for IoT distinguished security Figure 1. The Internet of Things (IoT) Stage
solutions will dramatically expand in 2017 as current
security providers aggressively retool present In the upstanding stage, sensors collect data from the
capabilities to address IoT security risks IoT devices environment or object under measurement and turn
have often limited resources and may be more it into useful data [9]. Actuators can also intervene to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


463
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

modify the physical conditions that generate the data. security problems. In the future, maybe around the
The thing may be a constant body (body sensors or year 2020 with IPv6 and the 5G network, billions of
RFID tags) or a dynamic vehicle (sensors and chips) miscellaneous things will be part of the IoT. Privacy
[10]. The actuating and sensing stage involucre and security will be the major factors of worry at that
everything from legacy industrial devices to robotic time [12]. IoT security threats and attacks will
camera systems, water-level detectors, heart rate escalate as the IoT devices become everyday events.
monitors, accelerometers, and air quality sensors. In The security threat is high enough for Gartner to
the secondly data collected on stage first should be estimate spending on IoT security is required to reach
stored and making it obtainable for analysis. If the near $600 million in 2018. In this [13] report,
thing has its own local memory, data can be stored prophecy that 30% of attacks in enterprises will
[7]. The different technologies are appropriate for include IoT. In this section, we are discussing
this purpose, depending on the scenario. The high security problems in three dimensions, components,
performance storage solutions are suggested for based on stage, and architecture.
companies performing data analysis with the
distributed system Hadoop. Normally, IoT 3.1 The Components Based Incursion
components are installed with low memory and low
processing potential [1]. The cloud takes over the As a technology, IoT is peerless since it has a role to
answerability for storing the data in the case of play in consumer, industrial worlds and enterprise.
stateless devices. In the thirdly stage, IoT analyzes The things are diverse in nature, communicating
the data stored in the cloud DCs and provides sensitive data over a distance [13]. The Internet of
intelligent services for work and life in hard actual Things (IoT) is extremely heterogeneous, highly
time as well as analyzing and responding to queries, dynamic, always available, [7] and consequently
then IoT also handle things. The IoT proposal for always vulnerable to attack. Apart from attenuation,
intelligent processing and control services to all stealing, loss, violation, and disaster, data can also be
things homogeneously. In the fourth, data [1] concocted and modified by compromised sensors.
transmission happen in all stage for instance sensors, The figure 2 shows the types of attacks at the
RFID tags [10], processors to controllers, devices, DCs component level in IoT.
to processing units, and end users. In the fifth stage,
delivery of processed data to things on time without
errors or transformation is a sensitive task that must
always accomplish.

III. INTERNET OF THINGS (IoT) THREATS

The IoT devices have several applications that are


designed to make life convenient and effortless.
Think of engineers being able to access a device,
perform remote diagnosis and remediation any
problem. This is after the device has informed the
engineering team of an imminent problem before it Figure 2. The Types of Attacks at the Component
becomes a major problem [11]. As you can predict, Level in Internet of Things (IoT)
with this data exchange over the internet come

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


464
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

3.2 The Stage Based Incursion needed to make sure connected devices on the IoT
can be trusted to be what they intend to be.
The diversify kind of attacks on the five stages of IoT. Accordingly, each IoT device needs a distinctive
Data effusion, sovereignty, [14] violation, and identity that can be authenticated when the device
authentication are the major concerns in the data effort to connect to a main server or gateway. With
perception phase. this distinctive identity in place [15], IT system
administrators can track every device throughout its
3.2.1 Data Effusion lifecycle, communicate securely with it, and inhibit it
from executing detrimental processes. If a device
The data is a precious resource, and if recent events shows unforeseen behavior, administrators can
surrounding the allegations against (for instance, simply revoke its perquisite [16]. In addition, it is
Cambridge Analytica) demonstrate anything, it is compulsory to verify that the data do not change
that you cannot always trust companies to manage during transit. Data authentication could endow
such data responsibly. Specifically, device originality and integrity.
manufacturers are in a prime position to potentially
abuse [14] IoT generating data. Data effusion can be 3.2.4 Data Mislay
internal or external, intentional or unintentional,
authorized or spiteful, involving hardware or Data mislay considered a primus risk for the Internet
software. Data effusion is a solemn threat to of Things. The data mislay dissimilar from data
reliability. As the cloud data move from one renter to effusion in that the latter is a sort of revenge taking
several other renter of the cloud, there is a sedate risk action on the employer or administrator. Data mislay
of data effusion. is losing the work accident due to hardware or
software, lack of success and natural calamity.
3.2.2 Data Sovereignty
3.2.5 Attack on Availability
Internet of Things (IoT) connections growing from 9
billion in 2017 to 24 billion by 2020, it is vital to The Internet of Things continues to increase, which
understand the potential influence data sovereignty in turn spread your organization’s attack surface.
will play for organizations collecting and sharing Many types of attacks have been on every side for a
customer’s individual information. International very long time. What is new is the scale and relative
organizations are now often moving big data across naivety of attacks in the Internet of Things the
borders for analysis and consolidation [2]. Data millions of devices that are a potential sufferer to
sovereignty is the idea that information which has traditional style cyber-attacks, but on a much huge
been transformed and stored in binary digital form is scale and often with limited, if any shield [17]. IoT is
subject to the laws of the country in which it is all about connecting and networking devices that up
located. The IoT encloses all things across the globe until now not on a mandatory basis been connected.
and is hence responsible to sovereignty. This means that all of those devices, whether it is
your new connected sensor or your connected
3.2.3 Data Authentication vehicle, are creating a new entry point in the
network and here upon posing an increasing security
The data can be perceived from any device at any and privacy hazard. In this context, denial of service
moment. The powerful IoT device authentication is (DoS) attack happens when a service that would

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


465
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

usually work is not available. There can be many of vicinal gets formed and these new vicinal [19] are
factors for unavailability, but it generally refers to an all from other end of wormhole tunnel and not in
infrastructure that cannot cope due to capacity transmission range of node therefore, during the
overload [18]. In a Distributed Denial of Service attack lots of control packets are going to exchange
(DDoS) attack, a huge number of systems spitefully from one end of the tunnel to the other in that
attack one target. This is often done via a botnet, vicinal advertisement. Wormhole attack is very
where many devices are programmed to entreaty a strange and arduous to identify. The figure 3 shows
service at the same time. the wormhole attack on IoT.

3.3 The Architecture Based Incursion

The IoT will generate data at various locations for


various end users, including the enterprise, its
subscriber and partners, network segmentation and
segment-based topologies are expected to protect
against extensively attacks. The various vendors and
applications adopt their own layers. Now, we are
discussing the possible threats to each layer in IoT. Figure 3. The Wormhole Attack in Internet of
Things
3.3.1 Exterior Attack
3.3.3 Brute-Force Attack
The IoT is a bit of a buzzword right now. It is a used
as a catch all for everything that layers physical A brute force attack is the cyber-attack compeer of
devices like as computing infrastructure, sensors, trying every key on your key ring, and in conclusion
networking, storage, application [14] capabilities. The finding the right one. This type of attack depends on
organizations purposely offload both confidential and a trial and error technique in order to get information
non-confidential data to obtain the services. like as user passwords or personal identification
However, they do not know of the location where number (PIN). The brute force attack uses [20]
their data will be processed or stored. It is possible automated software to generate a large number of
that the provider may share this information with sequential guesses to decrypt the cipher text. The
others, or the provider itself may use it for spiteful criminals to crack encrypted data may use brute force
actions. attacks.

3.3.2 Wormhole Attack 3.3.4 Man-in-the-Middle Attack

The wormhole attacks considered as the grievous The man-in-the-middle concept is where an attacker
attacks during IoT routing. In this attack, tunnel is or hacker is looking to intervene and violation
established between two nodes and the packet is communications between two distinct systems. It can
forwarded among each other. These distant spiteful be a hazardous attack because it is one where the
nodes make believe that they are very close to each attacker secretly intercepts [21] and transmits
other so that vicinal nodes forward packets via them. messages between two parties when they are under
If wormhole attack is triggered in the more number

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


466
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

the faith that they are communicating directly with


each other. As the attacker has the genuine The impersonation is a threat in which malicious
communication, they can trick the recipient into nodes alter the data flow route and tempt the nodes
thinking they are still getting a rightful message. to the wrong positions. In Sybil attack, a malicious
Man-in-the-middle attacks can be succinct in many user dissimulates to be a separate user after acquiring
ways, including MITM, MitM, MiM and MIM. multiple identities and tries to create a relationship
with a truthful user. If the malicious user is successful
3.3.5 Sinkhole Attack in compromising one of the truthful users, the
attacker gains unauthorized privileges that help in
The sensors, which are leave unattended in the the attacking process [24].
network for long periods, are mainly susceptible to
sinkhole attack. The compromised node attracts the 3.3.9 Social Engineering Attack
information from all the neighboring nodes. Thereby,
the intruder posts other attacks, [21] such as selective Social engineering is the act of manipulating people
forward, fabrication, and transformation. so they give up secret information [25]. The types of
information that criminals are seeking can vary, but
3.3.6 Botnet Attack when individuals are targeting, the criminals are
usually an effort to deceive the user into giving them
A botnet is a collection of Internet-connected passwords or bank information. Alternatively, they
computers whose security defenses have been infarct could be effort to access a computer in order to
and control ceded to a spiteful party. Each such stealthily install malicious software that will then
compromised device, known as “bot” is created, give them access to confidential information, as well
when a computer is infiltrated by [22] software from as giving them control over the computer.
a malware distribution, else cognize as malicious
software. The controller of a botnet is able to direct 3.3.10 Hello Flood Attack
the concern with of these compromised computers
via communication channels formed by standards- In Hello flood news attacks, every object will
based network protocols such as hypertext transfer familiarize with Hello messages to all the vicinal that
protocol (http) and internet relay chat (IRC). are reachable at its frequency level. A malicious node
will cover a wide frequency area, and hence it
3.3.7 Side-Channel Attack becomes a vicinal to all the nodes in the network.
Thereupon, this malicious node will also broadcast a
Most IoT objects, for security intent, will be Hello message to its entire vicinal, make an
integrated with some of security mechanisms such as impression the availability. Flooding attacks cause
an encryption to protect their confidential data. The non-availability [26] of resources to rightful users by
side channel attack, [23] is intended to break such distributing a large number of nonsense requests for
mechanisms by analyzing side channel information a few services.
emitted by IoT objects. In this context, power and
time analysis attacks are some examples of such type 3.3.11 Data Insertion Attack
of attacks.
During the process of transposing data transmitted
3.3.8 Sybil Attack between two objects equipped with NFC protocol, an

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


467
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

attacker could insert some data into this data only, if


the object needs a long time to reply [27]. The well-
turned insertion could only happen if the inserted
data can be conveyed, before the original device
starts with the answer. If both data streams overlap,
the data will be unserviceable.

3.3.12 Flash Crowd Attack

A flash crowd is fundamentally an unexpected


increase in the overall traffic to any specific web page
or website on the Internet and the unexpected Figure 4. The IP Spoof Attack
occurrence of any event that triggers that particular
massive traffic of people accessing that web page or 3.3.14 Eavesdropping Attack
website. Less robust sites are unable to cope with the
large increase in traffic and become not available Eavesdropping is obstructing of network traffic to
[14]. The general causes of flash crowd or lack of gain prohibited access. It can outcome in lack of
sufficient data bandwidth, servers that collapse to success of confidentiality. The man in the middle
cope with the huge number of requests, and traffic attack is also a category of eavesdropping.
quotas. Eavesdropping attack observes the packets
transmitted between objects and servers during the
3.3.13 IP Spoof Attack firmware [29] upgrade process. The attacker could
only get confidential data if the packets are either
Spoofing is a type of attack in which the attacker weakly keeping safe or not keep safe at all.
dissembles to be someone else in order to gain access
to prohibit resources or thieve information. This type 3.3.15 Object Tampering Attack
of attack can take a [28] diversify of different forms,
for example; an attacker can act the IP address of a The likelihood of accessing IoT objects physically by
authorize user to get into their account. IP address attackers is very high because some IoT objects may
spoofing, or IP spoofing, refers to the creation of IP bring into service in stoic environments.
packets with a counterfeit source IP address, called Consequently, such objects are vulnerable to
spoofing, with the purpose of concealing the identity hardware attack, the most [30] renowned ones are
of the sender or imitate another computing system. the extraction of cryptography keys, the modification
The figure 4 shows the IP Spoof attack. of operating system or firmware, and the circuit
transformation.

IV. INTERNET OF THINGS (IoT) SCHEME

The IoT is a computing idea that depicts the thought


of everyday physical objects being related to the
internet and being able to identify themselves to
other devices. It enables multifarious physical objects

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


468
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

to share information and coordinate decisions. In spite


of, the functionality and operations of the IoT heavily 4.3 Friend Relationship Sybil Scheme
depend on the underlying network connectivity
structure [1]. The IoT characteristic ubiquitous In a mobile network, due to the mobility and the
communication among all kinds of electronic devices, deficiency of global social graph information, Sybil
it inevitably raises security concerns due to seamless defense is completely different and arduous
infiltration and automated integration among all sorts compared with that in the online networks [24].
of applications. Consequently, efficient and effective Quercia and Hailes propose an MSD scheme to match
defense mechanisms are of the paramount importance mobile user communities and label the users from the
to make sure the reliability of the IoT [31]. Sybil community as Sybil attackers [35]. In, one
imagination is that each mobile maintains two lists,
4.1 Sequential Scheme
first familiar list containing the trusted mobile users,
and second foe list with the unfaithful users in it. In
A sequential defense scheme is proposed by that
view of the fact that two users are encountered in the
sequentially collects feedbacks from high degree
network, they match their communities. If a user
nodes for attack presumption. The benefit of
were not in the trusted communities, this user would
sequential defense is that there is no necessity to
be reckoned as a Sybil user. In [36], Chang et al. also
acquire feedbacks from all nodes, and it abolishes the
propose a Sybil defense scheme in MSNs, assuming
collection process once sufficient feedbacks have
that the Sybil users and normal users happen in
been collected for attack presumption. The huge
various communities, and rely on the community
network size renders concurrent data transmissions
analogous to detect the Sybil users [24]. Therefore,
infeasible, particularly for wireless networks with
leveraging rapprochement is an effective solution to
scarce radio resources [32]. Besides, due to the huge
detect Sybil attackers. Although, this type of FR-MSD
network size and limited computational power,
schemes need mobile users to maintain the faithful
analyzing the collected information from all nodes
community information in advance.
incurs terrible computation overheads and it may
default to [33] provide timely safeguard. It is
4.4 Fusion Scheme
reaffirming in that a relatively small fraction of
feedbacks is enough to detect fatal attacks on the
A fusion-based defense mechanism is proposed to
network prior to network interruption.
deduce the presence of an attack [37] based on the
response from each node. The response information
4.2 Topological Scheme
can be as simple as a binary status report reflecting
that each node is, or is not, under attack, based on the
A topological defence scheme permits alteration of
node-level detection capacity [38]. The network level
network topology to enhance network resilience [34].
attack presumption plans are carried out at the fusion
It has been established in that by swapping a small
center [39]. A two-player game between the
number of edges in the network topology, one is able
protector and the attacker is naturally formed, given
to significantly make better network resilience
the critical value of network resilience and the node-
without including extra edges. Moreover, the
level detection configurations. Intuitively, from the
proposed edge rewiring technique in can be executed
adversary’s point of view, too few node elimination
in a distributed fashion, which is especially
cause hardly any harm to the network connectivity,
advantageous for the IoT due to scalability.
while too many node elimination is prone to be

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


469
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

detected by the fusion center, which means that the arduous endeavor. On the physical level, device
attack is eventually in futile. From the protector enclosures are frequently not tamper proof devices
point of view, inferring attacks using all feedbacks can be unlocked and their hardware can be accessed
might treat the topological attack as a fabled alarm, via investigate and pin headers [45]. The device
since only a small subset of nodes are targeted. On central processing units (CPUs) are low-cost
the other hand, inferring attacks using only a few components that frequently have no sophisticated
feedbacks might tolerate from information means to protect their code, data, and tokens for
insufficiency and therefore fabled to detect the external access. This permits an attacker to clone
presence of attacks [40]. Accordingly, there exists a entire devices or manipulate software and data. If the
balance point at which both attacker and defender device is brought into service in an unsupervised
are contented with their own master plan, which is environment, it may be accessed and frame up by a
precisely the notion of Nash equilibrium in game malicious third party without information. Besides,
theory. As an outcome, the game payoff at game IoT devices are frequently based on low-power
equilibrium can be used to study the robustness of a hardware and may only be able to process tokens
network. with a low complexity. This can have an implication
on the robustness of a token, as it can be
V. INTERNET OF THINGS (IoT) RELIANCE reengineered or recovered through a brute force
attack. As an outcome of this, any faith management
The Internet of Things (IoT) can deliver important system for IoT deployments must have the ability to
advantage to our society and economy, enabling dynamically withdraw faith of individual devices.
superior health services, cleaner and secure societies, When faith and security credentials given out at the
more dexterous industries and [41] stimulating new time of manufacturing or deployment, a device is
businesses. On the other hand, data is a sensitive seen as initially honest. A network wide update
subject and customers and end users are wary of mechanism will ideally incorporate a smooth and
sharing data for apprehension that it will fall into the effective patching process, which contains robust
hands of criminals, or using in ways they consider integrity and authenticity examine, minimize service
inappropriate [42]. Faith is the oxygen, which will outages, and permits for a version withdraw if
breathe life into the internet of things. Industry needed.
necessity to show data is secure and that it is properly
treated [43]. Faith and security are based on tokens or 5.2 Secure Key Storage
credentials, provided by a faith management
infrastructure, which are embedded in and The secure storage provision to increase the
potentially shared between devices. The aim of this robustness of reliance tokens used both within an IoT
section is to investigate frameworks that ensure faith system and within its reliance management
as well as communication security between nodes in infrastructure [46]. Passive keystores endow a means
an IoT deployment [44]. to securely save and retrieve credentials cryptographic
operations are carried out [10] outside these stores by
5.1 Reliance and Security from a Device Point of the device’s CPU. Active keystores in contrast, permit
View the internal execution of cryptographic operations
through an application program interface (API), so
IoT devices are vulnerable in many sides, so the credentials are never disclosed.
providing and maintaining faith and security is an

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


470
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

5.3 Reliance and Security from a Network Point of the number one users of the Internet and necessity to
View digital identities for secure operation. As enterprises
seek to transform their business models to stay
During the operation of a network, devices set up competitive, swift adoption of IoT technologies is
stable or unstable communication links with other creating increasing demand for Public Key
peers. These links can be either point-to point or Infrastructures (PKIs) to confer digital certificates for
incorporate a group of nodes. From a device’s point of the increase number of devices and the software and
view, the challenge [47] is to validate the authenticity firmware they run [49]. The secure IoT deployments
and authority of the other peer and to set up a secure need not only faith the devices to be authentic, but
communication link to keep away from attack also faith that the data they collect is real and not
scenarios. For this objective, reliance tokens are changed. If one cannot fault the IoT devices and the
interchange and validated or new session tokens bring data, there is no point in collecting, running
into being. The promise of data integrity, optionally analytics, and executing decisions based on the
in combination with data confidentiality through [48] information gathered.
encryption, provides trustworthiness in the data a
node sends or receives. Data integrity and 6.1 The PKIs Elements
confidentiality provide a foundation to deal with the
principal attack vectors of interception, A public key infrastructure confers the revocation,
transformation and disruption. The data secrecy is distribution, revocation and verification of public
generally conferred through symmetric encryption keys used for public key encryption, and enables
with the advanced encryption standard algorithm as a linking of identities with public key certificates. The
de facto industry standard often implemented directly public key infrastructure describes the policies, the
in hardware. While data integrity confers through procedure, the hardware, the software, and the
message authentication codes or cryptographic hashes people that are used to handle digital certificates [50].
that are attached to the data payload [48]. The peer This means the process of creating, distributing,
authenticity belonging to the problem of how a peer handle, storing, and revoking certificates are all
can validate another peer’s identity before a encapsulate under PKI. The PKI also mention to the
communication link is established. Peer authenticity associations we make with a public key to a person,
can go hand in hand with system availability. For or a public key to a device. It is based on faith, and a
instance, denial-of-service (DoS) way attacks are certificate authority creates this faith. This key
typically external attacks, so the ability to qualify and management lifecycle starts with the creation of a
if necessary to discard data or connection requests [14] key [51]. We make up one's mind on a distinctive
at an early stage can help to reduce such attacks. Proof strength of the key, which would be a certain
of authorization provides commitment that a peer has number of bits, and we make up one's mind what
the authority to communicate with another peer and cipher we would like to use to create the key. Now,
conduct a few actions. we are discussing the PKI elements.

VI. PUBLIC KEY INFRASTRUCTURES (PKIs) IN 6.1.1 The Certification Authorities


INTERNET OF THINGS
A believable party confers the root of faith for all PKI
At present, there are more things (devices) online certificates and confers services that can be used to
than there are people on the planet. The devices are authenticate the identity of individuals, computers

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


471
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

and other existence. Generally known as certificate typically through online certificate status protocol
authorities (CA), these entities provide promise about services.
the parties identified in a PKI certificate. Each CA
keeps going its own root CA, for use only by the CA. 6.1.4 The Central Directories
The certification authority browser forum, also called
a CA & browser forum, is an industry coalition The central directories protect location in which are
founded in 2005 and whose members contain [51] stored and index keys [50]. The central directories
CAs, browser software publishers and other system make certificates accessible to other entities. Since
contributor, who use X.509 digital certificates for other data, such as policies or CRLs necessity to be
authentication and encryption. The CAs forms the published as well, central directories store and make
backbone and the faith anchors of a PKI. They all these data accessible. They are frequently
problem certificates and, in many cases, repeal status executing as lightweight directory access protocol
data regarding the certificates they problem, and servers.
publish both types of products.
6.1.5 The Time Stamping Authorities
6.1.2 The Registration Authorities
When the date and time of the phenomena is
The registration authorities (RAs) act as the front end recorded, we say that it is time stamped. A digital
of certification authorities. A registration authority, camera will record the time and date of a photo being
often called a subordinate CA, problem PKI taken, a computer will record the time and date of a
certificates. The registration authorities are certified document being saved and emended. These are all
by a root certificate authority and authorized to issue instances of a timestamp. Timestamps are essential for
certificates for distinguished uses permitted by the keeping records on when information is being
root. That they are accountable for identifying and reciprocity, created, or destroyed online. In many
authenticating entities that request certificates, and situations, these records are simply utilitarian for us
then dispatching certificate requests to CAs and to know about. However, in some situations, a
routing back the certificate(s) to the implore entity. timestamp is more valuable. The time stamping
In some instance, RAs are just a unique component of authorities are characterized by their ability to
CAs. problem PKI based believable timestamps. Believable
time stamping is a process that keeps track of the
6.1.3 The Validation Authorities creation and alteration of data. This data can be a
program or a document. This process is done in a
A certificate database stores information about trouble free manner and recorded so that no one cans
controversy certificates. Therewith, to the certificate alteration the data, including the owner, without
itself, the database includes validity period and the being informed and it assurance the integrity of the
status of each PKI certificate. The validation data.
authorities (VAs) permit for the validation of
certificates. Validating a certificate in fact comprises 6.1.6 The Certificate Revocation Authorities
many steps for instance; possibly acquire certificates,
verifying signatures, checking the revocation status. A certificate revocation list (CRL) contains digital
It’s normally supposed that VAs only provides certificates that have been invalidated by the
services in connection to check revocation status, emanate certificate authority before their scheduled

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


472
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

expiration date and should no longer be believed. The The connected IoT endpoints share coequal security
CRLs is a type of blacklist and are used by different requirements [47]. A faithful or device identity,
endpoints, including web browsers, to confirm confirm applications and data secured in motion and
whether a certificate is valid and trustworthy. In at rest. These needs translate to authentication,
general, invalidate duties are carried out by a signature and encryption [52]. The industrial sector
dedicated service that belongs to each certificate has placed greater trust on digital certificates for IoT
authority. While either the number of issuing system elements such as controls, applications,
certificates is high, or the complexity of invalidate sensors, devices, switches, and data [23]. The
procedures increases, or so does the number and countless commercial operations, systems and
diversification [51] of CAs, specialized authorities, infrastructure are before controlled through the
CRAs, come to play, whereby a single, centralized Internet. Many devices that use the sensors and
CRA can replace equivalent invalidate services on actuators should follow particular policy and proxy
multiple CAs. When a web browser makes a rules for authentication to authorize the sensors to
connection to a site using TLS, the web server's public their information [53]. Meantime, low cost
digital certificate is investigated for anomalies or solutions in this field have not been conferred as
difficulty, part of this process involves investigating much as needed. At the present, if we want to confer
that the certificate are not catalogued in a certificate the security for the sensors we have to use high-cost
revocation list. These investigate are arduous steps in solutions, which is a dispute with the primary goal of
any certificate-based transaction because they permit IoT to provide lightweight protocols. The need for
a user to calibrate the identity of the owner of the particular commitment and related controls is greatly
site and discover whether the CAs still considers the increased for both industrial and consumer-driven
digital certificate believable. IoT devices. The security is very critical, and one of
the key distinctions between an enterprise PKI and
6.2 The IoT Challenges in PKIs an IoT PKI is the necessity for protecting data and
safe authentication at multiple endpoints. In this
PKI for IoT needs to be dissimilar than an enterprise context, digital certificates are an increasingly famous
PKI. The majority of devices collect transmit and has solution for signing, encryption and authentication.
at least one piece of private, confidential or
proprietary information [47]. Some devices will 6.4 Need the IoT Device Authentication
cross-unencrypted network and cloud services with
varying security levels and requirements. Some The powerful IoT device authentication is needed to
devices will be physically inaccessible. Other devices make sure connected devices on the IoT can be
may be attractive to impersonate in order to [14] gain believed to be what they purport to be. Accordingly,
access to IoT system resources. Comprehension the each IoT device needs a distinctive identity that can
following difficult situation can help organizations be authenticated, when the device effort to connect to
plan certificate policies upon which the PKI a central server. With this distinctive ID in place, IT
environment will be based. system administrators can track each device via its
lifecycle, communicate securely with it, and inhibit it
6.3 In the IoT Authentication, Integrity & from executing detrimental processes. If devices,
Confidentiality manifest unforeseen behavior, administrators can
simply revoke its prerogative.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


473
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

6.5 Need the IoT Device Connected Trouble Freely certificate policy the suitable amount of promise so
that relying parties know why the certificate is
To trouble freely participate in the IoT, each considered faith, as well as to understand what the
connected device needs a distinctive identity even device is authorized to do and not to do.
before it has an IP address. This digital credential
establishes the root of faith for the device’s entire VII. TERNET OF THINGS (IoT) DESIGN IDEA
lifecycle, from initial design to deployment to FOR DIGITAL CERTIFICATES
freedom from work. The every device a distinctive
identity using the powerful cryptographic processing, In a digital identity certificate, both its owner and the
key shielding, and key management available. A CA that signed the certificate must be distinctively
digital certificate is injected into every device enable identified. While there will be a comparatively small
to firstly, authentication of each device introduced to number of CAs, there is a need for a scalable naming
the organization’s architecture. Secondly, verification scheme appropriate for billions of nodes. In the design
of the integrity of the operating system and idea, device identifier establishment plans can be
applications on the device level. Thirdly, secure based on a different technique. These techniques
communications between devices, gateway, and incorporate either a hierarchical identifier, the
cloud. Lastly, authorized software and firmware encoding of extra information, unsystematic data, and
updates, based on approved code. the use of cryptographic operations [54]. In the design
idea, certificate validity the X.509 certificates have a
6.6 The PKI Security Guarantee in the IoT limited life span, which is enciphered in the validity
field [55]. The field accommodates the two date’s
The guarantee in the PKI space can be defined as the notAfter and notBefore, both accommodates a
amount of confidence that a person or system has timestamp in the UTCTime encoding format.
that the identity being introduced in a certificate in Investigation the validity of a certificate requires
fact be suited to the device. Maintaining specific access to actual time [56], and since low cost
levels of guarantee across millions of deployed device oscillators found in embedded, systems have an
identities is no trivial task. Risks must be important drift in the order of up to many. Seconds
comprehensible and mitigation plans established and per day, the use of time synchronization protocols
executed. The IoT system may incorporate millions of like network time protocol or precision time protocol
diverse connected devices sharing data to complete should be believed. In the design idea, the public key
various tasks, it makes sense that peril be inclined to cryptosystems confer pairs of keys, whereby the
be greater for an IoT system than for an enterprise public encryption key dissimilar from the secret
use case. The higher IoT peril’s profile, the PKI decryption key [51]. Such cryptosystems are at the
environment needed to issue believe certificates core of PKI, as they confer a means to digitally, the
across the system endpoints warrants very careful hash value of a digital certificate using a CA’s private
security planning to make sure a distinctly defined key; provide a means to confirm the integrity of a
level of promise for device authentication. The digital certificate, through decoding the already. The
reliability promise applies to both enterprise and IoT encoded hash value using a CA’s public key and
PKIs and various use cases may rely on varying levels comparing it with the hash value calculated over the
of promise. Some use cases need to define, detailed presented certificate; and permit a device to digitally
and strict promise requirements, while others may sign or decrypt information. In the design idea, hash
require less definition. Organizations must define in functions are one-way functions that modify a bit

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


474
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

string of variable length into a fixed-length hash value. scenarios, a number of technologies have been
They are utilizing to digital signs a certificate [57]. In developed to achieve information privacy and security
the hash, functions have four essential mathematical objective, such as transport layer security, which
and algorithmic characteristic, first they should have a could also make better the confidentiality and
small computational complexity, second irreversible integrity of the IoT.
“one-way” functions, third infeasible to alter an input
without changing the hash and fourth it must be 8.1 In IoT Secure Device Provisioning and
infeasible to find two dissimilar inputs with the same Authentication Using Azure
hash. There are a number of various future-proof hash
algorithms in use, most particularly SHA-2 and SHA- The security token technique provides authentication
3 with customizable hash lengths of between 224 and for every call made by the device to the IoT hub by
512 bits. associating the symmetric key to each call. The X.509-
based authentication permits authentication of an IoT
VIII. INTERNET OF THINGS (IoT) PROTOCOLS device at the physical layer as part of the TLS
FOR ACCESS NETWORKS connection establishment [60]. The security token-
based method can be used without the X.509
The Internet of Things (IoT) goal to make better our authentication, which is a less safe pattern. The
lives by increasing the interconnectivity of an preference between the two techniques is primarily
increased variety of embedded computing devices dictated by how safe the device authentication needs
using components of existing Internet infrastructure to be, and availability of safe storage of the device. IoT
[1]. This will permit for communications between hub uses security tokens to authenticate devices and
sensors in cars, laptops, factory, home appliances, services to keep away from sending keys on the
mobile phones, machineries, and many other devices network. Besides, security tokens are limited in time
that are already capable of network access through validity and scope. Azure IoT SDKs automatically
existing protocols such as 3G, Wi-Fi, Bluetooth [58], generate tokens without need any special
and ZigBee. The IoT scenarios remain a challenge, configuration [61]. Some scenarios need the user to
mainly due to the large number of miscellaneous generate and use security tokens outright. These
devices as well as data exchanged via insecure scenarios contain the direct use of the AMQP, MQTT,
connections. Moreover, the concepts of security are and HTTP surfaces, or the implementation of the
extended not only to device-to-device token service pattern.
communications, but also to network aspects. As an
example, many hackers create fake networks (termed After each IoT hub has a recognize registry that can
botnets) to steal data and user privacy information. be used to create per-device resources in the service,
Normally, various security requirements should be like as a queue that contains in flight cloud-to-device
addressed to promise network, and data security. First, messages, and to permit access to the device facing
confidentiality is essential to limit network access and endpoints. The IoT hub identity registry provides safe
data only to authorized users. Second, data integrity storage of device identities and security keys for a
and authentication should be promise so that solution. Independent or groups of device identities
messages are triumphantly transmitted and are can be added to a permit list, or a block list, enabling
reliable to the receiver. In the end, data complete control over device access. The use of a
authentication and availability should be provided, as device-based X.509 certificate and its associated
well as detection of malicious interloper [59]. In IoT private and public key pair permits extra

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


475
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

authentication at the physical layer. The private key close to the gateway and negotiator, propagation
is stored securely in the device and is not detectable delays can be ignored. In addition, for the
outside the device. The X.509 certificate contains unidirectional case, the security keys have a valid
information about the device, like as device time interval sufficiently long to transmit one or more
identification, and other organizational details. A packets and to absorb possible retransmissions or any
signature of the certificate is originated by using the other undesirable latency.
private key. Internet connection between the IoT
device and IoT hub is secured using the transport IX. GUIDELINES FOR SECURE THE INTERNET OF
layer security standard. Azure provides for IoT THINGS DEVICES
transport layer security standard, namely TLS 1.2,
TLS 1.1, and TLS 1.0, in this order. The Internet of Things is comprised of an
indiscriminately diverse range of device types from
8.2 The Secure Access to Unidirectional Data in IoT small to huge, from simple to complex, from
consumer gadgets to state of the art systems found in
The unidirectional devices cannot perform any safe DoD, [1] utility and industrial and manufacturing
procedure for secure key exchange with the systems. IoT devices face the same types of privacy
negotiator. The transmitter just sends a message and security problem that many traditional end-user
without any feedback, that is, it lapses to receive any devices face. There are approx six million new things
signal, and is equipped with an internal clock, which being connected every day in 2016, as we head
is supposed not to be on the mark. Then, a toward more than 22 billion by 2020, according to
nonspecific non-IP unidirectional terminal runs the Gartner [5]. End users do not have the technical
following move to send data to the gateway and specialist to assess the privacy and security
negotiator in a safe way. In the first move, it implications of any particular IoT device, or they may
generates the encryption key locally, based on the lack interest in doing so [63]. The subscribers already
time measured by a local clock. In the second move, have trouble identifying and troubleshooting the
it creates the message and encrypts it with the devices that are currently connected to their home
generated key, this message includes the payload and networks [64]. IoT devices will worsen these
any other data to be used to make better security. In circumstances, as subscribers connect an increasingly
the third move, it computes the hash values using the wide variety of devices to their home networks. The
message text and the generated key and attaches consumer will likely lose track of what devices are
them to the message. At the end move, it sends the connected to the Internet over time, which will make
message to the gateway and negotiator. defend them even more challenging. In this section,
we are discussing guidelines for secure the IoT
8.3 The Secure Access to Bidirectional Data in IoT
devices.

For bidirectional terminals, the negotiator can 9.1 Don’t Connect Your Devices Unless You Necessity
periodically broadcast its clock timing in a dedicated
message, and its identification in the plain part of the The first step is to consider what functionality you
message. The terminals can align their local clocks to necessity from the device. In view of the, your TV or
the gateway and negotiator terminal, and then fridge can connect to the internet, doesn’t mean you
generate the security keys in accordance with the surely want to hook it up. Take a good look at the
algorithm already described [62]. Since devices are

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


476
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

features it offers and learn precisely what internet sent through email or SMS before access is permitted.
connectivity brings before you connect. When used properly, two-factor authentication can
halt the bad people obtain access to your accounts
9.2 To Secure Communications Using Encrypted and taking control of your IoT devices.
Protocols
9.5 Using the Latest Firmware
Encryption practices of IoT devices are lower and
unsafe. A small number devices use encrypted If you want to make sure you have the latest security
communications as part of their beginning patches and diminish the possibility of a successful
configuration. Instead, most use ordinary web attack, then you need to keep your firmware fully
protocols that communicate across the Internet in updated. Vulnerabilities and exploits will be
plain text, which makes them simple targets for extricated as they emerge, so your IoT devices and
hackers keep an eye on network traffic to identify your router need to be frequently updated. Automate
debility. At the very least, all webs, traffic should be this wherever possible or set a schedule to investigate
using HTTPS [65], transport layer security, secure file for updates every two months or so.
transfer protocol, DNS security extensions, and other
secure protocols for communications with 9.6 During Processing Using a Encrypt Data
management stations and across the Internet.
Therewith, devices that connect to mobile apps or Infrequently the party processing the data should not
other remote gateways should use encrypted be able to read the data or the computational
protocols as well as encrypt data stored on flash outcome. The operational data while they are in
drives. encrypted form. For instance, identical encryption is
a form of encryption that permits computations to be
9.3 Block Incoming Traffic When Possible carried out on cipher-text, thus generating an
encrypted outcome that, when decrypted, matches
The several IoT devices ship with open ports to the outcome of operations performed on the plain
support management functions rather than the text.
standard functionality available through a user
interface. Even some passwords allow telnet access 9.7 Create Impressive and Inoffensive Password
with only an IP address. Afterwards, the point here is Policies
to decrease your attack surface as much as feasibly
possible. That might mean perfectly blocking all Most network infrastructure needs the administrator’s
incoming traffic with a firewall. However, in other default password to be altered when first accessed. In
cases, that will mean only keeping open which TCP spite of that, most devices, like as home routers,
and UDP ports you necessity. network printers, and sensors, lack strong
authentication and access technique. Furthermore,
9.4 Two-Factor Authentication the concept of using multifactor authentication using
a diversity of mechanisms to log in besides an easy
Supposing any of your devices offer two-factor
password, such as with an SMS code sent to a cell
authentication, use it. Two-factor authentication is
phone is a rarity in the IoT world. Actually, some IoT
an extra security layer on top of a device password
devices do not need any authentication. A subscriber
that need secondary authentication a one-time code
can navigate with a web browser to a specific IP

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


477
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

address and control the device’s configuration and read up on the provider’s privacy policy and look for
operation. commitment about encryption and data safety.

9.8 Investigate if Physical Access Assents Intrusion 9.12 Put IoT Devices on Their Personal Firewalled
and Monitored Network
It is obliged to understand how your attack surface
dissimilar in the case that a hacker is remote versus, When it comes to linked consumer grade IoT devices
when they are corporeally in the office location. in the enterprise, you need to take a proactive

There are a number of attached devices that are procedure. You want to have them segmented away
vulnerable subsequently doing a hard reset. If there and rear a firewall. You can block incoming traffic to
are any, consider locking them away, when feasible. it so people cannot attack from the inboard and you
can handle and monitor it closely.
9.9 Decrease Data Granularity
9.13 Disable UPnP Characteristics

The IoT applications should appeal the minimum


The IoT devices tend to have Universal Plug in and
level of granularity that is needed to perform their
Play (UPnP) characteristics, enabling various devices
most important tasks. A higher level of granularity
to explore and connect to one another. Whilst this is
could lead to secondary data usage and in conclusion
favorable and removes the need to configure each
privacy violations.
device individually, the protocols rely on local
9.10 Ameliorate Failover Design networks to connect to each other and these are
vulnerable to third party attackers.
The devices should function when Internet
connectivity is vanished or interrupted. However, 9.14 Conduct Risk Assessment
few IoT devices are designed to face with the lack of
success, such as Internet continuity or data After choosing the devices, check-up the network and

disconnections. Failover design is especially vital for its potential lapse points, as well as the IoT and Cloud
IoT devices that involve user protection, like as video platform used for handling and storage of data. Many
monitoring, door lock mechanisms, and simpler IoT devices have no computing power and
environmental monitors and alarms. These devices communicate with a gateway and the idea should be
should have manual overrides or particular functions given to choosing a remote monitoring service or IoT
for disconnected operations. service platform.

9.11 Careful of the Cloud Services X. CONCLUSION


In the time to come, every object in our daily life will
Many IoT devices rely on cloud services, but the need be connecting to the Internet. In this context, mobile
for an internet connection in order for something to phones will be used as the center point or the remote

function can be a real difficulty. Not only will it not control for all objects in the physical world commonly
work when the network is very slow, but it may also called as IoT. The Internet of Things opens up a new
be synchronized sensitive data or offering another universe of connected and intelligent devices that can
potential route into your home and make sure you work together to provide virtually unlimited
capabilities, and the majority of these new capabilities

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


478
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

will be personalized. Much of the value of the IoT Technology,” International Transaction of
comes from the capability to customize products and Electrical and Computer Engineers System
services to a client individually and immediately, (ITECES), USA, ISSN (Print): 2373-1273 ISSN
necessity. The analysts have a prophecy that (Online): 2373-1281, Vol. 4, No. 1, page 26-38,
hundreds of thousands of new IoT services will May 2017, DOI: 10.12691/iteces-4-1-4.
connect billions of new IoT devices over the next [4]. Luigi Atzori, Antonio Iera, and Giacomo
decade. Industry and academia are both Morabito. The internet of things: A survey.
concentrating on moving ahead in attempts to Computer Networks, 54(15):2787–2805, 2010.
improve usability, maintainability, and security via [5]. Gartner Inc. Press Release (2014)
standardization and development of best practices. http://www.gartner.com/newsroom/id/2905717
The Internet of Things (IoT) embodies the [6]. G. Wu, S. Talwar, K. Johnsson,N. Himayat, and
convergence of the physical and virtual worlds. It is K. D. Johnson, “M2M: from mobile to embedded
the important nexus between data-oriented internet,” IEEE Communications Magazine, vol.
applications and device-oriented sensor networks 49, no. 4, pp. 36–43, 2011.
facilitated by Internet technologies. The biggest [7]. L. Atzori, A. Iera, and G. Morabito, “The
challenges to the IoT will come in securing sensitive internet of things: a survey,” Computer
information from unauthorized access as well as Networks, vol. 54, no. 15, pp. 2787–2805, 2010.
authorizing access to only the information we're [8]. Gartner Inc., “Hype cycle for the internet of
comfortable disclose. This paper provides an things 2017,” Technical report, July 2017.
overview of Internet of Things (IoT) threats, scheme [9]. V. Karagiannis, P. Chatzimisios, F. Vazquez-
and faith. Afterwards, we are briefly discussing Gallego, J. Alonso-Zarate, "A survey on
public key infrastructures and guidelines for secure application layer protocols for the internet of
the IoT devices. things", Transaction on IoT and Cloud
Computing, vol. 3, no. 1, pp. 11-17, 2015.
XI. REFERENCES [10]. Xiaolin Jia, Quanyuan Feng, Taihua Fan,
Quanshui L. , ”RFID technology and its
[1]. Yusuf Perwej, Mahmoud A. AbouGhaly, Bedine applications in Internet of Things (IoT)”, 2nd
K. Hani Ali M. Harb, “An Extended Review on International Conference on Consumer
Internet of Things (IoT) and Its Promising Electronics, Communications and Networks
Applications”, Communications on Applied (CECNet), IEEE, Yichang, China, April 2012
Electronics (CAE), ISSN: 2394-4714, Foundation [11]. Nasser S. A., Andrew J., Olga A.,” Internet of
of Computer Science FCS, New York, USA, Things Security: A Review of Risks and Threats
Volume 9, Number 26, Pages 8– 22, Feb 2019, to Healthcare Sector ”, IEEE International
DOI: 10.5120/cae2019652812 Conference on Internet of Things (iThings) and
[2]. Yusuf Perwej, “An Experiential Study of the Big IEEE Green Computing and Communications
Data,” International Transaction of Electrical (GreenCom) and IEEE Cyber, Physical and
and Computer Engineers System (ITECES), USA, Social Computing (CPSCom) and IEEE Smart
ISSN (Print): 2373-1273 ISSN (Online): 2373- Data (SmartData), Exeter, UK, June 2017
1281, Vol. 4, No. 1, page 14-25, March 2017, [12]. Wei Z., Yan Jia., Anni P., Yuqing Z., Peng
DOI:10.12691/iteces-4-1-3. L.,”The Effect of IoT New Features on Security
[3]. Nikhat Akhtar, Firoj Parwej, Dr. Yusuf Perwej, and Privacy: New Threats, Existing Solutions,
“A Perusal of Big Data Classification and Hadoop

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


479
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

and Challenges Yet to Be Solved”, IEEE Internet [22]. Elisa Bertino, Nayeem Islam,” Botnets and
of Things Journal, June 2018 Internet of Things Security ”. Computer, IEEE,
[13]. J. Gubbia, R. Buyya, S. Marusic, M. Palaniswami, Volume 50 , Issue 2, Feb 2017
"Internet of Things (IoT): A vision architectural [23]. A. Mohsen Nia and N. K. Jha, “A Comprehensive
elements and future directions" in Future Study of Security of Internet-of-Things,” IEEE
Generation Computer Systems, Elsevier, vol. 29, Transactions on Emerging Topics in Computing,
pp. 1645-1660, 2013. vol. PP, no. 99, p. d, 2016.
[14]. R. H. Weber, "Internet of Things New security [24]. Kuan Zhang, Xiaohui Liang, Rongxing Lu,
and privacy challenges", Computer Law & Xuemin Shen,” Sybil Attacks and Their Defenses
Security Review, vol. 26, no. 1, pp. 23-30, Jan. in the Internet of Things”, IEEE Internet of
2010. Things Journal , Volume 1 , Issue 5 , Oct. 2014
[15]. M. O. Lehtonen, F. Michahelles, E. Fleisch, [25]. Ian G. Harris,” Social Engineering Attacks on the
"Trust and Security in RFID-Based Product Internet of Things ”, September 7, 2016
Authentication Systems", IEEE Systems Journal, [26]. L. Atzori, A. Iera, G. Morabito, "The Internet of
vol. 1, no. 2, pp. 129-144, Dec. 2007. Things: A survey", Comput. Netw., vol. 54, no.
[16]. Yusuf Perwej, Kashiful H., Uruj J., Firoj Perwej, 15, pp. 2787-2805, 2010.
“Block ciphering in KSA, A major breakthrough [27]. E. Haselsteiner and K. Breitfuß, “Security in
in cryptography analysis in wireless networks” Near Near Field Communication (NFC)
International Transactions in Mathematical Strengths,” Semiconductors, vol. 11, no. 71, p.
Sciences and Computer, India, ISSN-0974-5068, 71, 2006.
vol. 2, No. 2, pages 369-385, July-December [28]. Seo JW, Lee SJ,”A study on the detection of
2009 DDoS attack using the IP Spoofing”, J Korea Inst
[17]. G. Gan, Z. Lu, J. Jiang, "Internet of Things Inf Secur Crytol 25(1):147-153, 2015
Security Analysis", 2011 International [29]. D. Miessler, “Securing the Internet of Things:
Conference on Internet Technology and Mapping Attack Surface Areas Using the
Applications, pp. 1-4, Aug. 2011. OWASP IoT Top 10.
[18]. I. Gudymenko, K. B. Pfitzmann, K. Tietze, [30]. G. Hernandez, O. Arias, D. Buentello, and Y. Jin,
"Privacy implications of the internet of things" “Smart Nest Thermostat : A Smart Spy in Your
in Constructing Ambient Intelligence, Springer, Home,” Black Hat USA, pp. 1–8, 2014.
pp. 280-286, 2012 [31]. Pin-Yu Chen, Shin-Ming Cheng, and Kwang-
[19]. Marianne Azer, Sherif El-Kassas, Magdy El- Cheng Chen. Optimal control of epidemic
Soudani, “A full of the Wormhole Attack”, information dissemination over networks. IEEE
International Journal of Computer Science and Trans. Cybern., 44(12):2316–2328, December
Information Security, 2009. 2014.
[20]. Kelly Jackson, “Hacker’s Choice: Top Six [32]. Pin-Yu Chen and Shin-Ming Cheng. Sequential
Database Attacks. defense against random intentional attacks in
[21]. Ş. Okul ; M. Ali Aydın,” Security Attacks on IoT complex networks. Phys. Rev. E, 91:022805,
”, International Conference on Computer February 2015.
Science and Engineering (UBMK), IEEE, [33]. Shan C, Hui X, Da L, et al. A Vision of IoT:
Antalya, Turkey, Oct. 2017 Applications, Challenges, and Opportunities
With China Perspective{J}. IEEE Internet of
Things Journal, 1(4), 349-359, 2014

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


480
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

[34]. Pin-Yu Chen and Alfred O. Hero. Assessing and [44]. J. Guo, R. Chen, "A Classification of Trust
safeguarding network resilience to nodal attacks. Computation Models for Service-Oriented
IEEE Commun.Mag., 52(11):138–143,November Internet of Things Systems", IEEE International
2014. Conference on Services Computing (SCC), pp.
[35]. D. Quercia and S. Hailes, “Sybil attacks against 324-331, June 2015.
mobile users: Friends and foes to the rescue,” in [45]. Ie yuan, Xiaoyong li,” A Reliable and
Proc. IEEE IEEE Conf. Comput. Commun. Lightweight Trust Computing Mechanism for
(INFOCOM), 2010, pp. 336–340. IoT Edge Devices Based on Multi-Source
[36]. W. Chang, J. Wu, C. Tan, and F. Li, “Sybil Feedback Information Fusion”, special section
defenses in mobile social networks,” in Proc. on security and trusted computing for industrial
IEEE Conf. Global Commun. (GLOBECOM), internet of things, volume 6, May 16, 2018
2013, pp. 1–6. [46]. P. Porambage, C. Schmitt, P. Kumar, A. Gurtov,
[37]. Pin-Yu Chen and Kwang-Cheng Chen. M. Ylianttila, "Two-phase authentication
Intentional attack and fusion-based defense protocol for wireless sensor networks in
strategy in complex networks. In Proc. IEEE distributed IoT applications", Proc. IEEE
Global Telecommunications Conference Wireless Commun. Netw. Conf. (WCNC), pp.
(GLOBECOM), pages 1–5, December 2011. 2728-2733, Sep. 2014
[38]. Pin-Yu Chen, Shin-Ming Cheng, and Kwang- [47]. E Vasilomanolakis, J Daubert, M Luthra, V
Cheng Chen. Information fusion to defend Gazis, A Wiesmaier, P. Kikiras, "On the Security
intentional attack in internet of things. IEEE and Privacy of Internet of Things Architectures
IoT-J., 1(4):337–348, August 2014. and Systems", 2015 International Workshop on
[39]. J. Tian, W. Zhao, R. Du, and Z. Zhang, “A New Secure Internet of Things (SIoT), pp. 49-57, Sep
Data Fusion Model of Intrusion Detection- 2015
IDSFP,” in Parallel and Distributed Processing [48]. Chen, D., Chang, G. R., Sun, D. W., Li, J. J., Jia.
and Applications, vol. 3758 of Lecture Notes in J., and Wang. X. W.,”TRM-IoT: A Trust
Computer Science, pp. 371–382, Springer Berlin Management Model Based on Fuzzy Reputation
Heidelberg, Berlin, Heidelberg, 2005. for Internet of Things”, Computer Science and
[40]. Martin Osborne and Ariel Rubinstein. A Course Information Systems.,1207-28, October 2010
in Game Theory. MIT, Cambridge, MA, 1999. [49]. Michael Schukat, Pablo Cortijo,” Public key
[41]. S. Sicari, A. Rizzardi, L. A. Grieco, A. Coen- infrastructures and digital certificates for the
Porisini, "Security privacy and trust in Internet Internet of things”, 26th Irish Signals and
of Things: The road ahead", Computer Systems Conference (ISSC),IEEE, Carlow,
Networks, vol. 76, pp. 146-164, Jan. 2015. Ireland, June 2015
[42]. T. Eder, D. Nachtmann, D. Schreckling, "Trust [50]. Zhiyong Zhang, Qingqi Pei, Jianfeng Ma, Lin
and Reputation in the Internet of Things", Yang, "Security and Trust in Digital Rights
Conference Seminar (SS2013) - Real Life Management: A Survey", International Journal
Security (5827HS), Dec. 2013. of Network Security, vol. 9, no. 3, pp. 247-263,
[43]. Z. Yan, P. Zhang, A. V. Vasilakos, "A survey on 2009.
trust management for Internet of Things", [51]. Younggyo Lee, Jeonghee Ahn, Seungjoo Kim,
Journal of Network and Computer Applications, Dongho Won, "A PKI System for Detecting the
vol. 42, pp. 120-134, Jan. 2014. Exposure of a User's Secret Key Public Key

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


481
Dr. Yusuf Perwej et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 462-482

Infrastructure Springer Berlin, Heidelberg", vol. [60]. Borislav S. Đorđević, Slobodan P. Jovanović,
4043, pp. 248-250, June 2006. Valentina V. Timčenko, ” Cloud Computing in
[52]. Yun-kyung Lee, Hong-il Ju, Jee-hye User, Amazon and Microsoft Azure platforms:
authentication mechanism using authentication Performance and service comparison ”, 22nd
server in homnetwork Advanced Telecommunications Forum Telfor
Communication Technology, pp. 504-506, 2006. (TELFOR),IEEE, Belgrade, Serbia, Nov. 2014
[53]. Gianmarco Baldini, Trevor Peirce, Maria Chiara [61]. R. Tudoran, A. Costan, G. Antoniu, L. Bougé, "A
Tallachini, "Internet of Things: IoT Governance performance evaluation of azure and nimbus
European Research Cluster on the Internet of clouds for scientific applications", Proceedings of
Things, Jan. 2014 the 2nd International Workshop on Cloud
[54]. M. Bauer, P. Chartier, K. Moessner, Catalogue of Computing Platforms, pp. 4, 2012.
IoT Naming, Addressing and discovery schemes [62]. R. Giuliano, A. Neri, and D. Valletta, “End-to-
in IERC projects V1.7, IERC-AC2-D1, 2013. end secure connection in heterogeneous
[55]. AS. Wazan, R. Laborde, F. Barrère, A. Benzekri, networks for critical scenarios”, WIFS 2012,
"Validating X. 509 certificates based on their Proc. of the 2012 IEEE Intl. Workshop on
quality", Young Computer Scientists 2008. Information Forensics and Security, pp. 264–
ICYCS 2008. The 9th International Conference 269, Tenerife, Spain.
for IEEE, pp. 2055-2060, November 2008. [63]. Ka-Ping Yee, "Aligning security and usability."
[56]. J. Shannon, H. Melvin, A. G. Ruzzelli, Dynamic IEEE Security & Privacy 2.5, pp 48-55, 2004
flooding time synchronization protocol for [64]. Rebecca E. Grinter, et al., "The work to make a
WSNs, IEEE GLOBECOM, 2012. home network work." ECSCW 2005. Springer
[57]. P. Camion, J. Patarin, "The knapsack hash Netherlands, 2005
function proposed at Crypto'89 can be broken", [65]. A. P. Castellani, M. Gheda, N. Bui, M. Rossi, and
Adv. in Cryptology Proc. Eurocrypt'91, pp. 39- M. Zorzi, "Web Services for the Internet of
53, 1991. Things through CoAP and EXI," in IEEE RWFI,
[58]. Yusuf Perwej, Kashiful H., Uruj J., Sharad S., Kyoto, Japan, Jun. 2011
“Some drastic improvements found in the
analysis of routing protocol for the Bluetooth Cite this article as :
technology using scatternet” International
Conference on Computing, Communications Dr. Yusuf Perwej, Firoj Parwej, Mumdouh Mirghani
and Information Technology Applications Mohamed Hassan, Nikhat Akhtar, "The Internet-of-
(CCITA-2010), Ubiquitous Computing and Things (IoT) Security : A Technological Perspective
Communication Journal (UBICC) Seoul, South and Review ", International Journal of Scientific
Korea, ISSN Online 1992-8424, ISSN Print 1994- Research in Computer Science, Engineering and
4608, Volume CCITA-2010, Number 5, pages Information Technology (IJSRCSEIT), ISSN : 2456-
86-95, 2010 3307, Volume 5 Issue 1, pp. 462-482, January-
[59]. J. Granjal, E. Monteiro, J. Sá Silva, "Security for February 2019. Available at doi :
the Internet of Things: A survey of existing https://doi.org/10.32628/CSEIT195193
protocols and open research issues", IEEE Journal URL : http://ijsrcseit.com/CSEIT195193
Commun. Surveys Tuts., vol. 17, no. 3, pp. 1294-
1312, 3rd Quart. 2015

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


482
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT195199
Secure Micro Payment Using Physical Unclonable Function
1
M. Jalasri, 2S. Nalini, 3N. Magesh Kumar, 4J. Elumalai
1,2,3,4
Assistant Professor, Information Technology, Jeppiaar Maamallan Engineering College, Sriperumbudur,
Tamil Nadu, India
ABSTRACT

In online aggressors regularly go for taking such client information. In these situations, malware that can take
card information when they are perused by the gadget has thrived. This framework gives secure micro-
payment and enhances date approaches in terms of security. In proposed system, utilize two principle capacities
PUF (Physical Unclonable Function) and FRoDo(Fraud Resilient Device for Off-Line Micro-Payments). These
two can be given security to installment process and client account. These produce double encryption to
protect the system from attackers. In proposed system gives more security for transaction.
Keywords : Mobile Secure Payment, Architecture, Protocols, Advance Encryption, Fraud-Resilience

I. INTRODUCTION identification and authentication but very difficult to


exchange key to card issuers [3].
Mobile commerce is one of the future research areas,
especially for mobile payment systems. In the FRoDO highlights a character component to confirm
existing system, the payments are performed on the the client, and a coin component where coins are not
fixed cellular network which degrades security. To privately put away, but rather are registered on-the
enhance security on transaction and payment process fly when required.
by performing e-payment systems in the distributed
development of the wireless ad-hoc network [1]. Payment transaction does not directly read customer
coins. The vendor only communicates with the
In Modern cryptographic, only authorized parties can identity element in order to identify the user.The
acquire secret keys and access information. Be that as main benefit is a simpler, faster, and more secure
it may, different sorts of altering techniques have interaction between the involved actors/entities. The
been concocted to separate secrecy keys from identity element used to improve the security of the
restrictive access frameworks, for example, users. This provide more secure off-line payments.
smartcards and ATMs. This server authenticates the .
PUF by two kinds of error probabilities such as false II. RELATED WORK
alarm rate (FAR) and false detection rate (FDR) [2].
A TPM (Trusted Platform Module) is a standard for
The main goal of this paper is to introduce the variety securing hardware device with the help of
of devices for influence magnetic stripe card data and incorporated cryptographic keys. Here identify
demonstrate magnetic stripe cards by fraudulently attacks only in memory and bus probing not real
cloned and analyze how such devices can be word attacks [4].
processed. This technique used for personal

CSEIT195199 | Received : 12 Feb 2019 | Accepted : 25 Feb 2019 | January-February -2019 [ 5 (1) : 483-486 ]
483
M. Jalasri et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 483-486

Near Field Communication (NFC) technology is a A. CLIENT MODULE:


short-range wireless communication between the
Fig. 2. Client module. User is registered first before
devices. NFC used in credit card, e-ticket, and mobile
login. User and Vendor are waiting for activation.
payment. It provides security for m-payment systems
Customer gives the payment request after Activation.
and also eases of use for the user. The main issue in
Users are activated by Frodo. Gives bank name,
this system is unable to receive phone calls when
permanent card number and its pin number to
processing payment operation [5].
registration. User Details are Stored in Database.

Symmetric key cryptography which reduces


computation at all the entities of the protocol.
Kungpisdan et al and Tellez said this security
protocol satisfies the existing standard protocols like
SET. But lack of complexity produced during
transaction and identification of users [6]. However,
the solutions are limited in that protocols require at Fig. 2. Client module
least one of the two parties to be online, i.e.
connected either to a trusted third party or to a B. VENDOR MODULE:
shared database.
In fig. 3 Vendor is registered first before login.
Vendor Details are Stored in Database. Vendor is
FORCE (Fully off-line secure credits for mobile
waiting for activation. Vendors get transaction after
micro payments ) exclusively perform the operations
activation. Users and Vendors are activated by Frodo.
at catching the topological impact introduced by the
Gives bank name, Branch and password to
wormhole. Detection of wormholes is successfully
registration.
done without any overhead or need of external
devices to monitor the network. But, difficult to
share database to trusted third party it didn’t notice
any real time attacks such as skimmers, scrapers and
data vulnerabilities [7].
Fig. 2. Vendor module
III. IMPLEMENTATION WORK
C. KEY GENERATION
Architecture Diagram for secure micro payment
using physical unclonable function in Figure 1. User Frodo view the user who gives request to activation
is registered first before login. Vendor is registered and Frodo activate the user. Identity element is
first before login. Vendor Details are Stored in generating private key to the user and coin element
Database. Vendor is waiting for activation. Vendors generate binary code for user account number into
get transaction after activation. Frodo view the user binary code. This two-step makes Payment more
who gives request to activation and Frodo activate secure. Key Generator used generate the private key
the user. After activation user login into account to of the identity element and symmetric and
gives payment request. After PUF verification asymmetric cryptographic algorithms applied to data
Payment transaction will success. received in input and sent as output by the identity
element in fig, 3.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


484
M. Jalasri et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 483-486

verification. After PUF verification Payment


transaction will success.

Fig. 3. Key Generation


Identity Element Coin Element

Vendor
FRoDo
Activate the
User

Generate
Private Key For Generate Binary
User Code for Account
FRoDo Login Number

Check Availability
Balance

Request Sent to
Fig. 4. Payment request
PUF

User
User Database User Login Enter the Payment Encrypted Cash
Registration
Amount

Vendor Login
In cryptography, (Advance Encryption Standard)AES
Payment Verification Identity and Coin
Element Verification Cash decryption
is broadly embraced and upheld in both equipment
Waiting for PUF Activate the
View Vendor
Verification Vendor

PUF Login and programming. Cryptanalytic attacks against AES


Payment Success

PUF
has not been discovered practically. A degree of
Fig. 4. Architecture Diagram for secure micro ‘future-proofing’ allowed by AES which has
payment using physical unclonable function. flexibility of key length.

The Rivest-Shamir-Adleman(RSA) public key In any case, similarly with respect to DES, the AES
cryptography algorithm which is used to encrypt a security is guaranteed just in the event that it is
message. The RSA algorithm performs public key effectively actualized and great key administration is
encryption and digital signatures. utilized.
1. Represent the message as a whole number in the
vicinity of 0 and (n-1). Huge messages can be E. PUF PROCESS:
separated into various squares. Each piece would then
be spoken to by a whole number in a similar range. Heart of FRoDO proposal lies a read-once strong
2. Encrypt the message by raising it to the eth physical unclonable function. In fig.5 PUF check
control modulo n. The outcome is a ciphertext vendors details involved in payment and activate the
message C. vendors.
3. To unscramble ciphertext message C, raise it to
another power d modulo n

The encryption key (e,n) is made open. The


unscrambling key (d,n) is kept private by the client.

D. PAYMENT REQUEST
Fig. 5. Puf process
After activation user login into account to gives
payment request in fig. 4. Enter the required In payment verification PUF check both Identity and
payment and Check the availability of payment for Coin Element. After verification payment decrypted
customer requirement. Required amount encrypted using public key and the payment transaction
using key and after payment request goes to PUF completed successfully and whole transaction will be

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


485
M. Jalasri et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 483-486

secure. Finally receives the Private Response, the last [2]. D. Lim, J. W. Lee, B. Gassend, G. E. Suh, M. van Dijk,
step only requires the coin just read to be validated. and S. Devadas, “Extracting secret keys from
Then, the whole payment transaction can be integrated circuits,” IEEE Trans. Very Large Scale
Integr. Syst., vol. 13, no. 10, pp. 1200–1205, Oct.
authorized and committed.
2005.
[3]. G. Hong and J. Bo, “Forensic analysis of skimming
IV. CONCLUSION
devices for credit fraud detection”, 2nd IEEE Int.
Conf. Inf. Financial Eng., Sep. 2010
Security analysis shows proposed system does not [4]. P. Choi and D. K. Kim, “Design of security enhanced
impose trustworthiness assumptions.Further, it is also TPM chipagainst invasive physical attacks”,IEEE Int.
the first solution in the literature where no customer Symp. Circuits Syst., pp. 1787–1790,2012.
device data attacks can be exploited to compromise [5]. W. Chen, G. Hancke, K. Mayes, Y. Lien, and J.-H.
the system This has been accomplished basically by Chiu, “Using 3G network components to enable NFC
utilizing a novel erasable PUF engineering and a mobile transactions and authentication”, IEEE Int.
Conf. Progress Informat. Comput., Dec. 2010, vol. 1,
novel convention outline. Furthermore, our proposal
pp. 441–448.
has been thoroughly discussed and compared against
[6]. V. C. Sekhar and S. Mrudula, “A complete secure
the state of the art. Our investigation demonstrates
customer centric anonymous payment in a digital
that it is the main suggestion that appreciates every ecosystem”, Int. Conf. Comput., Electron. Elect.
one of the properties required to a safe small scale Technol., 2012, pp. 1049–1054.
installment arrangement, while likewise presenting [7]. V. Daza, R. Di Pietro, F. Lombardi, and M. Signorini,
adaptability while thinking about the installment “FORCE: Fully off-line secure credits for mobile
medium (kinds of advanced coins). micro payments”, 11th Int. Conf. Security
Cryptography, 2014, pp. 125–136.
Finally, some open issues have been identified that [8]. S. Golovashych, “The technology of identification
and authentication of financial transactions. from
are left as future work. Specifically, examining the
smart cards to NFC-terminals”, IEEE Intell. Data
likelihood to enable computerized change to be spent
Acquisition Adv. Comput. Syst., Sep. 2005, pp. 407–
over various disconnected exchanges while keeping
412.
up a similar level of security and ease of use. In [9]. Slava Gomzin, “Hacking Point of Sale: Payment
FRoDO, develop to access money without standard Application Secrets, Threats, and Solutions,
coin element. In future, try to enhance the capability Computing & Technology 1st ed. New York, NY,
of the AES algorithm to a generate private key and USA: Wiley, 2014.
create more security for Electronic payments. [10]. Meng-Day Yu, Srinivas Devadas “Secure and robust
Another approach will involve smarter error correction for physical unclonablefunctions”,
implementation of secure offline transfer of money IEEE Design & Test of Computers , vol. 27, no. 1, pp.
48–65, Jan. 2010.
by using PUF.

Cite this article as : M. Jalasri, S. Nalini, N. Magesh Kumar,


V. REFERENCES
J. Elumalai, "Secure Micro Payment Using Physical
Unclonable Function", International Journal of Scientific
[1]. N. Chitra Kiran, G. Narendra Kumar, “Building Research in Computer Science, Engineering and
robust m-commerce payment system on offline Information Technology (IJSRCSEIT), ISSN : 2456-3307,
wireless network”,2011 Fifth IEEE International Volume 5 Issue 1, pp. 483-486, January-February 2019.
Available at doi : https://doi.org/10.32628/CSEIT195199
Conference on Advanced Telecommunication
Journal URL : http://ijsrcseit.com/CSEIT195199
Systems and Networks (ANTS) , March 2012

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


486
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951102

Reversible Data Hiding in Homomorphic Encrypted Domain by


Monitoring Ciphertext Group
V. Santhi *1, M. Abinaya 2
*1 Assistant Professor, Dept. of Computer Applications, Bon Secours College for Women Tamil Nadu, India
2 M.Sc Computer Science, Bon Secours College for Women, Thanjavur, Tamil Nadu, India

ABSTRACT
Since few years, a new problem is trying to combine in a single step, compression, encryption and data hiding.
So far, few solutions have been proposed to combine image encryption and compression for example. Nowadays,
a new challenge consists to embed data in encrypted images. Since the entropy of encrypted image is maximal,
the embedding step, considered like noise, is not possible by using standard data hiding algorithms. A new idea
is to apply reversible data hiding algorithms on encrypted images by wishing to remove the embedded data
before the image decryption. Recent reversible data hiding methods have been proposed with high capacity,
but these methods are not applicable on encrypted images. In this paper we propose an analysis of the local
standard deviation of the marked encrypted images in order to remove the embedded data during the
decryption step. We have applied our method on various images, and we show and analyze the obtained results.
Keywords: Standard Data Hiding, Embedding Images

I. INTRODUCTION two technologies can be used complementary and


mutually commutative. Sinha and Singh proposed a
Nowadays, the transmission of images is a daily technique to encrypt an image for secure image
routine and it is necessary to find an efficient way to transmission. In their approach the digital signature
transmit them over networks. To decrease the of the original image is added to the encoded version
transmission time, the data compression is necessary. of the original image. The encoding of the image is
The protection of this multimedia data can be done done using an appropriate error control code. At the
with encryption or data hiding algorithms. Since few receiver end, after the decryption of the image, the
years, a problem is to try to combine compression, digital signature can be used to verify the
encryption and data hiding in a single step. For authenticity of the image. Encryption and
example, some solutions were proposed in to watermarking algorithms rely on the Kickoffs
combine image encryption and compression. Two principle: all the details of the algorithm are known,
main groups of technologies have been developed for and only the key to encrypt and decrypt the data
this purpose. The first one is based on content should be secret.
protection through encryption. There are several
methods to encrypt binary images or gray level II. LITERATURE SURVEY
images. In this group, proper decryption of data
requires a key. The second group bases the protection Reversible data hiding techniques The quality of the
on digital watermarking or data hiding, aimed at image gets disturbed when the data is embedded into
secretly embedding a message into the data. These the image. So it is expected that after the data

CSEIT1951102 | Received : 12 Feb 2019 | Accepted : 25 Feb 2019 | January-February -2019 [ 5 (1) : 487-490 ]
487
V. Santhi et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 487-490

extraction the image quality should be maintained Reversible data hiding is a technique to embed
just like the original image. But the image which is additional message into some distortion-unacceptable
obtained contains some distortions. With regards to cover media, such as military or medical images, with
distortion in image, Kalker and Williams established a eversible way so that the novel cover content can
a rate-distortion copy for RDH, through which they be perfectly restored after extraction of the hidden
showed the rate-distortion bounds of RDH for message. As an effective and popular means for
without memory covers and proposed a recursive privacy fortification, encryption changes the ordinary
code development which, however, does not move signal into incomprehensible data, so that the general
towards the bound. Another promising strategy for signal processing typically takes place before
RDH is histogram shift (HS), in which the space is encryption or after decryption.
saved where data can be embedded by shifting the
bins of histogram of gray values. In this process, the However, in some circumstances that a content
embedding of data is done in three steps. Step1. The owner does not trust the supplier, the ability to
histogram is drawn. Step2. The peak point is taken influence the encrypted data when maintaining the
into consideration. Step3. The whole image is plain content secret is needed. When the secret data
scanned row by row. After these steps, the image is to be broadcasted are encrypted, a supplier without
scanned again. If the greyscale value 154 is any information of the cryptographic key may
encountered, then the embedded data sequence is compress the encrypted data due to the limited
checked & we get the marked image. Finally, the data channel resource. Some attempts on RDH in
extraction is done. To get the original quality of the encrypted images have been made. Zhang divided the
cover, the process of histogram shift is applied again. encrypted image into numerous blocks. By spinning
The original cover is then obtained back. Basically, LSBs of the half of pixels in every block, space can be
data hiding is the process to hide the data into some created for the embedded bit. The data extraction and
covering media i.e. it is the concatenation of two image recovery proceed by finding which part has
blocks of data, first is the embedding data & second is been spinned in one block. This process can be
the covering media. But in most of the cases the grasped with the help of spatial correlation in
covering media gets distorted after the data is decrypted image.
embedded & the covering media is not inverted back
to its original form after data is removed from it. Hong et al. ameliorated Zhang’s method at the
Some reversible data hiding methods use the concept decryption side by further making use of the spatial
of differential expansion transform which is based on correlation using a different estimation equation and
haar wavelet transform. Another concept used is the side match method to gain much lower error rate.
histogram shift. The differential expansion is the These two methods explained above rely on spatial
difference between two neighbouring pixels for correlation of original image to extract data. That is,
hiding one bit of data. In this process, the histograms the encrypted image should be decrypted first before
are drawn first. Then the peak values are taken into data extraction.
consideration. Then two peak values are considered
& difference is calculated. Then according to the Zhang et al. recovered the recursive code
result the bit by bit data is embedded into the image. development for binary covers and proved that this
In this way the distortion analysis is done & it is development can gain the rate-distortion bound as
helpful to remove the distortion in the covering long as the compacting algorithm reaches entropy,
media & to get the original cover back. which launches the correspondence between data

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


488
V. Santhi et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 487-490

compression and RDH for binary covers. A more IMPLEMENTATION RESULTS


popular method is based on difference expansion
(DE), in which the difference of each pixel group is Authentication module
expanded, e.g., multiplied by 2, and thus the least An authentication module is a plug-in that collects
significant bits (LSBs) of the difference are all-zero user information such as a user ID and password, and
and can be used for embedding messages. So in this compares the information against entries in a
way the additional data can be embedded into the database. If a user provides information that meets
covering media which is an improvement to the the authentication criteria, the user is validated and,
existing methods. assuming the appropriate policy configuration,
granted access to the requested resource. If the user
III. METHODS AND MATERIAL provides information that does not meet the
authentication criteria, the user is not validated and
EXISTING PROCESS denied access to the requested resource.
There are existing systems having the key tool for
information hiding which is Vacating the room after Data Analysis
encryption. It consists of problems such as, the The Data Analysis module will focus on strategies
extracted data may contain errors. If there is no and procedures, both quantitative and qualitative, for
availability of sufficient space then some data may be analyzing social data. The module will complement
lost & that is why the data is missing at the receiver the material on data collection covered in the
side which can be termed as data with error. Again Research Appreciation module. Though the
the un-availability of memory space is a big problem. preferable sequence of progression would be from
Some space is created at the time of data embedding Research Appreciation to the Data Analysis module,
which is a time consuming process. After data the latter module can stand by itself enabling if
extraction the image recovered does not contain the necessary students to undertake these modules in
qualities of the original cover. Some distortions are either sequence.
introduced into the image. But it is possible in future
that the quality may be improved as compared to Data Encryption
existing system. Data encryption translates data into another form, or
code, so that only people with access to a secret key
PROPOSED METHODOLOGY (formally called a decryption key) or password can
The extracted data may contain errors because if read it. Encrypted data is commonly referred to as
there is no availability of sufficient space then some cipher text, while unencrypted data is called
data may lost & that’s why there is data missing at the plaintext.
receiver side which may called as data with error.
Again the unavailability of memory space is the big Data Decryption
problem, as some space is created at the time of data The conversion of encrypted data into its original
embedding which is the time consuming process. form is called Decryption. It is generally a reverse
After data extraction the image recovered does not process of encryption. It decodes the encrypted
contain the qualities as was the original cover. Some information so that an authorized user can
distortions are there into that image. only decrypt the data because decryption requires a
secret key or password.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


489
V. Santhi et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 487-490

Mail [4]. W. Liu, W. Zeng, L. Dong, and Q. Yao, “Efficient


The Mail module enables the sending of emails from compression of encrypted grayscale images,” IEEE
within Magnolia. The module can be used to send Trans. Image Process., vol. 19, no. 4, pp. 1097-1102,
Apr. 2010.
plain text and HTML messages, and template-based
[5]. V. Sachnev, H. J. Kim, J. Nam, S. Suresh, and Y.-Q.
messages. Sending emails from Magnolia is typically
Shi, “Reversible watermarking algorithm using
an automated process. An event acts as a trigger. For
sorting and prediction, ”IEEE Trans. Circuits Syst.
example, a verification message is sent to a user when Video Technol., vol. 19, no. 7, pp. 989-999, Jul.
they fill a registration form. 2009.
[6]. W. Zhang, B. Chen, and N. Yu, “Capacity-
IV. CONCLUSION approaching codes for reversible data hiding,” in
Proc 13th Information Hiding(IH’2011),LNCS 6958,
Data hiding is gaining the area of interest due to its 2011, pp. 255-269, Springer-Verlag.
provision for secured environment. Data hiding in [7]. J. Tian, “Reversible data embedding using a
difference expansion, ”IEEE Trans. Circuits Syst.
reversible manner in encrypted images is providing
Video Technol., vol. 13, no. 8, pp. 890- 896, Aug.
double security for confidential data by using
2003.
techniques such as image encryption. The existing
[8]. D.M. Thodi and J. J. Rodriguez, “Expansion
system contains some disadvantages so the future embedding techniques for reversible watermarking,”
scope is to remove the disadvantages by adding IEEE Trans. Image Process., vol. 16, no. 3, pp. 721-
reversible manner i.e. data extraction and recovery of 730, Mar. 2007.
image are free of errors. The PSNR will be improved [9]. X. L. Li, B. Yang, and T. Y. Zeng, “Efficient
to get original cover back. reversible watermarking based on adaptive
prediction-error expansion and pixel selection,”
V. FUTURE ENHANCEMENT IEEE Trans. Image Process., vol. 20, no. 12, pp.
3524-3533, Dec. 2011.
[10]. Z. Ni, Y. Shi, N. Ansari, and S. Wei, “Reversible data
In future it may be possible that memory space can
hiding,” IEEE Trans. Circuits Syst.Video Technol.,
be reserved before encryption which requires less
vol. 16, no. 3, pp. 354- 362, Mar. 2006.
amount of time for data extraction & image recovery. [11]. L. Luo et al., “Reversible image watermarking using
interpolation technique,” IEEE Trans. Inf. Forensics
Security, vol. 5, no. 1, pp. 187-193, Mar. 2010.
VI. REFERENCES [12]. X. Zhang, “Reversible data hiding in encrypted
images,” IEEE Signal Process. Lett., vol. 18, no. 4,
[1]. P. Tsai, Y. C. Hu, and H. L. Yeh, “Reversible image pp. 255-258, Apr. 2011.
hiding scheme using predictive coding and
histogram shifting,” Signal Process.,vol. 89, pp. Cite this article as :
1129-1143, 2009.
[2]. M.Johnson, P.Ishwar, V.M.Prabhakaran, V. Santhi, M. Abinaya, "Reversible Data Hiding in
D.Schonberg, and K.Ramchandran, “On com- Homomorphic Encrypted Domain by Monitoring
pressing encrypted data," IEEE Trans. Signal Ciphertext Group ", International Journal of Scientific
Process, vol. 52, no. 10 Oct 2004., pp. 2992-3006. Research in Computer Science, Engineering and
[3]. W. Hong, T. Chen, and H.Wu, “An improved Information Technology (IJSRCSEIT), ISSN : 2456-3307,
reversible data hiding in encrypted images using Volume 5 Issue 1, pp. 487-490, January-February 2019.
side match,” IEEE Signal Process. Lett., vol. 19, no. Available at doi : https://doi.org/10.32628/CSEIT1951102
4, pp. 199-202, Apr. 2012. Journal URL : http://ijsrcseit.com/CSEIT1951102

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


490
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951118

A Review : Venue Booking System


Shruti N. Ghule*1, Harish M. Mohurle2, Akshay P. Shegaonkar3, Zeba Parvin Abdual Saleem4,
Pradhnya S. Borkar5
1-4Priyadarshini J. L. College of Engineering, Nagpur, India
5 Assistant Professor in Priyadarshini J.L. College of Engineering, Nagpur, India
ABSTRACT
Ecommerce refers to both online retail as well as electronic transactions. Ecommerce has hugely increased in
popularity over the last decades, and in ways, it's replacing traditional brick and mortar stores. This application
enables you to book venues as per the choice of customer and his / her requirements. Venue booking system is a
web application in which service providers or a venue owner can registered their venues and customers can
book venues for events as per their location, availability and interest. This makes the workload simple for
customers to book a venue for an event they need to search for it without visiting actual place. In this
application, the user needs to be registered and log in for further advantages. The customer can search for
venues by specifying names or by giving the location if venues are present in that particular location.
Keywords: E-Commerce, Venue Booking

I. INTRODUCTION enter into a single system for reservation which


provides availability and profile of different venues..
In the 21st century most of the users are using If a particular venue is not available they have to go
internet for purchasing the product online, paying for to other website which is a time consuming process.
bills, booking for tickets etc. Due to rapid use of venue and is time consuming as the user checks with
internet the user have reduced purchasing their multiple services on their own. All this consumes lot
product through stores. E-Commerce based websites of time. Consider a user who wish to book a wedding
are increasing massively. The venue or hotel Industry hall for the wedding. The first thing that user does is
is one of the largest and most profitable industries in to visit the owner of wedding hall, discuss with him
most countries. It helps to oust the economy of both about the services offered by the hall and then
developed and developing countries. negotiate the deal for booking the hall.

People are always looking around’ when vacationing Furthermore there may be some services which are
for the cheapest venues for celebration’s and arrange not provided by the wedding hall owner, so for that
bookings in advance to guarantee their stay for the the user has to inquiry and get there services from a
period of their celebration. However these activities third party, such services can be like decoration of
generally require individuals to do the bulk of the hall, orchestra, photographer etc. So in he proposed
search and booking. This is not only time consuming system we aim to gather the entire venue for all types
exercise, but can be frustrating and costly as well. of function and also the services. By this user will
have much more options under one roof.
The objective of the system is to reserve a venue for
functions, parties, weddings, meetings etc., user can

CSEIT1951118 | Received : 12 Feb 2019 | Accepted : 25 Feb 2019 | January-February -2019 [ 5 (1) : 491-494 ]
491
Shruti N. Ghule et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 491-494

II. LITERATURE SURVEY people are viewing it and more and more people are
queuing.
“LARS: A Location-Aware Recommender System”
authored by Justin J. Levandoski ,Mohamed Sarwat, In [4] the author tells about Position based Services
Ahmed Eldawy, Mohamed F. Mokbel contains a which offers many benefits to the mobile consumers
position alert recommender structure that uses to recover the information about their current
position-based ratings to generate references. LARS position and manner that data to get more useful
achieves Customer rating positions through user information near to their position. With the help of
partitioning, a technique that influences GPS in phones and through Net Services using GPRS,
recommendations with ratings spatially close to Position based Facilities can be fulfilled on Android
querying users in a manner that maximizes system based smart phones to provide these value-added
scalability while not sacrificing recommendation services like handset profiling, scheduling alarms.
quality. Position based services are a general class of
computer program level services used to include
“Position Detection and Tracking System” published specific controls for position and time data as control
by Mahesh Kadibagil, and Dr. H S Guruprasad features in computer programs.
proposed a Self-directed location recognition and
tracking method which enhances the correctness of III. PROPOSED
finding friends and family member’s locations by
using GPS and typical web technology. This design The main aim of the proposed concept is to develop a
includes a mobile client, a repository, a web client web application that will allow the user to search and
and a map service. The mobile user is used to book the menu. The application should be able to
discover location and conduct a notification SMS to provide a platform for venues providers to showcase
user when his/her friends or family members come their services. There should be the options for
nearby the user’s zone of path. This location material disbursing auto discount for existing customers. The
can be sent to the server and the same information application should provide details of all booked
can be achieved and observed using the web venues with dates and user can query if they want to.
consumer by other customers. Following fig1. shows work flow VBS which has two
parts service provider (venue owner) and second is
In [3] the author wants to convey that with user having their own functionality of booking venue
the hasty improvement of Internet, people's day-to- or registering the venue.
day life has basically been attached from the Internet.
After the Internet, without resolution, whether
entities or industries, both clients and industrialists,
are made on the network of expectation, this new
expectation is to get through the network of the
greatest things the inexpensive, most practical to find
the fastest information, the most money-making
products to sell up. As people's lives are getting better,
it comes with many lifestyle changes. As the film are Fig1. Work Flow of VBS
still the same and has not changed predilections and
as more and more film are churned out more an more

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


492
Shruti N. Ghule et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 491-494

• Admin
Admin can add, update the categories in which a
venue will be added. He/ She can manage all the
functionality related to user and venue owner.
The admin acts as a intermediate between this
two modules.
• Booking
The booking module will handle the booking
functionality also it manages all functions related
to the enquiries of different venues and
availability of dates according to bookings.
Fig 2. DFD for Customer • Category
Above data flow diagram(DFD) i.e. fig2 will show This module will used to saving the categories
how user can interact with the application by sign in, from admin module. It contains different
sign up, searching venues etc. categories related to venues which can be decided
by venue owner according to his / her venue.
• Discount
This module will handle the functionality related
with discount. It provides discounts to the
existing user which will decided by venue owner.
• Gateway
This is the dummy implementation of gateway
for payments. It would be the full payment to
owner or some amount of down payment.
• Messages
This will used for showing and storing the
Fig 3. DFD for Venue Owner inquiry messages. This is basically act as enquiry
Fig3 will show how venue owner can interact with section where user can ask their inquiries related
the application by registering their venues and to venues.
provide various services as per their venue. • User
The user module will provide functionality for
This application will consists of following modules. signup and login etc. This will be customer
1) Admin Module section.
2) Booking Module • Venue
3) Category Module The venue module will used for showing and
4) Discount Module managing the venues which are registered and
5) Payment Gateway their information and services provided by venue
6) Inquiry Module owner.
7) Customer Module • DAO
8) Venue Owner Module Data access object will used for managing the
Following information shows functionality of database connectivity
different modules.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


493
Shruti N. Ghule et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 491-494

The proposed system will be implanted as a web Pages 196-199 July 16-17, 2011 IEEEComputer
based application on J2EE platform of Java with Society Washington, DC, USA ©2011 ISBN:978-
MySQL as database. 0-7695-4443-4 10.1109/ISCCS.2011.61
[4]. VEDANG MOHOLKAR,PRATHAMESH
Thus we have been able to design a venue booking HULE,MANDAR KHULE,SUMIT SAURABH
system and have been able to carry on partial computer department,pvpit,pune,India
implementation as. “AUTOMATED LOCATION BASED
SERVICES”, International Journal of Advanced
IV. CONCLUSION Research in Computer Science and Software
Engineering, Volume 4, Issue 2, February 2014
In this paper we discuss about managing different ISSN: 2277 128 X.
venues according to user need. It helps user to book
venues as per their requirements. It can be different Cite this article as :
halls, lawns, meeting rooms etc. The user can show
their interest by booking venues. This application Shruti N. Ghule, Harish M. Mohurle, Akshay P.
will overcome the problem of previous system in Shegaonkar, Zeba Parvin Abdual Saleem, Pradhnya S.
which all the work were done manually. It gives Borkar, "A Review : Venue Booking System",
better results by booking venues directly without International Journal of Scientific Research in
visiting the venue. Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
V. REFERENCES 5 Issue 1, pp. 491-494, January-February 2019.
Available at doi :
[1]. Justin J. Levandoski ; Mohamed Sarwat ; Ahmed https://doi.org/10.32628/CSEIT1951118
Eldawy ; Mohamed F. Mokbel “LARS: A Journal URL : http://ijsrcseit.com/CSEIT1951118
Location-Aware Recommender System” ICDE
'12 Proceedings of the 2012 IEEE 28th
International Conference on Data Engineering
Pages 450-461 April 01 - 05, 2012 IEEE
Computer Society Washington, DC, USA ©2012
ISBN: 978-0-7695-4747-3 10.1109/ICDE.2012.54
[2]. MAHESH KADIBAGIL PG SCHOLAR, DEPT.
OF ISE, DR. H S GURUPRASAD PROFESSOR
AND HEAD, DEPT. OF CSE, BMSCE,
BANGALORE, INDIA POSITION DETECTION
AND TRACKING SYSTEM IRACST -
International Journal of Computer Science and
Information Technology & Security (IJCSITS),
Vol. 4, No. 3,June 2014
[3]. BO HANG “DESIGN AND IMPLEMENTATION
OF CINEMA ONLINE BOOKING SYSTEM”,
ISCCS ‘ 11 Proceedings of the 2011 International
Symposium on Computer Science and Society,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


494
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951103

A Review on IOT Based Garbage Monitoring System


Himanshu Gupta1, Amit Chaturkar1, Abhay Bagde1, Hritik Raj1, Ankur Hemane1, Prof. Prerana Deshmukh2
1 BE Students, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,
Maharashtra, India
2Assistant Professor, Department of Computer Technology, Priyadarshini College of Engineering, Nagpur,
Maharashtra, India
ABSTRACT

Ordinarily, in our city, we see that the garbage bins or dustbins put at open spots are over-burden. It makes
unhygienic conditions for individuals just as grotesqueness to that place leaving a terrible stench. To dodge
every single such circumstance we are going to execute a venture called IoT Based Smart Garbage and Waste
Collection Bins. As the populace is expanding step by step, the earth ought to be spotless and sterile. In a large
portion of the urban areas, the flooded garbage bins are making an unhygienic situation. This will additionally
prompt the emerging of various kinds of maladies. This will corrupt the way of life. To beat these circumstances
a productive savvy squanders collection framework must be created. As the extent of IoT is creating step by
step successful techniques can be discovered effectively. Different plans were proposed and have points of
interest just as drawbacks. This paper is a review dependent on Smart Waste Collection System dependent on
IoT.
Keywords : Internet of Things, Smart Garbage Monitoring, Smart City, Microcontroller

I. INTRODUCTION the fuel utilization, CO2 outflows and city traffic


decrease. Taking into account that the courses will be
The garbage collection issue on megalopolis turned advanced so as to keep away from a truck moving to a
into a genuine issue crosswise over numerous urban vacant waste. For instance to send a truck to a course
communities around the globe. One of the incredible in which the whole of the occupations of the dumps
regions of concentrate inside Smart City and Internet don't extrapolate the limit of the truck, which will
of Things (IOT), includes the proposition of answers permit enhancement of the quantity of trucks
for regular issues of the urban communities, one of utilized in a given territory. The metropolitan waste
these issues is accurately the Garbage Collection. collection has just been streamlined through
Pointers demonstrate that Brazil has a low exhibition programming and numerical techniques [3], anyway
with regards to the collection and transfer of urban the outcomes were not agreeable, because of the
waste [1]. substantial measure of factors that brought about an
incredible computational time which may make the
The inquiries identified with garbage collection and utilization of remote sensors difficult and reasonable.
that should be unraveled are a few, in this work the As per [4] as the Internet and different innovations
issue that will be tended to is simply the one that kept on creating and advance in the principal decade
bargains with the collection itself, that is, the age of of the twenty-first century, a few arrangements rose
the courses for the trucks so as to permit decrease in up out of monsters showcases that made Internet of

CSEIT1951103 | Received : 13 Feb 2019 | Accepted : 25 Feb 2019 | January-February -2019 [ 5 (1) : 495-500 ]
495
Himanshu Gupta et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 495-500

Things a suitable alternative for a decent number of developing and populated urban communities as it
urban communities. contains contaminants which result in flimsiness,
issue and uneasiness to the biological system.
The board of waste is a major testing issue in urban Numbness and absence of neatness are ruining nature.
regions for the greater part of the nations all through The correct waste expulsion and the executives are
the world and is seen in a large portion of the incredibly viable to improve the wellbeing and
creating nations than in the created nations. A prosperity of the city's populace.
productive administration of waste is a prerequisite
for keeping up a perfect and green condition as there The primary objective of this paper is to take a shot at
is increment in a wide range of squanders tossed by natural issues because of ill-advised waste transfer
numerous spots like mechanical, agrarian, home and unravel them for better wellbeing and
waste, and so on. Squander collection and reusing is cleanliness of the general population. The proposed
done through different innovations. Collection of framework fits into the classification of IoT
data is huge and unwieldy. The present development connected to outer and open situations and it satisfies
in country with vast local location and an interest for the accompanying essential prerequisites of IoT
modernization in the city makes a testing administrations:
undertaking for waste administration individuals [5].
• Unwavering quality: Communication is crucial in
India especially creates around 1, 33,760 tons of IoT for administration provisioning, connected to
Municipal strong waste (MSW) the executives every the outside and open condition. Dependable and
day, of which roughly 91,152 tones are gathered, and solid correspondence is required arranged by
a colossal entirety of cash is spent on collection [6]. completing correspondence between gadgets as
World waste generation is relied upon to be roughly this sort of IoT has an extensive administration
27 billion tons for each year by 2050, 33% of which space. Consequently, the receptacle utilized in
will originate from Asia, with significant the proposed framework associated with one
commitments from China and India [7]. Squander age another, in light of a remote work arranges
in urban regions of India will be 0.7 kg per individual (WMN), giving dependability.
every day in 2025, around four to multiple times • Portability: IOT gadgets might be required to
higher than in 1999 [8]. move in the external environment. The proposed
framework works with a battery rather than the
II. MOTIVATION AND BACKGROUND changeless power source, bringing about an
extraordinary dimension of versatility. The
In the previous history, since the human populace versatility of the framework is secured with a
thickness and dimensions of misuse of characteristic power-based power supply.
assets were less, the measure of strong waste created • Administration Continuity: Data communications
was noteworthy. Be that as it may, the expansion in and administrations ought to be directed
the populace in this day and age has expanded the flawlessly whenever and any area in IoT with an
garbage. To keep nature spotless and solid, there is a extensive administration space. Bins are situated
necessity of legitimate garbage transfer. Inappropriate at the customary space of separation to guarantee
garbage transfer raises contamination, medical the administration progression.
problems, different dangers and in result it influences • Client Convenience: The presentation of IoT has
nature. Contamination extremely influences the improved client comfort. For client facilitate, the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


496
Himanshu Gupta et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 495-500

proposed framework lessens the method postpone Another technique [10], there are different
the time of the existing garbage gathering dustbins situated all through the city or the Campus.
frameworks, which empowers clients to set free These dustbins are furnished with ease implanted
their garbage without a long pause and auspicious gadget which helps in following the dimension of the
end of filled bins. garbage bins and a one of a kind ID will be
• Vitality Efficiency: IoT connected to outer and accommodated each dustbin in the city so it is
open conditions depends on a dependable on the anything but difficult to recognize which garbage
framework and requires versatility, causing a lot container is full. The undertaking module is isolated
of vitality utilization. To take care of this issue, into two sections Transmitter area and recipient
the sensors work utilizing vitality proficient segment. Here in the transmitter segment we are
procedures, expanding their battery lifetimes. utilizing 8051 microcontrollers, RF Transmitter and
sensors these are appended to the dustbin. Where
III. LITERATURE REVIEW
sensor is utilized to identify the dimension in the
dustbin whether the dustbin is full or void.
The garbage the executives in urban communities
must be successfully and productively actualized. The
Another strategy [11] is that, when the garbage
different recommendations were advanced and some
achieves the edge level ultrasonic sensor will trigger
of them effectively actualized. Be that as it may, it
the GSM modem which will constantly alarm the
can't be considered as a powerful one. Therefore, an
required expert until the garbage in the dustbin is
overview was done among various proposition and
squashed. When the dustbin is squashed, individuals
this review paper incorporates study among various
can reuse the dustbin. At customary interims dustbin
strategies for Waste Collection System dependent on
will be squashed. In this strategy, GSM 900A modem
IoT.
is utilized to send the messages.

The paper [9] proposed waste collection framework


Another strategy for garbage the executives is
depends on waste dimension information from
presented [12] as pursues. A dustbin is interfaced
trashcans in a metropolitan territory. The
with microcontroller based framework having IR
information gathered by sensors is sent over the
remote frameworks alongside focal framework
Internet to a server where it is put away and handled.
demonstrating current status of garbage, on portable
The creator gathered information is then utilized for
internet browser with HTML page by Wi-Fi.
checking and enhancing the every day choice of
Consequently the status will be refreshed on to the
trashcans to be gathered, figuring the courses in like
HTML page.
manner. Consistently, the specialists get the recently
determined courses in their route gadgets. The key
In paper [13] Infrared sensor (IR sensor) is utilized
element of this framework is that it is intended to
which is a multipurpose sensor, which can recognize
gain as a matter of fact and to settle on choices on the
the dimension of garbage. IR sensor produces the
every day squander level status as well as on future
light, which is imperceptible to stripped eye yet the
state estimate, traffic clog, adjusted cost-proficiency
electronic segments can distinguish it.
capacities, and other influencing factors that from the
earlier people can't predict.
In Paper [14] System screens the garbage bins and
educates about the dimension of garbage gathered in
the garbage bins by means of a page. For this the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


497
Himanshu Gupta et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 495-500

framework utilizes ultrasonic sensors put over the partition from wrong dimension estimation and are
bins to identify the garbage level and contrast it and interfaced with the microcontroller.
the garbage bins profundity. The framework makes
utilization of AVR family microcontroller, LCD To distinguish awful stench a gas sensor is set at the
screen, Wi-Fi modem for sending information and a base of the dustbin and is adjacent to interfaced with
signal. the controller to remember it off the waste filled in
the dustbin. The two sensors send the signs to the
In paper [15] guarantees the cleaning of dustbins controller. Arduino accumulates information gotten
soon when the garbage level achieves its most by the gatherer and trade nearby page through the
extreme. In his administration framework IOT as the Ethernet shield.
working in the field for arranged radio-recurrence
distinguishing proof (RFID), following the collection
vehicle, Dustbin observing and other developing
detecting advancements.

Creators in [16] consider dynamic booking over a lot


of recently characterized collection trips. The
fundamental target of the methodology is to limit the
complete operational and fixed truck costs.

A numerical detailing procedure is proposed in [17]


building up an arrangement of administrative regions, Figure 1. System Architecture
characterizing directing, and planning booking
mulling over conceivable new elective arrangements The ultrasonic sensor is utilized to check the
in dealing with the framework in general. dimension status of a dustbin so as to pick whether it
is full or passed greatest limit esteem. Dynamic status
In [18] creators assess dynamic arranging strategies of dustbin is appeared on the site using relationship
connected for the waste collection of underground through the Ethernet shield. Checking the page will
bins. Demonstrate diminishes the measures of carbon help the waste collection office with following for the
dioxide discharged in the earth from trucks by correct area and proportion of the junk. The waste
making dynamic steering progressively powerful. vehicles would then have the ability to debilitate the
garbage from a specific domain.
IV.PROPOSED METHODOLOGY
The farthest point of GSM module is to establish a
In this framework, dustbins are organized at level 1 of connection on the waste social event division when it
a building made under savvy city activity. It will gets full. The garbage bins put at level 1 of the
gather the misfortune through keen pipe framework structures can be effectively emptied utilizing engines
set in the building. The perceptive dustbins are to pivot it by 180⁰ while the gatherer truck is at an
interfaced with the web through GSM to get the area under dimension 1 (ground level).
present status. Two sensors are settled and no more
lifted inspiration driving the dustbin to keep up a key

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


498
Himanshu Gupta et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 495-500

V. CONCLUSION [8] Modak P, Jiemian Y, Hogyuan Yu, Mohanty CR


2010 Municipal solid waste management
Despite the fact that online pharmacy is more subject turning waste into resources in Shanghai
than nearby pharmacy particularly in remote regions, manual: a guide for sustainable urban
it ought to be planned to build up standards and development in 21st century, pp. 1-36
direction like transfer of filtered solutions. Patients [9] Kumar JS, Subbaiah KV, Rao PVVP. 2014
ought to know on utilization of doctor prescribed Municipal solid waste management in India.
prescriptions. It is great to embrace the plans and Austr J. Eng. Res. 2, 1-8 (DOI: 10.7603/s40632-
projects of driving nations to structure and build up 014-0001-4)
the e-pharmacy framework. [10] Jose M. Gutierrez, Smart Waste Collection
System Based on Location Intelligence.
VI. REFERENCES Procedia Computer Science 61 ( 2015 ) 120 –
127..
[1] Melo, Alexander Bento, et al. "Optimization of [11] Parkash, Prabu, IoT Based Waste Management
Garbage Collection Using Genetic Algorithm." for Smart City. IJECS Vol. 4, Issue 02 February
Mobile Ad Hoc and Sensor Systems (MASS), 2016.
2017 IEEE 14th International Conference on. [12] Monika K, Smart Dustbin-An Efficient Garbage
IEEE, 2017. Monitoring System. IJECS Volume 6 Issue No.
[2] C. A. Mucelin, “LIXO E IMPACTOS 06 June 2016.
AMBIENTAIS PERCEPTÍVEIS NO [13] S.S. Navghane, IoT Based Garbage, and Waste
ECOSSISTEMA Collection Bin. IJARECE Volume 5, Issue 5,
[3] URBANO Garbage and perceptible May 2016.
environmental impacts in urban ecosystem,” [14] Meghana KC, Dr. K R Nataraj IOT Based
vol. 20, no. 1, pp. 111–124, 2008. Intelligent Bin for Smart Cities. IJRITCC May
[4] R. Fujdiak, P. Masek, P. Mlynek, J. Misurec, 2016.
and E. Olshannikova, “Using Genetic [15] Abdul Atif Khan, Study Of Smart City Using
Algorithm for Advanced Municipal Waste Internet Of Things. Ijritcc March 2016.
Collection in Smart City,” 2016. [16] Vishesh Kumar Kurrel, Smart Garbage
[5] M. N. K. Boulos and N. M. Al-shorbaji, “On the Collection Bin Overflows Indicator using the
Internet of Things , smart cities and the WHO Internet of Things. Volume 3, Issue 05 May
Healthy Cities,” pp. 1–6, 2014. 2016.
[6] Jain, Aaditya, and Ranu Bagherwal. "Design [17] J. Q. Li, D. Borenstein, P. B. Mirchandani,
and implementation of a smart solid waste “Truck Scheduling forSolid Waste Collection in
monitoring and collection system based on the City of Porto Alegre, Brazil”, Omega,
Internet of Things." Computing, Elsevier, 2008, vol. 36, pp. 1133-1149.
Communication and Networking Technologies [18] P. T. R. Ramos, M. I. Gomes, and A. P. B.
(ICCCNT), 2017 8th International Conference Povoa, “Assessing and improving Management
on. IEEE, 2017. Practices when Planning PackagingWaste
[7] Mirchandani, Sahil, et al. "IoT enabled Collection Systems”, Resources Conservation
dustbins." Big Data, IoT and Data Science, 2017 and recycling, Elsevier, 2014, vol. 85, pp. 116-
International Conference on. IEEE, 2017. 129.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


499
Himanshu Gupta et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 495-500

[19] A. Stellingwerff, “Dynamic Waste Collection:


Assessing the Usage of Dynamic Routing
Methodologies”, Master Thesis, Industrial
Engineering & Management, University of
Twente, Twente Milieu, 2011.

Cite this article as :

Himanshu Gupta, Amit Chaturkar, Abhay Bagde,


Hritik Raj, Ankur Hemane, Prof. Prerana Deshmukh,
"A Review on IOT Based Garbage Monitoring
System", International Journal of Scientific Research
in Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 495-500, January-February 2019.
Journal URL : http://ijsrcseit.com/CSEIT1951103

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


500
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951122
A Survey on Rice Crop Yield Prediction in India Using Improved Classification
Technique
Kolin Sukhadia*1, M. B. Chaudhari2
*1PG Scholar, Department of Computer Engineering, Government Engineering College Modasa, Modasa, Gujarat, India
2Head of Department, Department of Computer Engineering, Government Engineering College Modasa, Modasa, Gujarat,
India

ABSTRACT

India is an agricultural country. Agriculture is the important contributor to the Indian economy. There are
many classification techniques like Support Vector Machine(SVM), LADTree, Naïve Bayes, Bayesnet, K Nearest
Neighbour(KNN), Locally Weighted Learning(LWL) on rice crop production datasets. They have some
drawbacks like low accuracy and more errors. To achieve more significant result, To increase classification
accuracy and reducing classification errors, our research uses classification method Bayesnet based adaboost
will be proposed in work. Rice crop yield depend on environment’s parameters like Rainfall, minimum
temperature, average temperature, Maximum temperature, Vapour Pressure, potential evapotranspiration,
reference crop evapotranspiration, cloud cover, wet day frequency for the kharif season. our dataset containing
these environmental parameters for accurate prediction of Rice crop yield.
Keywords : AdaBoost Technique, BayesNet Classifier, Rice Crop Yield Prediction, Kharif Season

I. INTRODUCTION used by Government, policy makers, agro-based


industries, traders and agriculturists. The
Rice crop production contributes more than 40% to Government use crop yield prediction in
overall crop production. Its production is reliant on procurement, distribution, buffer-stocking, import
climatic conditions. Improving the ability to predict export, price fixation and marketing of agricultural
crop productivity in under different climatic commodities[8]. Agriculture is the most important
parameters like rainfall, temperature, cloud cover, branch in Indian economy and 70 percentage of rural
vapour pressure, potential evapotranspirsation etc. population livelihood depends on agricultural work.
can assist farmers and other stakeholders in making Crop yield depends on environment’s factors like
important decisions in terms of agronomy and crop precipitation, temperature, evapotranspiration, etc.
choice[43]. The prediction of agricultural yield is a Based on previous experience, Farmers cultivate crop.
challenging and desirable task for every nation. But nowadays, the uncertainty increased in
Prediction of yield can be done based on historic crop environment. So, accurate analysis of historic data of
cultivation data and environment data and environment parameters should be done for
environment data crop yield can be predicted. Crop successful farming. To get more harvest, we should
yield prediction is the forecasting of harvest in Kg per also do the analysis of previous cultivation data. The
Hectare. To analyse data various data mining Prediction of crop yield can be done based on historic
technique can be used. Crop yield prediction can be

CSEIT1951122 | Received : 15 Feb 2019 | Accepted : 26 Feb 2019 | January-February -2019 [ 5 (1) : 501-507]
501
Kolin Sukhadia, M. B. Chaudhari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 501-507

crop cultivation data and weather data using data Regression AdaBoost (Adaptive Boosting), Support
mining methods[14]. Vector Machine Regression and a Modified
Nonlinear Regression (MNR) equation and then
There are many classification techniques like Support comparing the modified equation with the other
Vector Machine(SVM), Naïve Bayes, BayesNet, K three methods to check its accuracy For Aus and
NearestNeighbour (KNN), Locally Weighted Boro Rice our modified Nonlinear Regression
Learning(LWL), LAD Tree are applied on rice crop equation not only yields better RMSE, MSE and MAE
production datasets. The provision of accurate and values, it also has the highest Rsquare value, which
timely information can assist farmers to make the proves that the MNR equation is the best fit for this
best decision for their cropping situations This could study. this study also proves that weather conditions
benefit them to attain greater crop productivity[42]. play very vital role in predicting the yield. In [2],
This study examines the application of machine author used classifiers for research are J48, LWL,
learning for the prediction of rice crop yield. The LAD Tree and IBK. The experimental results showed
current study use the most rice producing that J48, LADTree achieved the highest accuracy,
states(Haryana, Punjab, Tamilnadu, Andhra Pradesh, sensitivity and specificity. In [3], author used
Bihar). The dataset parameters are state name, district classifiers were BayesNet and NaiveBayes. The results
name, crop year, Rainfall, minimum temperature, showed that the performance of BayesNet was much
average temperature, maximum temperature, vapour better compared with NaiveBayes for the dataset. In
pressure, reference opevapo transpiration, [4], proposed approach builds a model to predict the
Potential evapotranspiration, diurnal temperature weather parameters applying Neural Networks, then
range, area, production, yield, class. The dataset was estimates the rice yields applying Support Vector
then sorted on the basis of yield to classify the Regression. In [5], The authors have used pre-
records in to low, moderate and high. The low yield processed data that was clustered using k-means
was from 0. 1 to 1 tonnes/hectare, moderate from 1 to clustering algorithm. The association rule mining
2 tonnes/hectare and high from 2 to 5 process will apply on clustered data to find the rules.
tonnes/hectare. The training phase ends with number of generated
rules. In the testing phase, the yield value is predicted
The further document is as follows. In Section 2 based on the generated rules. The parameters
(Literature Survey), the literature review done on considered for research were precipitation, minimum
Rice Crop Yield Prediction, Bayesnet Method, temperature, reference crop evapotranspiration
AdaBoost method and Rice Crop. In Section 3 average temperature, maximum temperature.
(Conclusion) conclusion derived based on literature production, area and yield for the Kharif season
survey. including months from June to November.

II. LITERATURE SURVEY This research discusses the experimental results


obtained by applying SMO classifier using the WEKA
A. CROP YIELD PREDICTION tool on the dataset of 27 districts of Maharashtra state,
India in [6]. In [7], The parameters considered for the
IN [1] Umid Kumar Dey, Abdullah Hasan Masud, present study were precipitation, reference crop
Mohammed Nazim Uddin does research to predict evapotranspiration average temperature, minimum
the yield of rice in the regions of Bangladesh during temperature, maximum temperature, area,
the aforementioned seasons using Multiple Linear production and yield for the Kharifs season (June to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com | UGC Approved Journal [Journal No : 64718]
502
Kolin Sukhadia, M. B. Chaudhari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 501-507

November) for the years 1998 to 2002. In this Nigeria. The climate data used was for a period of 36
research WEKA tool is used. A Multilayer Perceptron years from 1980 to 2015. This study[14] describes the
Neural Network was implemented as ANN. Cross role of data mining in Agriculture and crop yield
validation method was used to validate data. In [8], prediction. This paper also describes Groundnut crop
The proposed method uses yield and weather data yield prediction analysis and Naive Bayes Method.
collected from United States Department of This Study [15] describes multi-model heterogeneous
Agriculture. The parameters in the dataset are rainfall, ensemble approach. This study [16] examine the
humidity, temperature, yield and. regression-based introduction about Bayesian network, Directed
algorithms to predict the crop yield. These were acyclic graph, Conditional Probability table. This
multivariate polynomial regression, support vector study [17] examine various data mining classification,
machine regression and random forest. support clustering and association rule mining technique.
vector machine regression to obtain the best possible This study [18] examine detail information about
results for predicting the crop yield. In [9], The naïve bayes classification technique. This study [19]
prominent classification methods in this study are the examine Bayesian network classification technique,
Support VectorMachine(SVM) and Naive Bayes. In structure learning technique, parameter learning
this paper, AdaSVM and AdaNaive are the proposed technique. This study[20] examine various adaptive
ensemble model use to project the crop production Learning Algorithms for Bayesian Network Classifiers.
over a period of time. This ensemble model is This study[21] examine Bayesian Network Classifiers
compared with Naive Bayes and SVM methods. in Weka. This study[22] examine the introduction
accuracy and the classification error are two about bayesian approach to Probability and Statistics.
parameters used for output predictions. The finding This study[23] examine bayesian method for
yields that AdaSVM and AdaNaive are agreeable than constructing probabilistic networks from databases.
SVM and Naive Bayes for the data set analyzed. In they focus on constructing Bayesian belief networks.
this study[10], regression based adaptive boosting This study[24] examine the overview of rice crop
prediction model is presented, There are three coastal yield prediction. Examines Different data mining
districts belonging to Odisha are located in India techniques utilized for foreseeing rice crop yield. In
using the datasets of Rabi and Kharif seasons. This this study[25], there is overview of rice crop yield
study experiments on weak regressors, such as: SVR prediction. This paper Examines Different data
regression, linear, lasso, ridge proposes strong mining techniques utilized for foreseeing rice crop
predictors by avoiding the shortcomingss of yield. In this study[26], rice yield prediction is
individual weak regressors. In [11] study examines performed by back propagation algorithm. In this
the variations of climate variables and rice yield and study[27], there is Comparative Study of
examines the relationships using multiple linear Classification Algorithms for Forecasting Rainfall.
regression, principal component analysis, and support This study[28] examine Modeling Rainfall Prediction
vector machine (SVM) analysis in southwest Nigeria. Using Data Mining Method: A Bayesian Approach
In this study[12], Adaptive boosting of weak data mining method for modeling rainfall prediction.
regressors for forecasting of crop production In [29], Rice Crop Yield Forecasting of Tropical Wet
considering climatic variability. In this study[13] and Dry Climatic Zone of India using data mining
examines the variations of climate variables and rice Techniques is applied. For data set results show that
yield and quantifies the relationships using multiple J48 and LADTree achieved the highest accuracy,
linear regression, principal component analysis, and sensitivity and specificity. Classification by LWL
support vector machine (SVM) analysis in southwest classifier displayed the lowest accuracy, sensitivity

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com | UGC Approved Journal [Journal No : 64718]
503
Kolin Sukhadia, M. B. Chaudhari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 501-507

and specificity results. In [30], This study presents an Accuracy is highest in case of Soybean. A study by
analysis using data mining techniques for estimating [37] has applied association rules for Tamilnadu. Rice
the future yield prediction in tea cultivation with crop yield predicted using regression analysis,
climatic change trends observed in last 30 years experiment implemented in MATLAB tool. Yield
(1977-2006). The patterns of crop production in predicted from Annual Rainfall, Area under
response to the climatic (rainfall, temperature, Cultivation, and Food Price Index. Relationship
relative humidity, evaporation and sunshine) effect established between AR, AUC and FPI[38]. A study
across the four tea growing regions (South Bank, [39] described literature review in tabular form. Some
North Bank, Upper Assam and Cachar) of Assam research study only summarized the data mining
were developed using Multiple Linear Regression techniques for crop yield prediction. Support Vector
(MLR) technique. Machine, Random Forest, Neural Network, REPTree,
Bagging, Bayes algorithm described in the study [40].
The tea production estimation equations developed These classifiers used for Soybean Crop Yield
for the regions were validated for the future yield Prediction and results are compared. Bagging
prediction (2007, 2009 and 2010) and were found to algorithm gives maximum accuracy for soybean crop
be significant. yield prediction among REPTree, Random Forest,
Neural Network, SMO classifiers. Bayesian modelling
A study by [31] has done Crop Yield Prediction of used for wheat crop yield prediction using small
Wheat Using Fuzzy C Means Clustering and Neural dataset[41]. A study also identified drought related
Network. A research work [32] has implemented agronomic factors that significantly affect the
artificial neural networks (ANNs) for predicting crop variation in the rain fed wheat yield and relative
yields for an agricultural region in Nepal. In research, importance of these factors.
Agricultural data was collected from paddy field
cultivation in the Siraha district, an eastern region in Applications
Nepal. climatic parameters are rainfall, maximum
temperature and minimum temperature, fertilizer use Predicts crop yield production of rice for most Rice
were also used as input values. The result shows that producing states, India.
the trained neural network produced a minimum Predicting productivity of crop in various climatic
error. In [33] researcher has implemented neural conditions can help farmer and other partners in
networks for rice yield prediction in mountainous essential basic leadership as far as agronomy and
regions. A study by [34] reaching review of research product decision. It also helps to increase the yield of
devoted to applications of machine learning in crop.
agrarian generation frameworks. The review reports
the application of machine learning techniques with III. PERFORMANCE EVALUATION METRICS
artificial neural networks, Bayesian networks,
support vector machines and k-means. In [35], Accuracy: It is defined as the overall success rate of
researchers predict Rice Production in Phimai the classifier.
District, Thailand Decision Support System Using Accuracy = (TP + TN) / (TP + FP + FN + TN).
Artificial Neural Network. A study by [36] Where, True Positive (TP) depicts the number of
implemented C 4. 5 Algorithm to predict yield of instances where system detects for condition when it
crops like paddy, maize, wheat, soybean. Interactive is really present[1].
web pages developed for crop yield prediction.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com | UGC Approved Journal [Journal No : 64718]
504
Kolin Sukhadia, M. B. Chaudhari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 501-507

True Negative (TN) depicts the number of instances V. REFERENCES


where system does not detect a condition when it is
absent[1]. [1]. Shruti Mishra, Priyanka Paygude, Snehal
False Negative (FN) depicts the number of instances Chaudhary, Sonali Idate"Use of Data Mining in
where system does not detect a condition when Crop Yield Prediction", Proceedings of the
actually it is present[1]. Second International Conference on Inventive
False Positive (FP) depicts the number of instances Systems and Control (ICISC 2018), pp. 796-802,
where system detects a condition when it is really 2018
[2]. Umid Ku mar Dey, Abdullah Hasan Masud,
absent[1].
Mohammed Nazim Uddin, "Rice Yield Prediction
RMSE(Root Mean Square Error): It is defined as the
Model Using Data Mining", International
difference between the values predicted by the model
Conference on Electrical, Computer and
and the actual values noted[1].
Communication Engineering (ECCE), pp. 321-
MAE(Mean Absolute Error): It is another factor in
326, 2017
statistics which measures the difference between two
[3]. Niketa Gandhi, Leisa J. Armstrong, Manisha
continuous variables[1]. Nandawadekar, "Application Of Data mining
RAE(Relative Absolute Error): This measure gives the Techniques For Predicting Rice Crop Yield Semi-
total absolute error between the variables[1]. Arid Climate Zone Of India", 2017 IEEE
International Conference on Technological
IV. CONCLUSION AND FUTURE SCOPE Innovations in ICT for agriculture and Rural
Development, pp. 116-120, 2017
In rice crop yield prediction, various classification [4]. Ayush Shah, Akash Dubey, Vishesh Hemnani,
techniques are applied like support vector machine, Divye Gala, D. R. Kalbande, "Smart Farming
naïve bayes, bayesnet, locally weighted learning, System: Crop Yield Prediction Using Regression
ladtree etc. are applied on rice crop production Techniques", Proceedings of International
datasets. They have some drawbacks of low Conference on Wireless Communication, pp. 49-
classification accuracy and high errors. To get more 56, 2018
significant result and to get better classification [5]. Niketa Gandhi, Leisa J. Armstrong, Owaiz Petkar,
accuracy and to reduce errors like RMSE(Root Mean Amiya Kumar Tripathy, "Rice Crop Yield
Square Error), MAE(Mean Absolute Error), Prediction in India using Support Vector
Machines", 2016 International Joint Conference
RAE(Relative Absolute Error), boosting technique
on Computer Science and Software Engineering
Adaptive boosting technique is used in conjuction
(JCSSE)
with classification method Bayesnet in my research
[6]. Niketa Gandhi, Leisa J. Armstrong, Owaiz Petkar,
work and that will determine better accuracy of
"Predicting Rice Crop Yield Using Bayesian
prediction of rice yield in future. This survey
Networks", 2016 Intl. Conference on Advances in
concludes that the more accurate rice crop yield
Computing, Communications and Informatics
prediction model can be constructed by using (ICACCI), pp. 795-799, 2016
different data mining techniques. [7]. Akter Hossain, Mohammed Nazim Uddin,
Mohammad Arif Hossain, Yeong Min Jang,
"Predicting Rice Yield For Bangladesh By
Exploiting Weather Condition", 2017
International Conference on Information and

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com | UGC Approved Journal [Journal No : 64718]
505
Kolin Sukhadia, M. B. Chaudhari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 501-507

Communication Technology Convergence Journal of Applied Research in Higher Education,


(ICTC), pp.589-594, 2017 Vol. 10 Issue: 1, pp.61-75
[8]. Dipika H. Zala, M.B. Chaudhari, ” Review on Use [16]. ADNAN MASOOD, "Bayesian Networks", A
of “BAGGING” Technique in Agriculture Crop BRIEF INTRODUCTION
Yield Prediction”, International Journal for [17]. Jiawei Han, Micheline Kamber, "Data
Scientific Research & Development| Vol. 6, Issue Mining:Concepts and Techniques"
08, 2018, pp.675-677. [18]. https://en.wikipedia.org/wiki/Naive_Bayes_classif
[9]. Niketa Gandhi, Leisa J. Armstrong, "Applying ier
Data Mining Techniques to Predict Yield Of Rice [19]. Friedman, N., Geiger, D. and Goldszmidt M.
in Humid Subtropical Climatic Zone Of India", (1997), ” Bayesian network classifiers”, Machine
2016 International Conference on Computing for Learning 29, 139–164.
Sustainable Global Development, pp.1901-1906, [20]. Castillo, G. (2006) , ”Adaptive Learning
2016 Algorithms for Bayesian Network Classifiers”,
[10]. Sagar S. Nikam , ”A Comparative Study of PhD Thesis, Chapter 3.
Classification Techniques in Data Mining [21]. Bouckaert, R. (2007), ” Bayesian Network
Algorithms”, ISSN April 2015. Classifiers in Weka”, Technical Report
[11]. Narayanan Balakrishnan, Dr.Govindarajan [22]. Heckerman, D. (1996), ” A Tutorial on Learning
Muthukumarasamy, ”crop Production - Ensemble with Bayesian Networks”, Microsoft Technical
Machine Learning Model for Prediction”, Report MSR-TR-95-06.
International Journal of Computer Science and [23]. G. Cooper, E. Herskovits (1992), ”A Bayesian
Software Engineering (IJCSSE), Volume 5, Issue 7, method for the induction of probabilistic
July 2016, pp. 148-153. networks from data, ”Machine Learning. 9(4):309-
[12]. Subhadra Mishra, Debahuti Mishra, Gour Hari 347.
Santra, "Adaptive boosting of weak regressors for [24]. N.Gandhi, L.J. Armstrong and O. Petkar, “Rice
forecasting of crop production considering Crop Yield Prediction in India using Machine
climatic variability: An empirical assessment", Learning Techniques”, communicated, 2016.
Journal of King Saud University - Computer and [25]. Weka 3:Data Mining Software in Java, Machine
Information Sciences, pp. 1-23. Learning Group at the University of Waikato,
[13]. Philip G. Oguntunde, Gunnar Lischeid, Ottfried Official
Dietrich, "Relationship between rice yield and Web:http://www.cs.waikato.ac.nz/ml/weka/index
climate variables in southwest Nigeria using .html, accessed on 26th March 2016.
multiple linear regression and support vector [26]. S. Puteh, M. Rizon, M. Juhari, J. Nor Khairah, S.
machine analysis", August 2017, springer. Siti Kamarudin, B. Aryati, R. Nursalasawati, “Back
[14]. Haresh L. Siju, Pinal J. Patel, "Review on Crop propagation algorithm for rice yield prediction”,
Yield Prediction using Data Mining Focusing on Proceedings of 9th of the Ninth International
Groundnut Crop and Naive Bayes Technique", Symposium on Artificial Life and Robotics,
International Journal of Scientific Research in Beppu, Japan, Oita, pp. 586-589, 2004.
Computer Science, Engineering and Information [27]. Deepti Gupta, Udayan Ghose, " A Comparative
Technology, Volume 3 | Issue 1, ISSN : 2456- Study of Classification Algorithms for Forecasting
3307, pp.1947-1954 Rainfall ", IEEE 4th International conference on
[15]. Olugbenga Wilson Adejo, Thomas Connolly, Reliability, Infocom Technologies and
"Predicting student academic performance using Optimization (ICRITO) (Trends and Future
multi-model heterogeneous ensemble approach", Directions), 2015, pp. 1-6.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com | UGC Approved Journal [Journal No : 64718]
506
Kolin Sukhadia, M. B. Chaudhari Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 501-507

[28]. V. B. Nikam and B. B. Meshram, "Modeling [36]. Veenadhari, B. Misra, and C. D. Singh, "Machine
Rainfall Prediction Using Data Mining Method: A learning approach for forecasting crop yield based
Bayesian Approach", Fifth International on climatic parameters, " 2014 Int. Conf. Comput.
Conference on Computational Intelligence, Commun. Informatics Ushering Technol.
Modelling and Simulation, Seoul, 2013, pp. 132- Tomorrow, Today, ICCCI 2014, pp. 1-5, 2014.
136. [37]. E. Manjula, S. Djodiltachoumy, ” A Model for
[29]. Niketa Gandhi, Leisa J. Armstrong, “Rice Crop Prediction of Crop Yield”, International Journal
Yield Forecasting of Tropical Wet and Dry of Computational Intelligence and Informatics,
Climatic Zone of India using data mining Vol. 6: No. 4, March 2017, pp.298-305.
Techniques”, 2016 IEEE International Conference [38]. Sellam and E. Poovammal, "Prediction of Crop
on Advances in Computer Applications (ICACA) Yield using Regression Analysis, " vol. 9, no.
[30]. Rupanjali D. Baruah, R.M. Bhagat, Sudipta Roy, October, 2016.
L.N. Sethi, ”Use of Data Mining Technique for [39]. B. H. Dhivya, R. Manjula, S. B. S, and R.
Prediction of Tea Yield in the Face of Climate Madhumathi, "A Survey on Crop Yield Prediction
Change of Assam, India”, 2016 International based on Agricultural Data, " pp. 4177-4183, 2017.
Conference on Information Technology, pp. 265- [40]. Savla and A. Mandholia, "Survey of classification
269. algorithms for formulating yield prediction
[31]. Akanksha Verma, Aman Jatain, Shalini Bajaj, ” accuracy in precision agriculture, " 2015.
Crop Yield Prediction of Wheat Using Fuzzy C [41]. Yildirak, Z. Kalaylioglu, and A. Mermer,
Means Clustering and Neural Network”, "Bayesian estimation of crop yield function :
International Journal of Applied Engineering drought based wheat prediction model for tigem
Research ISSN 0973-4562 Volume 13, Number 11 farms, " Environ. Ecol. Stat., vol. 22, no. 4, pp.
(2018) pp. 9816-9821. 693-704, 2015.
[32]. Tirtha Ranjeet, Leisa Armstrong, ” An Artificial [42]. Mrs. K. E. Eswari, Ms. L.Vinitha, ” Crop Yield
Neural Network for Predicting Crops Yield in Prediction in Tamil Nadu using Baysian
Nepal”, Proceedings of AFITA 2014 - Copyright Network”, International Journal of Intellectual
ASICTA Inc, pp. 376-386. Advancements and Research in Engineering
[33]. B. J I , Y. SUN, S.YANG, J. WAN1, ” Artificial Computations, Volume-6, Issue-2, pp.1572-1576.
neural networks for rice yield prediction in [43]. Niketa Gandhi, Owaiz Petkar, Leisa J. Armstrong,
mountainous regions”, Journal of Agricultural "Rice Crop Yield Prediction Using Artificial
Science (2007), pp. 249–261. Neural Networks", 2016 IEEE International
[34]. Ankita Bissa, Meena Kushwaha , Mayank Patel, Conference on Technological Innovations in ICT
”A Survey of Machine Learning Appliacations In For Agriculture and Rural Development (TIAR
Decision Making To Improve Farming”, 2016), pp.105-110
International Journal of Computer Sciences and
Engineering, Vol.-6, Issue-9, Sept. 2018, pp. 669- Cite this article as : Kolin Sukhadia, M. B. Chaudhari, "A
683. Survey on Rice Crop Yield Prediction in India Using
[35]. Saisunee Jabjone, Sura Wannasang, ” Decision Improved Classification Technique ", International Journal
Support System Using Artificial Neural Network of Scientific Research in Computer Science, Engineering
and Information Technology (IJSRCSEIT), ISSN : 2456-
to Predict Rice Production in Phimai District,
3307, Volume 5 Issue 1, pp. 501-507, January-February
Thailand”, International Journal of Computer and
2019.
Electrical Engineering, Vol. 6, No. 2, April 2014,
Available at doi : https://doi.org/10.32628/CSEIT1951122
pp. 162-166.
Journal URL : http://ijsrcseit.com/CSEIT1951122

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com | UGC Approved Journal [Journal No : 64718]
507
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951131

A Review on Various Road Safety Systems


Chandrashekhar Rathod1, Sneha Wakde1, Roshan Punwatkar1, Rupesh Sharma1, Rupal Patil1, Prof. Rohan
Kokate2
1BE Student, Department Computer Science and Engineering, J. D. College of Engineering and Management,

Nagpur, Maharashtra, India


2 Assistant Professor, Department Computer Science and Engineering, J. D. College of Engineering and
Management, Nagpur, Maharashtra, India
ABSTRACT

The Safe System (SS) approach to road safety emphasizes safety-by-design through ensuring safe vehicles, road
networks, and road users. With a strong motivation from the World Health Organization (WHO), this
approach is increasingly adopted worldwide. Considerations in SS, however, are made for the medium-to-long
term. Our interest in this work is to complement the approach with a short-to-medium term dynamic
assessment of road safety. Toward this end, we introduce a novel, cost-effective Internet of Things (IoT)
architecture that facilitates the realization of a robust and dynamic computational core in assessing the safety of
a road network and its elements. In doing so, we discuss about the various existing mechanisms for road safety.
Keywords : Internet of Things, Smartphone, Road Safety, Arduino, G-Sensor, Accelerometer

I. INTRODUCTION star Corporation provides subscription based


communications, in–vehicle security, hand free
Road traffic crashes are one of the world’s largest calling, turn-by-turn navigation, and remote
public health and injury prevention diagnostics systems throughout the united states,
problem .According to the World Health Canada and China turn-by-turn navigation and road
Organisation (WHO), more people die on road in side assistance .The motivation behind the work
India than anywhere else in the world. At least 13 revolution is an attempt to make an embedded
people die every hour in road accidents in our system to bring a positive difference in the field of
country; the latest report of the national crime record road safety and road discipline.
Dureau reveals.
The work tackles some major causes of road accidents
In 2007, 1.14 lakhs people in India lost their lives in such as breaking traffic signal and horning in No
road mishaps improper road infrastructure failure to horning zone. It also has a major objective of
follow the speed limit, and increase in drinking and exercising road discipline such as speed control in
driving habits are among the major factors different areas and horn control in horn prohibited
contributing to deaths from road crashes, WHO said zones .This paper presents vehicle speed control in
in its report on ‘Decade of Action for Road Safety variable zone in this feature ;speed of the vehicle is
2011-2010’.Currently road safety systems are controlled in different areas such as
available in high end luxury cars such as Audi, flyovers ,bridges ,highways ,schools ,hospitals ,cities
Mercedes, Benz, etc., to name a few. Example: On and suburbs .Horn control of vehicle. Horn control of

CSEIT1951131 | Received : 01 Feb 2019 | Accepted : 16 Feb 2019 | January-February -2019 [ 5 (1) : 508-513 ]
508
Chandrashekhar Rathod et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 508-513

vehicles in no honking zone–controlling unwanted cushioned impact at a speed of more than 30 km/h,
disturbances in horn prohibited zone such as with lower speeds resulting in either death or serious
hospitals, public libraries ,courts, schools, etc., injury [1, 2]. The objective of the SS approach is that
the three model elements should be designed and
The Safe System (SS) approach to transport networks monitored to proactively prevent deadly speeds from
originated with the “Safe Road Transport System” happening and allow for a reduced emergency
model developed by the Swedish Transport Agency. response time in the event of an accident.
In its essence, the approach migrates from the view
that accidents are largely and automatically the Elements of the SS approach are as follows.
driver’s fault to a view that identifies and evaluates
the true causes for accidents. Through the (1) Safe Vehicle. Emphasis on vehicle safety is
categorization of safety into the safety of three verified through mandated regulatory testing and
elements (vehicle, road, and road user), SS minimizes rating, as well as technologies such as electronic
fatalities and injuries by controlling speeds and stability control. Beyond this, enforced checks (e.g.,
facilitating prompt emergency response. The model upon license renewals) combined with on the road
has been widely adopted since its introduction and is reporting work to review the status of vehicle safety.
currently motivated by the WHO as a basis for road
safety planning, policy-making, and enforcement. (2) Safe Road. The assessment of road (or road
network) safety is multifaceted. Road inspection
enables clear and direct observation of the state of the
road and assesses the need for repairs or
modifications. The structure of the road network is
amenable to safety assessment through partitioning
into what is called “Traffic Analysis Zones (TAZs)”
[4]. In addition, considerations for crash data and
other supporting data offer further insights into
general safety assessment.

In 2011, the European Road Assessment Programme


(EuroRAP) generated the European Road Safety Atlas
for EU countries [5]. The atlas indicated the safety
level of roads with a star rating based on specially
equipped vehicles for multimedia-based data
aggregation [6]. The EuroRAP efforts continue to
Figure 1 : The Safe-System-Based Safe Road
implement an SS approach across the EU, along with
Transport Systems
several other national programmes within the
International RAP, or iRAP, initiative [7].
An illustration of the model is provided in Figure 1.
A central emphasis is given to speed in the SS
(3) Safe Road User. There are several aspects to road
approach as it is the strongest and most fundamental
user safety, including measures for education and
variable in the outcome of fatality. The fragility of
awareness, travel distance, exposure, licensure,
the human body makes it unlikely to survive a
enforcement, and sober driving [3]. The need for

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


509
Chandrashekhar Rathod et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 508-513

such characterization rises substantially as the Anto bennet et al., [13] had proposed that a review
findings of crash report analysis in cities typically on an Embedded System in Passenger Car for Road
note a critical dependence on either driver behavior Safety using GPS and GSM.GPS Receiver gets the
or driver awareness [8]. A great need is further location information from satellite in the form of
established in these studies for innovative latitude and longitude. The system can be
mechanisms to instill safe driving at the licensing and interconnected with the alcohol detection and alert
post-licensing stages. the owner on his mobile phone.

II. LITERATURE REVIEW Anto bennet et al., [14] had proposed that a review
on Alcohol Detection and Accident Avoidance Using
Pramodshelke et al. [9] had proposed that a review on locking With Tracking. This system will detect
Arduino Based embedded System in Car for Road drunker driver by alcohol sensor through driver
Safety using RFID .In this work when the RFID tag is breathe fitted on steering in front of driver ,the
detected by the RFID reader the speed of the motor message is sent to the police through GSM system
reduces in specified areas and it avoids accidents. and provides GPS base system to track those cars.

Ch . Sonali Shankar et al., [10] had proposed that a III. RELATED WORK
review on Arduino based Embedded System in
Passenger Car for Road Safety .In this work the sound Many countries around the world ban the use of
of the horn stops and speed is also reduced when the hand-held devices while driving, including the
RFID tag is detected by the RFID reader. United States, Canada, Germany, Israel, Greece,
Norway, and more. While smartphones can be a
Deepa et al., [11] had proposed that a review on dangerous distraction, correct use of cell phone apps
Embedded System in Passenger Car for Road can reduce risk and promote safer driving behavior.
Safety.in this work the proposed features are Apps used with hands-free devices can monitor
automatic collision notification that gives notification distractions, protect young drivers, call for help when
to the victim’s relative, red light traffic control makes needed, and prevent speeding. Following are few Top
sure the vehicle doesn’t break signal, speed control Road Safety Apps.
alters speed in different zones, horn control prevents
honking in horn prohibited zone, alcohol detection Kruzr – Driving Assistant for Distracted Driving
detects drunk driving and vehicle security is used to Kruzr Road Safety App Kruzr is a free android-based
prevent theft Suhas Katkar et al., [12] had proposed app that detects when you’re driving and
that a review on An Embedded System In passenger automatically screens incoming calls and messages.
Car For Road Safety. This project is designed to Kruzr tells senders you’re on the road and will
inform about an accident or theft that has occurred to respond later. Don’t worry, you won’t miss critical
the vehicle, to the family members of the travelling calls. Kruzr automatic driving assistant allows urgent
persons and concerned authorities. Anto bennet et al., calls to proceed, while less important calls and
had proposed that a review on An Embedded System messages can be retrieved later.
in Passenger Car for Road Safety using GPS and GSM.
This work is designed to inform about an accident Safely Home – Detects Accidents
that is occurred to a vehicle to the family members of Safety Home Road Safety App Safely Home is a free
the travelling persons. Android app that automatically detects road accidents

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


510
Chandrashekhar Rathod et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 508-513

without any human intervention. If you’re in an such as “Driving right now, will get back to you
accident, the app notifies your emergency contacts. shortly”.
Safely Home includes a “Save a Life” feature to help
injured accident victims get help quickly. The app Fleetsafer
also provides traffic updates, and GPS-guided Fleetsafer is aimed at employers who want to reduce
Roadside Assistance for mechanical problems. the risk of employee liability for accidents on the job.
The app appears on company-issued tablets and
Drive Mode: Safe Driving App smartphones. It detects the driving state and
Drive Mode Road Safety app Drive Mode is an automatically puts the device into safe mode. Only
Android app that includes a voice-control option for emergency calls may be made.
streamlined interfacing with your favorite
applications, as well as simple swipe or tap MamaBear
functionality. With Drive Mode voice control, MamaBear Family Safety is a free app for Android
drivers can easily access navigation, music and and Apple devices. Recognized as the “Best Parenting
messaging apps. App” in 2017 by Mumi Family Awards, MamaBear
Drive Mode can be setup to launch when you start provides parental alerts for speeding, and provides
driving. The app will ignore calls and messages in GPS locations for young drivers. MamaBear also
“Do Not Disturb” mode, send text message auto allows parents to monitor their child’s text messages,
replies, or you can choose to reply to calls and social media, and arrival and departure times at
messages using seamless voice commands, and much locations away from home.
more.
Drive Safely Pro Road Safety App
AT&T DriveMode The DriveSafe.ly app is designed to prevent distracted
ATT Road Safety App driving accidents by providing hands-free access to
AT&T Drive Mode is a free app for Android and emails and text messages. Powered by a text-to-
Apple users and is available on all carriers. This app speech program, Drive Safe.ly Pro converts messages
silences incoming messages when the GPS detects in 27 languages. The app reads text messages and
you are driving 15mph or faster, and turns off when emails aloud in real time, allowing drivers to keep
you drop below 15 mph for two or three minutes. It their hands on the wheel and eyes on the road.
offers you a list of up to five contacts you can call
hands-free. Parents of young drivers can receive In this system, various sensors such as soil moisture,
alerts if the app is turned off. AT&T Drive Mode is DHT11 sensors are connected to the input pins of
also available in Spanish. arduino microcontroller. The sensed values from the
sensors are displayed in LCD. If the sensed value goes
iBolt Dock’n Drive beyond the threshold values set in the program, the
iBolt Road Safety App The iBOLT Dock’n Drive app pump will be automatically switched ON/OFF by the
requires the use of an iBolt docking station. This relay circuit and it is connected to the driver circuit
Android-based app provides automated answering of which helps to switch the voltage. The farmer will be
incoming calls. The “It Can Wait” feature places intimated about the current field condition through
incoming text messages on hold while the phone is GSM module and also updated in the web page. By
docked. Messages are delivered to the phone when using this system, the farmer can access the details
it’s not docked. You can set the auto reply message, about the condition of the field anywhere at any time.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


511
Chandrashekhar Rathod et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 508-513

IV. CONCLUSIONS [7] International Road Assessment Programme


(iRAP), “iRAP,” https://www.irap.org/.
This work illustrates the viability of an economic [8] H. M. Hassan and H. Al-Faleh, “Exploring the
road safety monitoring and assessment solution risk factors associated with the size and severity
through exploiting advances in the Internet of Things of roadway crashes in Riyadh,” Journal of
(IoT) within the context of smart cities. The Safety Research, vol. 47, pp. 67–74, 2013. View
introduced architecture facilitates robust and at Publisher·
dynamic road safety assessment that complements [9] Prof.Pramodshelke, Ram Anjankar,
the Safe System approach motivated by the World SakesheeWanare, Ganesh Surve“A Review on
Health Organization (WHO), which has been Accident Avoidance by Eye Blink Sensor and
increasingly adopted worldwide. An application of Alcohol Sensor Using GSM” E-ISSN:2321-9637,
the dynamic assessment framework for route 09th April 2017
planning is also demonstrated. Future work involves [10] Sonali Shankar Chalwad, Snehal Bhimrao
exploring further applications, especially in the Gaikwad, PrashantA.Chougule “Accident
context of raising driver awareness of the road safety Prevention Using Eye Blink Sensor”ISSN:2393-
conditions during their trips. 2835, Volume-4,Issue-1, January-2017
[11] SuhasKatkar,MaheshManikKumbhar, Priti
V. REFERENCES Navanath Kadam “Accident Prevention System
Using Eye Blink Sensor” Volume:03, Issue:05,
[1] F. Wegman, “The future of road safety: A May-2016
worldwide perspective,” IATSS Research, vol. [12] Pratik S. Danage, Ajay B.Dongare,
40, no. 2, pp. 66–71, 2017. View at Publisher · SangramsinhS.Dongare, Rahul B.Ghogare,
View at Google Scholar · Ganesh N.Khare “Automatic Breaking Sysetm
[2] World Health Organization, “Save LIVES - A Using Eye Blinking Sensor”Vol.2, Issue 2,
road safety technical package,” 2017. pp(166- 170),October2015- March-2016
[3] W. E. Marshall, “Understanding international [13] Deepa K. B, Chaitra M, Ankit Kumar
road safety disparities: Why is Australia so ShrmaSreedar V. S, Prashant Kumar H. R.”
much safer than the United States?” Accident Accident prevention by eye binking sensor and
Analysis & Prevention, vol. 111, pp. 251–265, alcohol detector” Volume-4 , Issue-7 , pp: 351-
2018. View at Publisher · 354, July-2015
[4] X. Wang, X. Wu, M. Abdel-Aty, and P. J. [14] Dr.AntoBennet, M , Sankaranarayanan S,
Tremont, “Investigation of road network Ashokram S ,Dinesh Kumar T R,“Testing of
features and safety performance,” Accident Error Containment Capability in can Network”,
Analysis & Prevention, vol. 56, pp. 22–31, 2013. International Journal of Applied Engineering
View at Publisher · View at Google Scholar · Research, Volume 9, Number 19 (2014) pp.
[5] European Road Assessment Program 6045-6054
(EuroRAP), “European Road Safety Atlas,” [15] Dr.AntoBennet, M, SankarBabu G, Natarajan S,
http://atlas.eurorap.org/. “Reverse Room Techniques for Irreversible
[6] European Road Assessment Programme Data Hiding”, Journal of Chemical and
(EuroRAP), “Star Ratings,” Pharmaceutical Sciences 08(03): 469-475,
http://www.eurorap.org/protocols/star-ratings/. September 2015.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


512
Chandrashekhar Rathod et al. Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 508-513

[16] Dr. AntoBennet, M ,Sankaranarayanan S,


SankarBabu G, “ Performance & Analysis of
Effective Iris Recognition System Using
Independent Component Analysis”, Journal of
Chemical and Pharmaceutical Sciences 08(03):
571-576, August 2015.r.

Cite this article as :

Chandrashekhar Rathod, Sneha Wakde, Roshan


Punwatkar, Rupesh Sharma, Rupal Patil, Prof. Rohan
Kokate, "A Review on Various Road Safety Systems",
International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 508-513, January-February 2019.
Journal URL : http://ijsrcseit.com/CSEIT1951131

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


513
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : 10.32628/CSEIT1951129

Review on Various Tools and Techniques for


Plagiarism Detection
Kedar Bhujade1, Tushar Bhandirge1, Rohan Khobragade1, Saurabh Shelare1, Jayprakash Shukla1,
Prof. R. S. Thakur2
BE, Department of Computer Science & Engineering, Dr. Babasaheb College of Engineering & Research,
1

Nagpur, Maharashtra, India


2Assistant Professor, Department of Computer Science & Engineering, Dr. Babasaheb College of Engineering &
Research, Nagpur, Maharashtra, India

ABSTRACT

Being a developing issue, plagiarism is commonly described as literature theft and academic dishonest nature in
the writing, and it must be avoided and adhere to the moral standards. Plagiarism occur in scholastics, paper
publication, music, work of art developing quickly, so the recognizing plagiarism is essential. While the most
recent couple of year’s plagiarism detection tools have been utilized predominantly in research conditions,
refined plagiarism programming and instruments are presently quickly rising. In this paper, we give an outline
of various plagiarism programming and apparatuses to take care of the plagiarism issue. We propose an element
classification conspire that can be utilized to examine plagiarism discovery programming and plagiarism
recognition instruments. This plan depends on the product's general qualities, devices qualities, and apparatuses
property.
Keywords : Plagiarism Detection, Plagiarism Types, Plagiarism Techniques, Plagiarism Algorithms

I. INTRODUCTION content, and reuse thoughts with/without referring


to them and others. Recognition of plagiarized record
The term “plagiarization” is characterized by the fact takes on vital jobs in many applications, for example,
that one takes thoughts, archives, code, etc. from document the executives, copyright assurance, and
some another and passes them without specific plagiarism aversion. Existing conventions expect that
reference as one`s own. Plagiarism is therefore a the substance of records put away on a server is
global issue in many aspects of our lives. There are a straightforwardly open. This presumption constrains
wide variety of types of plagiarism, and plagiarism in progressively down to earth applications, e.g.,
academies can be a deeply deterrent to educators and identifying copied reports between two meetings,
undergraduate study. On the off chance that where entries are confidential [2].
plagiarism isn't tended to adequately, literary thieves
could increase undeserved preferred standpoint, for Plagiarism can be one of the best known types, for
example, more checks for their tasks with less efforts. example, to replicate the whole or some portion of
the record, to rephrase the same substance in
Different types of plagiarism [1] are included: Use different words, to use the thoughts and ideas of
source without valid reference to them, summarize others to refer to incorrect or non-existent source [3].

CSEIT1951129 | Received : 15 Feb 2019 | Accepted : 26 Feb 2019 | January-February -2019 [ 5 (1) : 514-518 ]
514
Kedar Bhujade et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 514-518

Various plagiarism methods include deciphered 1. Web enabled systems


plagiarism, in which substance is interpreted and Web-enabled system are more widely used because
used without specific reference to the first work, they easily and reliably extended their search for
masterful plagiarism, in which distinguishable media, plagiarized resources to the global web. The
including image and recordings, show other people`s following are the two web-enabled screening systems,
work without legitimate reference [3].
✓ EduBirdie
A plagiarized code (also referred to as clone code)
It provides us with a report showing the percentage
which can be described as a reuse of the source code
of content that is unique. One can check any form of
without explicit consent or reference. So a plagiarized
text using this tool, whether it is an essay, academic
program can be described as a program developed
paper, Technical descriptions, case studies, product
with a few routines from some other program
details or white paper.
changes, routine changes, regularly message
substitutions, don't require a definite comprehension
✓ Safe Assign
of the program. Tragically, significant class sizes have
made the plagiarism of programming tasks less
It compares uploaded text document with a set of
demanding. research papers to identify areas of overlapping
between the uploaded document and available work.
In extensive university courses, the plagiarism of PC It is based on a distinctive text matching algorithm
projects can be very fairly normal. A plagiarized that detects the exact and inaccurate matching of a
program with an alternative visual appearance can be paper with the uploaded document. It compares
delivered with a bunch of fundamental manager submissions to several databases such as Inform
activities. This makes it very difficult to manually journal Database.
recognize plagiarized programs in significant classes.
Every one of this plagiarism procedures has a 2. Stand-alone systems
significant impact on the process of education. This software can be installed in the computer system.
Therefore, how can we ensure the management of Some of stand-alone system are, Their paper showed
how the instrument and also the experience of
plagiarism frameworks and the recognition and
using it in four different programming task in the
management of plagiarism? A basic issue requires
last 12 years.
Computer researchers to arrange.
Hermann et al. [7] state that plagiarization is
II. PLAGIARISM IN DOCUMENTS robbing someone else's work of credit. As per the
creators, content plagiarism implies that a creator
Documentary plagiarism is more relevant to the simply duplicates it with giving it the real credit.
academic purpose of the student community, They represent the main attempts to acknowledge
especially the postgraduate who modifies the plagiarized portions of a content using measurable
available documents and presents it as his own. This models of dialect and perplexity. The
should be prevented as it affects the quality of the investigations were carried out on two specific and
ability of the students themselves. It is therefore academic corporation. The first record and
necessary to detect plagiarism in documents first and linguistic form and stemmed adaptation were
for this purpose the following systems can be used, contained in the two specific works. The

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


515
Kedar Bhujade et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 514-518

plagiarism on these reports was distinguished and institutions such as Ohio University, Umass Boston
the results were checked. and Trinity College Dublin.

Jinan et al. [8] concentrated on the instructive setting and ✓ WCopyFind


confronted comparable difficulties. They show the most
competent method for checking cases of plagiarism.
What's more, they intended to fabricate learning networks
This system is plagiaristic between two or more tasks.
of understudies, educators, organization, and personnel
and staff all teaming up and developing solid connections
Many commercial tools for detecting plagiarism are
that give the establishment to understudies to accomplish available. Table 1 presents the tools` comparative
their objectives with more noteworthy achievement. They analysis.
also advanced the sharing of data. In a straightforward,
customizable and reusable way, they gave consistent
Table 1. Comparison software based on their
combination heritage and various different applications.
different features
Learning gateway may give a help device to this
learning framework. In any case, fabricating and
adjusting learning gateway is certifiably not a simple
errand. This paper recognizes the plagiarism of java
understudy assignments in the product.

Plagiarism can be differentiated by an understanding


of sentences and paragraphs from paper, which can
also be found with the help of web indexes. They
pointed this out to create a free software that can be
used to identify plagiarism in their classed by any
teacher or encouraging partner. Nathaniel et al. [9]
characterize plagiarism as a major problem affecting
III. LITERATURE REVIEW
copyrighted records/materials. They state that
plagiarism is expanded nowadays because of the
Allan et al. [5] exhibited a system for identification of
productions in on the web. They suggested a new
plagiarism. The development of the web, with
discovery technique for plagiarism called SimPaD.
bottomless data online, exacerbates the issue even.
The reason for this strategy is to contrast sentence by
The creators have discovered four diverse approaches
sentence to create similarities between two archives.
to approach plagiarism discovery. They continued to
Examinations show that SimPaD increasingly
pursue comprehensive looking and took the center
accurately identifies plagiarized reports and outstrips
ground technique instead of thoroughly or
existing approaches to plagiarism recognition.
accidentally searching for sentences on the web in a
study paper. They found the manifestation of thought
✓ Plagiarisma Checker X
they had acquired.
It is a simple system for undergrads, educators,
content producers, SEO experts and site owners to Francisco et al. [6] state that research facility work
verify that others have copied their work. According assignments are essential for software engineering
to the programmer, its clients include educational learning. The study showed that 400 understudies
duplicate a similar research in illustrating their

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


516
Kedar Bhujade et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 514-518

assignment at the same time during the last 12 years. document is mark as plagiarism document and
This has made the instructors to give careful similarity percentage is calculated.
consideration on finding the plagiarism. In this way,
they constructed a discovery device for plagiarism.
This device had the full range of tools to help
administrator to manage the work of the research
facility. To quantify the similarities between two
assignments, they used four comparability criteria.

IV. PROPOSED SYSTEM

Fig. 1 shows the proposed system architecture. In our


proposed approach user inputs a single document for
plagiarism checking. Initially pre-processing is
performed on document in which unnecessary space
within document, special characters, etc. are
removed and then stopword removal process is
performed in which the keywords such as a, an, the,
numbers in documents & other stopword are
removed. Then stemming processed is performed in
which ing, ed, etc. of each keyword is removed. At
the end only dictionary keywords are remain in input
document.

After getting dictionary keyword from document,


important keywords separated out (keywords having
count greater than threshold k). These top k keyword
set is passed to neural network classifier which
performs classification on previously stored
documents in database in two classes such as
documents containing top k keywords (say class 1)
and documents which don’t contain top k keywords
(say class 0). Then we use documents containing top
k keywords (class 1) for further processing.

Figure 1. System Architecture


After this, the document vector of Input document
and class 1 document is generated. Then TF-IDF of
V. CONCLUSION
all document is generated and finally cosine
similarity is calculated between input document and
In this paper we study plagiarism detection is very
class 1 documents. If similarity is found between
important not only in academics, but also in industry,
input document and any other document then input
music, artworks, etc. In this study, in particular, it

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


517
Kedar Bhujade et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 514-518

was shown how the problem of plagiarism can be Plagiarism in Programming Assignments,” IEEE
addressed using different techniques and tools. In this Transactions on Education, vol. 51, no. 2, pp.
paper, we have seen that different software and tools 174-183, 2008.
are available to check plagiarism. Comparison of [7]. Hermann M., Frank K., and Bilal Z.,
software and tools has shown that they still have no “Plagiarism -A Survey,” Universal Computer
software and tool that can detect that the document Science, vol. 12, no. 8, pp. 1050-1084, 2006.
has been plagiarized at 100 percent, since each [8]. Jinan F., Alkhanjari Z., Mohammed S., and
software and tool has advantages and limitations Alhinai R., “Designing a Portlet for Plagiarism
according to the features and performance described Detections within a Campus Portal,” Journal of
in the table. However, this software has limitations, Science, vol. 1, no. 1, pp. 83-88, 2005.
tools that have a significant impact on the success of [9]. Nathaniel G., Maria P., and Yiu N., “Nowhere
plagiarism detection. We also presented our proposed to Hide: Finding Plagiarized Documents Based
to the detection of plagiarism. on Sentence Similarity,” in Proceedings of
IEEE/WIC/ACM
[10]. International Conference on Web Intelligence
VI. REFERENCES and Intelligent Agent Technology, NSW, pp.
690-696, 2008.
[1]. P. OGR, “What is Plagiarism?”, [On Line]
http://www.plagiarism.org/,Retrieved Nov. 15, Cite this article as :
2010
[2]. C. Lyon, R. Barrett, and J. Malcolm, “Plagiarism Kedar Bhujade, Tushar Bhandirge, Rohan
is Easy, but also easy to detect.” Cross- Khobragade, Saurabh Shelare, Jayprakash Shukla,
Disciplinary Studies in Plagiarism, Fabrication, Prof. R. S. Thakur , "Review on Various Tools and
and Falsification, 2006. Techniques for Plagiarism Detection", International
[3]. L. Romans, G. Vita, and G. Janis, “Computer- Journal of Scientific Research in Computer Science,
based plagiarism detection methods and tools: Engineering and Information Technology
an overview,” the 2007 international (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
conference on Computer systems and 514-518, January-February 2019.
technologies. 2007, ACM: Bulgaria. Journal URL : http://ijsrcseit.com/CSEIT1951129
[4]. S. Mann and Z. Frew, “Similarity and
originality in code: plagiarism and normal
variation in student assignments,” the 8th
Australian conference on computing education,
2006.
[5]. Allan K., Kevin A., and Bruce B., “An
Automated System for Plagiarism Detection
Using the Internet,” in Proceedings of World
Conference on Educational Multimedia,
Hypermedia and Telecommunications,
Chesapeake, pp. 3619-3625, 2004.
[6]. Francisco R., Antonio G., Santiago R., Jose L.,
Pedraza M., and Manuel N., “Detection of

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


518
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951116
An Elementary Study on Various Techniques Involved in Face Recognition
Systems : A Review
Payal Maken
School of Computer Science & Engineering, Shri Mata Vaishno Devi University, Kakryal, katra, Jammu &
Kashmir, India
ABSTRACT

Face recognition has now become one of the interesting fields of research and has received a substantial
attention of researchers from all over the world. Face recognition techniques has been mostly used in the
discipline of image analysis, image processing, etc. One of the face recognition techniques is used to develop a
face recognition system to detect a human face in an image. In face recognition system a digital image with a
human face is given as an input which extracts the significant features of face such as (eyes, nose, chin, cheeks,
etc) to recognize a face in a digital image which is an exhausting task. Security of information is very salient
feature and is difficult to achieve. Security cameras are present in offices, universities, banks, ATMs, etc. All
these security cameras are embedded with face recognition systems. There are various algorithms which are
used to solve this problem. This paper provides an overview of various techniques which are often used for this
face recognition in a face recognition system. This paper is divided into five parts, first section concludes
various face detection techniques, second section describes about image processing ,third section have details
about face recognition techniques, fourth section describes various classification methods and last section
concludes all of these sections.
Keywords : Face Recognition, Face Detection, Features Extraction, Pre-Processing

I. INTRODUCTION passwords; person identity numbers (PINs), etc. But


identity cards can be easily stolen or misplaced;
Face recognition techniques is used in wide variety of passwords and PINs can be forgotten. However
face recognition systems. A face recognition person’s biological features can never change,
technique involves identification of faces and then misplaced or stolen. So, this technology can give
comparing it with the images in the database. Human solution to this problem since a face of a person is
beings can easily differentiate the human face on the undeniably connected to a person only except in the
basis of various factors. But it is complicated to case of twins. For the security purpose biometric
develop such a computer system which can imitate techniques gives better results than the identity cards,
human capabilities of recognition and detection of a passwords, etc. Various biometric techniques
face in a digital image. In recent years we can see that comprises of fingerprint impression, iris detection,
researches related to face recognition have been DNA detection and face recognition. Among all Face
escalating, and this is so because of the need of recognition is most challenging and strenuous area in
security in various areas. Person’s identification is this area.
most important for the purpose of security. We can
identify a person through its identity cards,

CSEIT1951116 | Received : 15 Feb 2019 | Accepted : 27 Feb 2019 | January-February -2019 [ 5 (1) : 519-525 ]
519
Payal Maken Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 519-525

Face recognition system basically consists of Each year new techniques are developing for this
following blocks: purpose.

II. RELATED WORK


1. Input image
2. Facial detection A. Face Detection
3. Image processing Face detection is the first step in any face analysis
4. Facial recognition system. Face detection involves finding a face in any
a. Features extraction given input image. So, basically it means to find out
5. Verification/Validation where is the face irrespective of the background. For
example in a figure 2, there is a photo frame over
A. Input Image
there but what it tries to do is extract from an image
A physical or behavioral sample of image is captured just a segment which represents a face.
from any physical image capturing device, which
serves as an input the system.

B. Facial detection
It basically means locating a face in a given image.

C. Image processing
Images are cropped to avoid facial image remains and
Figure 2. Detecting a face in an image
images are converted into black and white and then
There are number of techniques for detecting a face
to gray images.
in an image.
D. Facial recognition
1. Voila Jones face detection
It means to extract the features of a face (eyes, nose,
This technique was given by Paul Viola and Michael
etc) and make a template of it.
Jones. It is also known as haar cascade algorithm, this
E. Verification algorithm detect a face in an image on the basis of
System recognizes a given individual and gives a various features like eyes, nose, upper body, lower
decision. body. A system is designed by giving input some faces
and some non-faces and training a classifier. This
F. Validation
classifier is known as haar cascade classifier. Now a
In this system compares a given individual face with
cascade classifier is something that tells what we have
faces in the database to identify a given face.
to look in an image for detecting a face. For example,
if we are distinguishing whether a person is
overweight or not, then weight is the feature that is
to be considered not their age or skin color [1]. So we
basically train a classifier to detect a face. The data is
Figure1. Block diagram of face recognition system then stored in a file in the system. By referring to
Figure 1 shows the block diagram of face recognition that trained data one can detect a face.
system. Nevertheless ample work has been done on This algorithm includes following:
face recognition problem, but still not any work is up • Haar features
to the mark for the implementation point of view. • Integral image

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


520
Payal Maken Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 519-525

• Adaboost this we understand that this pattern is absorbed in


• Cascading this picture. Viola Jones uses 24*24 window as a base
window size to start evaluating these features in any
a. Haar features
given image. Considering all possible parameters of
A haar cascade is a series of “haar-like-features” that the haar like features we get about 16000+ features in
are combined together to shape a classifier. this window.

b. Integral image
In an integral image the value at pixel (x, y) is the
sum of pixels above and to the left of (x, y).

Figure3. Haar-like features

Figure5. Input image to Integral image


So haar-like-features are on the whole are
rectangular pattern data. These haar-like features c. Adaboost
have some sort of resemblance to facial features or Adaboost eliminates the redundant or irrelevant
characteristics of face (eyes, nose, cheeks, etc). These features and narrow it down to some 1000 features.
features are applied one by one to whole image and So that we have to apply only useful features to an
where the face features matches with it, it gives high image instead of all the features to reduce the
intensity value. The sum of pixels under white complexity and to make it time efficient. Adaboost
rectangular is subtracted from the sum of pixels find the best features among all the 16000+ features.
under black rectangular gives a single value. Figure 3
shows the Haar- like- features for face detection. d. Cascading
In every 24*24 we need to evaluate 2500 features that
we obtained after performing adaboost. But, instead
of calculating 2500 features in every 24* 24 window,
we use cascades. That is out of 2500 features, we
make a set of 10 features in one classifier, then the
next 20 or 30 features in next classifier and then 200
or 300 in next classifier and so on. Thus cascading
results in a strong classifier as a linear combination of
weak classifiers.
Figure4. Usage of haar-like-features

In figure 4 two haar features are applied to an image


and these features are able to extract the nose and
eyes from the image and this is done by applying the
feature all over the face and get the high value only
at the pixels where this pattern matches exactly, from
Figure6. Cascaded classifiers

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


521
Payal Maken Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 519-525

2. Face detection using skin color 1. Background removal/subtraction or cropping


Human skin color is proven to be most efficient 2. Converting RGB into gray image
method of detecting a human face than any other 3. Converting grayscale into binary image
algorithms that uses facial features[2]. Using skin 4. Filtering or smoothing
color for detection has several advantages, that the
1. Background subtraction or cropping
processing is much faster. The image for which face
In background subtraction, the face is detected from
has to be detected must be in color picture not in
the whole image, and this is done by face detection
grayscale. Although, different people have different
method with the help of features extraction (eyes,
skin colors and the difference lies between their
nose, lips, etc).
intensities rather than chrominance [3]. Since the
input color image is in RGB format, so this technique
usually uses the color components in other color
spaces such as YCbCr, because of the RGB
components which may fail to detect the face.
In YCbCr format, luminance information is
contained in Y component and the chrominance
information is in Cr and Cb components. Therefore Figure 7. Background removal or subtraction
luminance information of the skin can be easily de-
2. Converting RGB into gray image
embedded and this will indicates the skin region.
Now the face detection in skin model involves RGB (red, green and blue) image is known as true
detecting the skin areas. Later the skin region is color image, is stored as m*n*3 data array having red,
localized using: green and blue color components for each pixel. The
color of each pixel is decided by the combination of
• Template matching
red, green and blue intensities. Now, Gray images are
• Neural networks
those in which the only colors are the shades of gray.
B. Image preprocessing In ‘gray’ color the red, green and blue components
have equal intensity. The reason for converting RGB
Normally the images that are obtained from image
into grayscale is that less information needs to be
acquisition are not suitable for face recognition.
provided for each pixel. RGB image can be converted
Quality of image plays a crucial role in face
into Grayscale by applying weighted method or
recognition rate. Quality of image is compromised
luminous method:
due to the certain factors such as noisy images, poor
a. Get the RGB value of pixel.
resolution, etc. So to overcome these problems image
b. Use the formula: Grayscale = 0.3*R + 0.59*G +
preprocessing is done before feature extraction (face
0.11*B
recognition) which will increase the face recognition
c. Replace the R, G and B value of the pixel
rate.
with the Grayscale value calculated at
There are various image preprocessing techniques the
previous step.
can be used to improve the quality of input images
such as image normalization, de-noising, image
filtering, background subtraction etc.
Various steps involved in image pre-processing are
[4]:

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


522
Payal Maken Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 519-525

C. Face recognition
Face recognition primarily means extracting the
features from face i.e. eyes, nose, chin, etc. Various
techniques which are used for this purpose are
following:

1. Principal component analysis (PCA) or Eigen faces


Figure 8. RGB to Grayscale image It is mostly used tool for data analysis making a
3. Converting grayscale into binary image predictive model (face recognition system). Face
recognition technique based on PCA generally make
Binary image is a digital image that has only two
use of an Eigenfaces. In face recognition system,
possible values for each pixel (0 for black and 1 for
training set consists of a set of M images, and each of
white). It uses only two colors black and white.
these images is of m*m dimensions, so there will be
Grayscale image can be converted into binary image
m2 pixels. The PCA transforms the set of images into
by calculating its pixel value, if the intensity value/
low dimensional pictures known as Eigen faces. So
pixel value is greater than 127 then change it to 255
basically, PCA is a mathematical method which use
or 1 and if the intensity value is less than 127 change
an orthogonal transformation to convert a set of
it to 0, then the image is converted into binary image.
values of possibly correlated M variables(face images
in the training dataset) into a set of values of K
uncorrelated variables called principal
components(Eigen faces). The numbers of Eigen faces
is always less than or equal to the number of test
images in the training set i.e. K< =M.

Figure 9. Grayscale to binary image

4. Filtering or smoothing
Filtering or smoothing is done to enhance the quality
of the image by the removal of the noise (blurring,
poor resolution, etc) from the image.
Median filter: Filtering set the value of Figure 10. Face images to Eigen faces
the output pixel as the median of all the In Eigenfaces the first principal component shows the
neighborhood pixels that is taken as input. Other most dominant feature of the dataset and each
filters replace the pixels values by mean and succeeding component in turn shows the next
average value: dominant component.
a. Take the median value of pixels in a window in
the neighborhood.
Median= sort all the pixels in an increasing
order, take the middle one.
b. Replace the pixel value with the median value.

Figure 11. Eigen faces

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


523
Payal Maken Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 519-525

To reduce the calculations that are needed for finding separable and produces the good results even under
the principal components, the dimensionality of the varying illumination conditions and images with
original dataset is reduced before they are calculated. different poses and expressions.
Since each preceding components how less features
and more noise, only few first principal components
are selected and rest of the last are discarded.

Figure14. Fisher Faces

Figure12. Selected k useful Eigenfaces 3. Independent Component Analysis


Each image in the original dataset can be represented ICA is an abstract principle of PCA. PCA uses second
as a linear combination (weighted sum) of Eigenfaces. order statistics while ICA uses high order statistics to
form a better set of basis vector of face images. It
represents faces as a linear combination of
independent sources. It treats the images as random
variables and pixels as outcomes. So, it constructs the
face without face class information.

D. Verification/Validation
Verification and validation is the process of
Figure13. Weighted sum of Eigen faces recognizing an individual.
So, PCA reduces the number of values (from M to K) 1. Distance-wise classification
needed to recognize a face.
Once the features are extracted from face recognition
2. Linear Discriminant Analysis or Fisher Faces techniques, it is then to classify whether the test face
LDA is also known as Fisher faces. LDA is derived belongs to the training set or not. It is done by
from Fisher’s Linear Discriminant (FLD) technique. It measuring the distance between the training dataset
is considered as a classical technique in pattern and the image. If the distance is minimum, the test
recognition. It is used to find a linear combination of image is verified and the person is recognized.
features which separate two or more classes into 2. K-Nearest neighbor
objects [5]. Data is assumed to be uniformly
In this method, nearest neighbor is searched, and if
distributed in each class. It is used to convert the
maximum number of neighbors belongs to one class,
training samples into bases vectors. Linear
we can classify test image to that class. With
combination that is obtained using FLD is known as
increases k value, the classification rate decreases.
Fisher faces while in PCA it is known as Eigen faces.
LDA is preferred over PCA because it groups the 3. Neural Networks
classes into projected faces that can be linearly

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


524
Payal Maken Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 519-525

It is mostly used because of its similarities with the [4]. A. Sengupta and P. Mondal, “Estimating The
human brain. Neural networks make the mapping of Age Of Human Face In Image Processing Using
the images with the person’s expressions. Matlab,” vol. 4, no. 3, pp. 10899–10902, 2015.
[5]. M. N. S. Zainudin et al., “Face Recognition
III. CONCLUSION using Principle Component Analysis ( PCA )
and Linear Discriminant Analysis ( LDA ),” no.
The research in face recognition has been an exciting 05, 2012.
area and it will keep attracting many engineers and
scientists for many years in this field. In this paper Cite this article as :
different techniques that can be used in face
recognition are discussed. It also gives the whole Payal Maken, "An Elementary Study on Various
concept of face recognition system. According to this Techniques Involved in Face Recognition Systems : A
literature survey, Viola Jones gives real time Review", International Journal of Scientific Research
performance for face detection. Skin detection in Computer Science, Engineering and Information
technique also works well. Image preprocessing is Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
necessary to reduce the noise from the images and to 5 Issue 1, pp. 519-525, January-February 2019.
give better recognition rate. PCA and LDA works Available at doi :
better for features extraction. Among PCA and ICA, https://doi.org/10.32628/CSEIT1951116
ICA uses high order statistics. Neural networks, k- Journal URL : http://ijsrcseit.com/CSEIT1951116
nearest neighbor techniques and Distance-wise
classification methods are used for further
recognition and classification. Combining all these
techniques, a face recognition application or system
can be developed.

IV. REFERENCES

[1]. D. Saini and R. Kaur, “Assembling of Human


Beings in an Image to Detect a Group on the
basis of Distance using Digital Image
Processing,” International Journal of Current
Engineering and Technology, vol. 6, no. 2, pp.
462–466, 2016.
[2]. H. M. Z. Hewa Majeed Zangana, “A New
Algorithm for Human Face Detection Using
Skin Color Tone,” IOSR Journal of Computer
Engineering, vol. 11, no. 6, pp. 31–38, 2013.
[3]. A. Amine, S. Ghouzali, and M. Rziza, “Face
Detection in Still Color Images Using Skin
Color Information,” Network, vol. 24, no. July
2014, p. 4, 2002.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


525
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951132

E Return Tool
Pooja Aher*, Manish Kirnapure, Shubhangi Nemade, Tejaswita Shrivastava, Lulkaran Chohale
Department of computer science and Engineering, Priyadarshini J.L. College of Engineering, Nagpur,
Maharashtra, India
ABSTRACT

Every organization has need to pay salary to their employees, this work is done every month. Each and every
organization has to maintain the record monthly for future use. Now a day as we know the countries goes
become digit so this work is also performed digitally. This project E RETURN TOOL involves everything
regarding to employee payment and related to salary tax. This also includes delivering of taxes to the
government and also keep track of employee payments. This tool will facilitate to the employers to pre-validate
the ECR and also carry out amendments in the ECR file before uploading on Unified Portal. The return tool has
member wise details of wages and the contributions including basic details for the new and existing members’
return Tools also maintain the record of EPF member. Employer can check monthly or yearly details of every
EPF member. When a company hires its first employee it started its e return tool. In the United States, every
new employee must be reported to the state along with a completed W-4 tax form. The W-4 determines how
many allowances the employee qualifies for when calculating the federal income tax that should be withheld
from each check. Generally, the employee has to pay income tax depending upon their dependents.
Keywords : E Return Tool, ECR, EPF, EPFO, SME

I. INTRODUCTION existing members. E return Tools also maintain the


record of EPF member. Employer can check monthly
E Return Tool an Offline Tool for employers for and yearly details of every EPF member.
validating the ECR (electronic challan cum return)
before uploading it at the unified portal. Now a day, The E RETURN TOOL process typically includes
EPFO has not produced any software for validating calculating employee pay, recording payroll
the ECR or EPF return. Many employers are facing transactions and determining and paying taxes. A
this problem how to validating ECR for uploading at company must have in place of a timekeeping system
the unified portal because unified portal has only that accurately reflects the hours put in by
accept validating ECR "In order to assist the employer nonexcempted employees as well as the regular salary
to create the correct ECR file every month, E Return payments for exempt workers. Typically, employee
Tool provided an offline tool for employers for has to pay income from their earning; at the end of
validating ECR before uploading at Unified Portal," the year, they must report all wages, tips and
this tool will facilitate the employers to pre-validate othercompensation paid. Companies also must
the ECR and also carry out amendments in the ECR withhold the Social Security and a Medicare
file before uploading on Unified Portal. The return contribution from employees' wages and have to pay
has member wise details of the wages and a matching amount.
contributions including basic details for the new and

CSEIT1951132 | Received : 15 Feb 2019 | Accepted : 28 Feb 2019 | January-February -2019 [ 5 (1) : 526-530 ]
526
Pooja Aher et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 526-530

II. METHODS AND MATERIAL Form 5 are mandatory. In case a member has exited,
you can enter the date of exit and in such case the
E Return Tool software does a lot of the heavy lifting reason for exit will become mandatory. There are
of salary management for you. Like outsourced some additional details which also can be filled up.
payroll services, this software requires very little These details may become helpful later when
input from the employer, just employee wage member applies for claim. However, they are not
information and days. All of the tricky calculations mandatory.
and withholdings are calculated by the software,
which receives automatic updates every time tax laws D. Salary Information
are changed. The software can also fill out
This module taking the input from employer like
employment tax forms for you and remind you when
number of days and rate. Basis of that input it
to file them.
calculate the employee deduction and stored their
record in database
Module
E. Summery:
A. Registration:
This screen after the entry of the Return Month takes
Registration is first module that is authentication to
you to the actual data entry/upload screen for the
user means in Registration module software create
monthly return. This screen also displays the
the username and password
summary, account wise and month wise of the
already entered/uploaded wage details. In this
B. Log-in:
complete summery will be recorded in the form of
This is the screen where you have to enter your user A/c 1, A/c 2, A/c 10, A/c 21, A/c 22, Month, Year and
id and password for login. You have come here after total. The screen will ask you to enter in the Return
selecting an added establishment from the earlier Month and Return Year fields.
screen. The screen at top shows the Establishment ID
you have selected. In case you have selected a wrong F. Monthly Return:
establishment from the list, if multiple establishments
In this module EPF details is recorded and the based
have been added, please click BACK instead of login.
on this details EPF return fill in Unified Portal. This
It will take you back to the first screen for selecting
screen is to be used for entry of wages for each
the correct establishment.
member for a month.

The system allows the admin user to create/delete


1. Data can be entered for each member one by one.
new/existing users and to change password. However,
2. Data can be imported through csv file
the admin user cannot be deleted and only his
password can be changed. G. ECR and Report:
a. ECR:
C. Employee Registration
In this screen, thevalidated text file will be generated
This screen is used for upload/entry of the details of on the basis of the wages and other information
the members as required for filing the Return in entered by the user through the various functions of
Form 5, Form 10, for International Workers and for the software andbefore uploading the file and
Disabled Workers. The data that are required for the remitting the amounts in the bank against the

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


527
Pooja Aher et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 526-530

uploaded return, the employers should satisfy Flowchart :


themselves with the correctness of the data in the
text file.

b. Report:
i. Form 5:
In this module Form 5 is generated between pf range
or between Joining date. Form 5 is nothing but the
complete basic information of employee.

ii. Form 3A:


Form 3A is monthly details of individual employee
break in service.
iii. Form 6A:
Form 6A is yearly details of individual employees
including break in service.

III. PROBLEM FORMULATION


H. Parameter:
The use of traditional spreadsheet applications such
In this module user can easily change the parameter.
as Excel in the past few years has proved to be an
This is most important for user because some time
inefficient way of keeping faculties’ records,
employer has paid previous challan like Arrears.
calculating faculties’ enumerations and deducting
obligatory contributions from salaries. No option to
I. Back up and Reset:
restore or get the backup of all databases of faculties
This screen is to be used by admin only to reset data by just clicking a button Small and mid-sized
for the particular month. User can easily take backup enterprises (SMEs) are most affected as they
of complete file and store them to another location sometimes tend to use low end applications even as
for future used. they expand in business activities. This manual
method of calculating faculties’ salaries and carrying
out deductions impossible task due to its incapacity to
handle large amount of data without error. No bar
graphs to show the financial profit or loss of colleges.

IV. PROPOSED SYSTEM

This project is an online tool which is used to create


the validating ECR to upload in Unified Portal for
creating the EPF challan, this project is used for
involves everything that has to do with the payment
of employees and the filing of employment taxes.
calculating wages, taxes and other deductions,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


528
Pooja Aher et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 526-530

printing and delivering checks and paying statutory


employment taxes to the government. compliances
This tool will facilitate the employers to pre-validate like PT,
PF,ESIC,
the ECR and also carry out the ECR file before
TDS & IT
uploading on Unified Portal. The return has member
etc.
wise details of the wages and contributions including
basic details for the new and existing members. 3 ECR In this tool In this There is [4
return Tools also maintain the record of EPF TEMP used to tool no record ]
member. Employer can check monthly or yearly LATE create an used to will store
details of every EPF member. 2.0 ECR file in create
text format an If the
for filling Arrears employee
V. LITERATURE SURVEY
the EPF file in is cross 58
return at text year then
S Appli Features Proble Algorith R
unified format it’s not
r cation m m ef portal for created
Stateme filling accurate
N nt
the EPF text file
o
return
1 Mont In this 3A not bubble [1 at
hly E system generat sort and ] unified
return employer e binary portal
Softw can store the search
are EPF details KYC algorithm
of employee file not is used in
and generate create existing
ECR in text system VI. CONCLUSION
format to fill
EPF return Although a conclusion may review the main points of
on EPFO the paper, do not replicate the abstract as the
website conclusion. A conclusion might elaborate on the
importance of the work or suggest applications and
2 Comp A- 1) No Quick [2
extensions. Authors are strongly encouraged not to
uter PAYshould option sort and ]
Based take care of to binary call out multiple figures or tables in the conclusion
Payro Personnel restore search these should be referenced in the body of the paper.
ll Management or get used in
Syste , Payroll, the existing VII. REFERENCES
m Attendance backup system
Management of all [1]. "Automated Payroll System (A-PAY)", by
using databas
Arjun V. Singh , Siddesh V. Chaphekar ,Yogesh
biometric es of
S. Sawant, Volume 03, Issue 02, [February –
machine, faculties
Leave by just 2016] , ISSN (Online):2349–9745; ISSN
Management clicking (Print):2393-8161.
, SMS a button [2]. "Computer Based Payroll System
intimation, Implementation For E-Governance at Punjab

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


529
Pooja Aher et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 526-530

Agricultural University", by Poonamdeep Kaur,


Dr. Dinesh Grover, e-ISSN: 2278-067X, p-ISSN
: 2278-800X, Volume 5, Issue 3 (December
2012), PP. 55-60.
[3]. Taking EPF Parameter from
https://www.epfindia.gov.in/site_docs/PDFs/Mi
scPDFs/ContributionRate.pdf
[4]. ECR TEMPLATE 2.0 and ARREARS
TEMPLATE 2.0 available on
https://drive.google.com/file/d/0BwN9ZAwr4e
nzVVNUbU43S1ljcVk/view

Cite this article as :

Pooja Aher, Manish Kirnapure, Shubhangi Nemade,


Tejaswita Shrivastava, Lulkaran Chohale, "E Return
Tool", International Journal of Scientific Research in
Computer Science, Engineering and Information
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
5 Issue 1, pp. 526-530, January-February 2019.
Available at doi :
https://doi.org/10.32628/CSEIT1951132
Journal URL : http://ijsrcseit.com/CSEIT1951132

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


530
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951128

IoT Based LPG Gas Leakage Detector


Prof. Pranay Meshram*1, Nancy Shukla2, Stuti Mendhekar3, Renuka Gadge4 , Shivani Kanaskar5
*12345Computer Science &Engineering, Priyadarshini J. L. College of Engineering, Nagpur, Maharashtra, India

ABSTRACT

Gas leakage is a major problem in the industrial sector, in residential locations etc. One of the preventive
methods to stop the incident related to the gas leak is to install a gas leak detection kit at vulnerable locations.
The goal of this paper is to propose a system that can detect, alert and automatically control gas leaks. In
particular, a gas sensor has been used which has a high sensitivity to gases such as propane and butane together
with LPG. There is an alarm that is triggered once the LPG has been detected. The gas leakage system consists
of a Wi-Fi module that alerts the user by sending an SMS message.
Keywords : LPG gas, IoT (Internet of Things)

I. INTRODUCTION introduced so that it can also be used with alternating


current. To support this last case, it has a bridge
In homes and in various commercial areas, LPG is rectifier with a capacitor filter. This is followed by a
used for various purposes, such as cooking, Heating, controller that provides regulated power of + 5V.
Lighting, Cooling, etc. This energy source consists MQ-5 gas is used to detect LPG. This sensor can be
mainly of propane and butane, highly flammable used at + 5V. the sensitivity of this sensor is very high
chemical compounds. LPG loss can occur, albeit and its response time is fast.
rarely, in a home, business or gas vehicle. The loss of
this gas can be dangerous because it increases the risk There are about 30 crore users of LPG in the country
of explosion. An odor such as ethanol is added to the where the majority of 40% of the population. Several
LPG so that the losses can be easily detected by most standards have been put in place for the gas leak
people. However, some people with a reduced sense detection system. The existing system provides an
of smell may not rely on this intrinsic safety alarm system primarily to detect a gas leak in
mechanism. In these cases, a gas leak detector domestic and commercial premises. The main
becomes vital and helps protect people from the objective of this system is to monitor the leakage of
dangers of gas leaks. A number of research papers liquefied petroleum gas (LPG) in order to avoid
have been published on gas leak detection serious fire accidents and facilitate safety measures.
techniques. We have studied a research article about The system detects LPG leaks from the gas sensor and
our topic "LPG GAS WARNING DETECTION AND informs the consumer of the gas leak by sending a
LOSS SYSTEM". This document presents the text message.
detection and alarm system for LPG gases to prevent
fire accidents and ensure the safety of the home. The When the system determines that the LPG
LPG leak detection and reporting system has been concentration in the air reaches the specified level,

CSEIT1951128 | Received : 15 Feb 2019 | Accepted : 28 Feb 2019 | January-February -2019 [ 5 (1) : 531-534 ]
531
Prof. Pranay Meshram et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 531-534

tell the consumer by sending text messages to the Arduino is an open-source electronics platform based
registered mobile phone and informing people in on easy-to-use hardware and software. It
your home by activating the alarm including the has 14 digital input/output pins (of which 6 can be
alarm and the message display the message on the used as PWM outputs), 6 analog inputs, a 16 MHz
LCD screen to take required action. quartz crystal, a USB connection, a power jack, an
ICSP header and a reset button. Arduino boards are
WHAT IS IoT? able to read inputs - light on a sensor and turn it into
an output - turning on an LED, activating buzzer and
From homes and industries to enterprises, Internet
WiFi module for sending SMS. You can tell your
of Things have become commonplace in all walks of
board what to do by sending a set of instructions to
life. Connecting everything and everyone, IoT is
the microcontroller on the board. To do so you use
making the world smarter and better than ever
the Arduino programming language (based
before. However, this is just the beginning.
on Wiring), and the Arduino Software (IDE), based
Unlocking the full potential of Internet of Things
on Processing.
requires that businesses understand the
opportunities for value creation and systematically
Gas Sensor (Mq5) :
address the underlying challenges.
Generally, semiconductor sensors are used to detect
Internet of Things (IoT) is an ecosystem of connected LPG gas. MQ5 semiconductor sensor is used in this
physical objects that are accessible through the system. Sensitive material of MQ-5 gas sensor is SnO2,
internet. The ‘thing’ in IoT could be a person with a which with lower conductivity in clean air. When
heart monitor or an automobile with built-in-sensors, the target combustible gas exist. The sensor
i.e. objects that have been assigned an IP address and conductivity increases along with the rising gas
have the ability to collect and transfer data over a concentration. MQ-5 gas sensor has high sensitivity
network without manual assistance or intervention. to Propane, Butane and LPG, also response to Natural
The embedded technology in the objects helps them gas. The sensor could be used to detect different
to interact with internal states or the external combustible gas, especially Methane; it is with low
environment, which in turn affects the decisions cost and suitable for different application. The MQ-5
taken. Devices and objects with built in sensors are can detect gas concentrations anywhere from 200 to
connected to an Internet of Things platform, which 10000 ppm. The sensor's output is an analog
integrates data from the different devices and applies resistance.
analytics to share the most valuable information with
applications built to address specific needs. WI-FI Module :
The WI-FI technology is radio wireless networking
These powerful IoT platforms can pinpoint exactly technology. It is mainly used for communication
what information is useful and what can safely be purpose. In this project, the WI-FI system is used to
ignored. This information can be used to detect communicate with the user. When the gas leakage
patterns, make recommendations, and detect possible occurs, the microcontroller stops the leakage and
problems before they occur. alerts the surroundings. Then, the information about
the leakage has to be informed to the user. For this
II. METHODS AND MATERIAL purpose, WIFI is used. Using WI-FI, a warning is sent
to the user. The method of communication is
Arduino Kit : asynchronous serial communication. The

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


532
Prof. Pranay Meshram et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 531-534

corresponding code has to be loaded into the is connected to the Arduino kit. As soon as the gas
microcontroller, to which WI-FI is connected. level has exceeded this threshold value, the LED
lights up and the system warns the environment by
Buzzer: sending out a sound that is issued by the buzzer. At
It most commonly consists of a number of switches or the same time, the sound is sent via a WiFi module to
sensors connected to a control unit that determines if the user's mobile phone in order to inform the user of
and which button was pushed or a preset time has the loss. The user can avoid dangerous events such as
lapsed, and usually illuminates a light on the fire with the help of an exhaust fan.
appropriate button or control panel, and sounds a
warning in the form of a continuous or intermittent SYSTEM SENSO INDICATI FEATURE
buzzing or beeping sound. R ON
ARM Based MQ-6 Inform GSM
III. LITERATURE SURVEY Gas through Module is
Monitoring SMS used to
To overcome the issue of LPG leakage, many system System notify the
has been propossed. Gas leak detection is the process user
of identifying potentially hazardous gas leaks by Hazardous MQ- LCD Detects the
means of various sensors [2]. Several designs of LPG gas detection 4,MQ-7 Displays concentrati
detection and alert system have been proposed in the using arduino on
literature. Apeh et al. designed kitchen gas leakage Gas leakage MQ-2 Wireless Sends the
detection and automatic gas shut off system [3]. detector built Module informatio
T.Soundaryaet al. presented the cylinder LPG gas on n through
leakage detection system [4]. Wireless and GSM microcontroll wireless
technology [5] based gas detectors have also been ers module
proposed. This paper presents a LPG leakage IOT Based Electron Through Customer
detection and alert system to avoid fire accidents and real time ic sensor Social can check
to provide house safety. monitoring media the
gas detection previous
IV. PROPOSED SYSTEM system data and
determine
The proposed system uses the Arduino kit with an at which
integrated microcontroller. This Arduino kit forms condition
the core of our proposed system. Arduino is emergency
connected to other components that complement the occurred
system. These components are the MQ5 sensor, the
Gas leakage MQ-5 Through Notify the
LED light, the buzzer and the WiFi module. The
detection and buzzers and user
system uses an MQ5 gas sensor to detect LP gas. A
monitoring alarms through
threshold value (180 ppm for the proposed system) is
using IoT alarms
set in the MQ5 sensor. The sensor continuously
detects the surrounding atmosphere. For this
continuous air evaluation, the 5V power supply is
supplied to the Arduino via an AC adapter. A buzzer

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


533
Prof. Pranay Meshram et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 531-534

in April 2013.
[5] Online Monitoring of Green House Gas
Leakage inIndustries Angel College of
Engineering and Technology, Tirupur, Tamil
Nadu, India. In DecJan, 2014.
[6] Development of Movable Gas Tanker Leakage
Detection Using Embedded System
Mr.S.B.Patil, Dr.A.J.Patil Dec 2012.

Fig 1. System Block Diagram Cite this article as :

V. CONCLUSION Prof. Pranay Meshram, Nancy Shukla, Stuti


Mendhekar, Renuka Gadge, Shivani Kanaskar, "IoT
In this paper, LPG gas leakage is sensed by the MQ5 Based LPG Gas Leakage Detector", International
sensor. As soon as it detects LPG content in air more Journal of Scientific Research in Computer Science,
the threshold value, LED light glows and sound is Engineering and Information Technology
produced. Also, a SMS is send to user’s cell phone (IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
and thus notifies the user about the incident if the 531-534, January-February 2019. Available at doi :
user is out of the range of sound. The system can be https://doi.org/10.32628/CSEIT1951128
used to avoid fire accident and provide house safety. Journal URL : http://ijsrcseit.com/CSEIT1951128

VI. REFERENCES

[1] Shruti Bhoir1, Snehanjali Goregaonkar2, Prof.


ssShilpa Satre“ IoT Based Gas Detection
System” International Journal of Engineering
Science and Computing, April 2017. Volume 7
Issue No.4
[2] E. Jebamalar Leavline1, D. Asir Antony Gnana
Singh2 , B. Abinaya3 H. Deepika4 “LPG Gas
Leakage Detection and Alert System”
International Journal of Electronics
Engineering Research. ISSN 0975-6450 Volume
9, Number 7 (2017) pp. 1095-1097
[3] S.RAJITHA1 ,T.SWAPNA2 Electronics and
Communication Engineering, AP, India in
September-October 2012.
[4] Automatic LPG Booking, Leakage Detection
And Real Time Gas Measurement Monitoring
System B. D. Jolhe, P. A. Potdukhe, N. S. Gawai

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


534
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
Machine learning algorithm for Cyber Security - A Review
Mohammad Asif*, Pratap M. Mohite, Prof. P. D. Satya
Computer Science and Engineering, S.Y.C.E.T. Aurangabad, BATU, Lonere, Maharashtra, India

ABSTRACT

The computer networks are exposed to increasingly safety threats. With new kinds of attacks appearing usually,
growing flexible and adaptive protection-oriented strategies is a severe undertaking. In this context, anomaly-
primarily based community intrusion detection techniques are a precious era to guard target structures and
networks in opposition to malicious sports. Threats the internets are posing higher threat on IDS safety of
statistics. The primary concept is to utilize auditing programs to extract an in-depth set of capabilities that
describe each network connection or host session and practice statistics mining applications to learn rules that
correctly capture the behavior of intrusions and normal activities. Now Intrusion Detection has end up the
priority and on the crucial assignment of statistics protection administrators. A device deployed in a network is
at risk of numerous assaults and desires to be blanketed towards assaults. Intrusion detection machine is a
necessity of these days’ information safety area. It performs a vital function in detection of anomalous site
visitors in a community and indicators the network administrators to manage such visitors. The painting
supplied in this thesis is an attempt to locate such visitor’s anomalies in the networks through generating and
reading the site visitors float information.
Keywords : IDS (Intrusion Detection System), HIDS (Host Based Intrusion Detection System), ML (Machine
Learning), NIDS(Network Based Intrusion Detection System)

I. INTRODUCTION accelerated connectivity of laptop systems gives


greater get admission to outsiders and makes it less
This IDS supplied in this thesis implements the k- complicated for intruders to avoid detection. IDS are
means method of data mining for intrusion detection based on the perception that an intruder’s conduct
and the outlier detection technique the usage of can be exceptionally extraordinary from that of a
community outlier factor to locate the anomalies legitimate consumer. In view that many modern IDSs
present within the site visitors glide. The are built by using manual encoding of expert
performance of these two processes is compared expertise, modifications to IDSs are high priced and
through numerous confusion matrix and overall sluggish. in this paper, we describe a records mining
performance metrics like fake superb charge, framework for adaptively building Intrusion
sensitivity, specificity, type charge and precision, and Detection (ID) fashions. This record attempts a
an evaluation is accomplished to find out that which comprehensive compilation and categorization of
one of the tactics is higher for use for intrusion available intrusion detection gadget (IDS) products.
detection using site visitor’s waft. Intrusion detection lengthy studied and prototyped in academic and
is the hassle of identifying unauthorized use, misuse, authority’s circles, IDS have most effective within the
and abuse of computer systems by both machine last few years all started to emerge as a feasible and
insiders and outside penetrators. Namely, the useful commercial alternative. The primary industrial

CSEIT1951141 | Received : 20 Feb 2019 | Accepted : 28 Feb 2019 | January-February -2019 [ 5 (1) : 535-545 ]
535
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

IDS product becomes launched in 1991, with a ✓ It must impose a minimal overhead at the
relative handful emerging within the subsequent half system wherein it is walking; with a view to no
of dozen years. Then, inside the remaining couple of longer intervene with its ordinary operation.
years, the sphere underwent explosive increase. Even ✓ It needs to be capable of being configured in
after the obvious failure of numerous early releases, step with the security policies of the gadget this
there nevertheless remains at the least seventeen is being monitored.
extant products that claim to provide powerful ✓ It ought to be able to adapt to changes in device
intrusion detection in a networked environment. and person conduct over the years (e.g., new
Given this recent growth, and the reported improved applications being mounted, customers
utility of corporate assets to these products [1, 2, and converting from one hobby to another or new
3], the time has come for a complete assessment of assets being to be had those purpose
the subject. modifications in gadget aid utilization patterns).
✓ Anomalies are deviations from everyday
Intrusion detection structures try to discover laptop consumer behavior. Misuses, alternatively, are
misuse. Misuse is the overall performance of an recognized patterns of attack [5]. at the same
motion that isn't always favored by the machine time as misuse styles are frequently easier to
proprietor; one that doesn't comply with the device’s process and locate, it is frequently the paradox
ideal use and/or security coverage. IDSs mechanically patterns with a purpose to help to discover
examine online person hobby for forbidden (i.e., problems. As misuses are recognized styles of
invalid) and anomalous (i.e., strange, inconsistent) assault, the detection machine tends to fail
conduct. they're primarily based on the speculation whilst novel assault techniques are
that monitoring and reading community implemented. Detection of anomaly patterns is
transmissions, machine audit information, software computationally pricey due to the overhead of
audit data, device configuration, facts files, and keeping track of, and possibly updating several
different information can locate misuse. This record machine profile metrics, because it need to be
encompasses widespread portions of information, tailor-made gadget to device, and every now
effective analysis calls for detection1 specialized and and then even used to a person, due to the
constantly honed expertise, and at the least close to reality conduct styles and gadget usage vary
real-time of misuse is frequently important. significantly.
In [4], the subsequent characteristics are identified as ✓ Different IDSs had been designed to do
suited for IDS: dispensed collection and analysis of records. A
✓ It needs to run constantly with minimal human hierarchical device is defined in [6], and [7]
supervision. describes a cooperative gadget without a central
✓ It has to be fault tolerant in the sense that it has authority. these structures clear up maximum
to be capable of recover from gadget crashes, of the problems mentioned except for the
both unintended or because of malicious reconfiguration or adding skills to the IDS,
activity. Upon startup, the IDS have to be which aren't described in either of the 2
capable of recovering its previous country and designs.
resume its operation unaffected.
✓ The boom of the net has added excellent
✓ It ought to face up to subversion. The IDS need
benefits to society at the identical time the
to be able to screen itself and stumble on if it's
growing attacks on the IT Infrastructure are
been modified by way of an attacker.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


536
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

getting an increasing number of serious issues The growth of Internet has brought great benefits to
and wishes to be addressed. Together with the the society at the same time the growing attacks on
growth of the internet, attacks also are growing the IT Infrastructure are becoming an increasingly
in parallel. serious issue and needs to be addressed. Along with
✓ Host-based intrusion detection system pursuits the growth of Internet attacks are also growing in
at detecting the intrusions at the host stage. parallel. In earlier days, the attacker should have a
They function in my opinion at each host of the good knowledge about the target infrastructure and
community. Consequently, they are able to knowledge on the Network, Operating Systems &
operate on distinctive type and quantity of Applications. Whereas today there are lots of open
records on the equal time. The host-based tools available in the Internet which can trigger
machine is dependent on the host operating automated attacks.
system. Any vulnerability inside the host-
primarily based OS can weaken the integrity of
the host-primarily based.

An outsider can exploit this vulnerability to release


an assault on the way to be hard to be detected
through the HIDS. as a result, a robust HIDS have
to be supported with the non-susceptible host OS.
The bodily deployment of HIDS in a community
may be visible the traffic enters from the internet
via the router or switch or firewall to the local
network. It is the responsibility of the HIDS in the
host to detect possible intrusions in the traffic flow.
Figure 2. Growth of Internet in terms of Host Count

WannaCry attack: In might also 2017, the


WannaCry Ransomware spread via the net, the use
of a make the most vector named EternalBlue. The
ransomware attack infected greater than 230,000
computer systems in over a hundred and fifty
countries the usage of 20 specific languages to
demand cash from users the usage of Bitcoin crypto
currency. WannaCry demanded US$300 in line
with pc. [9]

Petya attack: Petya computer virus spread in the


Figure 1 : Host based intrusion detection system course of April 2016, this malware infected the
grasp boot record of the laptop through encrypting
II. RELATED WORK the report tables of the NTFS file system. as soon as
infected on the next boot expects a ransom is paid.
again in the month of June 2017, a modified model

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


537
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

of Petya the usage of Eternal Blue exploit and this consideration as the collection deviation from
was aimed to create disruption alternatively to normal profile sequence. thus, this device works
generate income. [9] offline the use of previously accrued information
and implements the view desk set of rules for
We located that there are many benefits of gaining knowledge of software profiles significantly.
C4.5algorithms for special attacks occurs on your Duan et al. [12] have concentrated on identifying
dataset and C4.5 will detect the R2L and U2R compromised machines which can be recruited to
attacks and the neural network is the use of for hit upon junk mail zombies. An method SPOT is
detecting the DOS and Probe attacks and many proposed to experiment sequentially outgoing
greater. In this, we are the usage of four algorithms. messages by means of enforcing SPRT (Sequential
The primary one is okay-manner clustering and the possibility Ratio check). This method quick
second steps are fuzzy good judgment 1/3 steps are estimates whether a number is compromised or not.
SVM and the remaining and very last step is C4.5. figuring out compromised machines using malware
The blessings of all this set of rules are to detect the contamination device is said via Bot hunter [13].
assaults from the datasets. After that, it will
examine it with the SVM and C4.5 classifiers to This machine has massive no of steps that allow
find out how many attacks are coming about in the intrusion detection alarms correlation caused using
dataset at the same time as transferring the dataset inbound visitors with outgoing message trade
from supply to destination. [8] sample consequences. Bot Sniffer [14] explained in
his paintings approximately compromised device
This center also evolved a subsequent-era traits which might be a uniform temporal-spatial
mechanism which includes audit profiles of conduct for detecting zombies. This approach
person’s and may monitor the modern-day identifies zombies by using combining flows based
reputation of the consumer, if any change takes totally on server connections and searching flows
place with person’s hobby as compared with audit with comparable conduct respectively. Kumar and
profile of person then it will generate an alarm. Goyal [15] have explained implements genetic
algorithms in dataset training to categories the
Haystack [10] later developed a framework to labels that are smurf attacked and achieves a low
estimate an intrusion detection approach based false high-quality ratio of 0.2%. in addition,
totally on person and anomaly strategies. Six paintings were completed through Abdullah [16]
varieties of intrusion have been detected and and co-people elaborated intrusion detection class
people include the masquerade assaults, malicious regulations using genetic algorithms. Intrusion
use, leakage, carrier denial, the unauthorized detection policies the usage of genetic algorithms
consumer’s wreck-ins try, and get admission to becomes also the have a look at made through
control of a protection system. The source fire Ojugo et al. [17]. This approach uses health
developed suggests a community-based intrusion characteristic for estimating the rules. system
detection and prevention mechanism referred to as learning strategies are also applied to hit upon the
snicker machine that is an open supply. Forrest [11] intrusion. current device mastering strategies
in 1996 created an ordinary profile based on (artificial Neural Networks - ANN) for intrusion
studying the call sequences between intrusion detection was described by way of Roshani group
detection and protection in opposition to a human [18]. Gaikwad et al [19] added a technique based on
machine. An assault on this device is taken into fuzzy clustering and ANN approach.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


538
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

which the statistics is destined can offer the


PROBLEM STATEMENT :- A few of the existing attacker the possibility of acting Insertion and
network- and host-based totally IDSs [20, 21] Evasion assaults. those assaults employ
perform records collection and analysis centrally mismatched assumptions inside the community
using a monolithic architecture. by using this, we protocol stacks of different hosts to hide the
mean that the facts is accrued through a single assaults or create a denial of- provider attacks.
host, both from audit trails or via tracking packets different IDSs had been designed to do
in a network, and analyzed by means of a dispensed collection and analysis of facts. A
unmarried module the usage of one of a kind hierarchical device is defined in [25], and [26]
techniques. other IDSs [22, 23] perform dispensed describes a cooperative gadget without a
facts collection (and some preprocessing) with the government. those systems resolve most of the
aid of the use of modules disbursed inside the hosts problems cited besides for the reconfiguration
which are being monitored, however the or including abilities to the IDS, which aren't
accumulated statistics remains shipped to a crucial described in either of the 2 designs.
vicinity in which it is analyzed through a ✓ Make a larger framework to resource
monolithic engine. an awesome overview of unsupervised ML: The feature choice" step
systems that take both methods is presented in within the framework currently requires
[24]. There are a number of issues with those labeled datasets to discover the most
architectures: discriminative talents. Destiny art work will
✓ The imperative analyzer is a single factor of dispose of this framework requirement to
failure. If an outsider can by some means save manual unsupervised ML with unlabeled
you it from operating (for instance, by crashing records. big effort in this thesis went into
or slowing down the host wherein it runs), the ensuring experiments have been representative
complete community is without safety. of actual-worldwide situations. This ensured
✓ Scalability is constrained. Processing all of the our outcomes have been considerable in
statistics at a unmarried host implies a limit on modern-day networks our datasets have been
the dimensions of the network that can be snapshots of community site visitors which
monitored. After that limit the vital analyzer may be analyzed opine in batch mode. For the
turns into not able to preserve up with the detectors to artwork on a live network.
town of statistics. Disbursed data collection can
also motive problems with excessive facts
visitors within the network.
✓ It’s far hard to reconfigure or add abilities to
the IDS. Changes and additions are generally
carried out by way of editing a configuration
report, including an entry to a desk or putting
in a brand new module. The IDS commonly
must be restarted to make the modifications
take impact.
✓ Analysis of network facts can be awed. As
shown in [20], appearing a set of network
information in a number other than the only to

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


539
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

supervised scheme implementing basis of K-means


and KDD CUP algorithms.

Logs documents need to be tested to understand any


compromised bills, originating IP’s, and all sources
accessed via using the attacker. All related sports want
to be collected and examined several weeks or even
months in advance than the detected event. capability
areas of future paintings are computerized correlation
and assessment of the log facts from cyber-attacks.
additional machine mastering algorithm.

PROPOSED SYSTEM: -According to base paper the


machine learning deal with two concept i.e.
supervised learning and unsupervised learning. Both
concept have some limitations, to overcome this
limitation we can concern with the concept of semi
supervised approach. In semi-supervised approach
concern two scenarios. In this approach front-end to
be implementing using of K-means Algorithm and
back-end to implementing using of KDDCUP 99. In
k-means algorithm concern to secure to be user
interface and KDD CUP 99 concern as to centralized
data storages on the system or host.
There are numerous procedures and methods utilized
in id. each method has deserved and demerits.
therefore, this paper highlights the similar
Figure 3 : Classification of ml algorithms for cyber distribution of attacks nature with the aid of using ok-
security applications [28] way and also the powerful accuracy of the Random
forest set of rules in detecting intrusions. This paper
Machine learning algorithms fragmented as describes complete sample popularity and machine
supervised learning and unsupervised learning. learning set of rules overall performance for the four
Behave of the scheme as to be defined as, supervised attack categories, which include Denial-of-service
learning which to be deal with more sophistication (DoS) attacks (deny legitimate request to a system),
rather than unsupervised learning but it has too costly Probing attacks (statistics gathering attacks),
and complexity. consumer-to-root (U2R) assaults (unauthorized access
to nearby splendid-consumer), and faraway-to-
But in case of unsupervised learning it takes a huge
neighborhood (R2L) attacks (unauthorized
amount of input but produce low amount of desirable
neighborhood get admission to from a faraway device)
output. The supervised scheme work under the expert
proven inside the KDD Cup 99 intrusion detection
system.
dataset.
To overcome the limitation of this, we can go with
semi supervised learning scheme. This semi

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


540
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

channeled effectively, vastly, it feasible to deal with


size of network facts and heterogeneity [29].

In this project, we use k-means algorithm to cluster


dataset connections. The k-means set of rules is one
of the widely recognized clustering tools.
K-means agencies the information according to
with their characteristic values right into a person-
particular variety of ok wonderful clusters. Facts
categorized into the same cluster have identical
characteristic values. k, the fantastic integer
denoting the number of clusters, desires to be
Figure 4 : - Overview of IDS furnished earlier. the steps involved in a k-means
algorithm are given consequently: [30]
III. METHODOLOGY
This project consists of the communication of the 1 1. K points denoting the data to be clustered are

algorithm of data mining class methods. Those is k- placed into the space. These points denote the
means primary group centroids.
2. The data are assigned to the group that is
K-means Clustering Algorithm: - adjacent to
the centroid.
Clustering, primarily based on distance 3. The positions of all the K centroids are
measurements carried out on items, and classifying recalculated
gadgets (invasions) into clusters. not like type, as soon as all the data are assigned.
classification because there may be no information 4. Repeat steps 2 and 3 until the centroid
about the label of studying statistics is an unattended unchanged.
getting to know system. For anomalous detection, we
will use welding and in-intensity evaluation to guide This consequence within the partition of facts into
the identity model. Dimension of distance or groups. The preprocessed dataset partition is
similarity performs a critical function in collecting achieved the usage of the ok-means set of rules
observations into homogeneous corporations. with k value as 5. due to the fact we have the
dataset that contains regular and 4 assault
Jacquard affinity size, the longest not unusual order categories which include DoS, Probe, U2R, R2L.
scale (LCS), is vital that the occasion is to evoke the
scale to decide if regular or odd. Euclidean distance is
about two vectors X and Y in space Euclidean n-
dimensions, the size of the distance widely used for
vector area. Euclidean distance can be defined
because the rectangular root of the whole difference
of the same vector dimension. Subsequently,
grouping and category algorithms want to be

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


541
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

KDDCup’99 dataset has variations of education


dataset; one is a complete education set having 5
million connections and the opposite is 10% of
this schooling set having 494021 connections.
since the whole dataset is huge, the test has been
completed on its smaller amount of dataset this is
10% of KDD. moreover, the KDDCup’99 dataset
consists of many assault behaviors, categorized
into 4 companies: Probe, Denial of provider (DoS),
user to Root (U2R), and far-flung to local (R2L)
[33]. these may be visible in desk I. normal
connections are created to profile than predicted
in a military network. The detailed information of
the two versions of schooling dataset may be seen
in table 1.

The KDDCUP 99 was simulated in a army


community environment and used for The 0.33
global understanding Discovery and information
Mining equipment opposition, which became held
alongside KDDCUP 99 The 5th international
convention on knowledge Discovery and facts
Mining. The competition challenge changed into
KDD Cup 99 Dataset: - The assessment of any to study a predictive model or a classifier able to
intrusion detection algorithm on actual network distinguishing among valid and illegitimate
data is extraordinarily tough particularly because connections in a computer community. This
of the high fee of acquiring proper labeling of dataset includes one form of normal records and
community connections. due to the actual pattern 24 distinctive sorts of assaults which might be
Table be gotten for intrusion detection, the categorized into four kinds such:
KDDCup’99 datasets are used as the sample to
confirm the overall performance of the misuse
detection model. The KDDCup’99 datasets,
referred by way of Columbia college, became
arranged from intrusions simulated in military
community surroundings on the DARPA in 1998.
It includes network connections obtained from a IV. CONCLUSION AND FUTURE SCOPE
sniffer that facts all network visitors the use of the In this project we can sophisticate more on K-
TCP unload layout. The overall simulated length is means rather than KDDCUP 99, because it
seven weeks. It was carried out in the MIT Lincoln valuable to aspect of semi supervisor concept. This
Labs after which announced at the UCI KDD Cup scheme facility to labeled and unlabeled dataset as
1999 Archive [32]. efficiency.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


542
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

This paper provides a comparative evaluation [6] S. Staniford-Chen, S. Cheung, R. Crawford,M.


hybrid system getting to know approach to detect Dilger, J. Frank, J. Hoagland, K. Levitt,C. Wee,
the Denial of service (DoS) assaults, Probing R. Yip, and D. Zerkle. GrIDS: A graphbased
intrusion detection system for large
(Probe) assaults, consumer-to-Root (U2R) attacks,
networks.In Proceedings of the 19th National
and far off-to-neighborhood (R2L) attacks. we can
InformationSystems Security Conference,
understand the same nature of assault organization volume 1, pages 361-370. National Institute of
by the use of ok-manner algorithm. And then we Standardsand Technology, October 1996.
use to categories ordinary and attack connections. [7] Gregory B. White, Eric A. Fisch, and Udo
The experiments display that KDD Cup 99 dataset W.Pooch. Cooperating security managers: A
can be carried out as an effective benchmark peer basedintrusion detection system. IEEE
dataset to help researchers evaluate distinctive Network, pages 20-23, January/February 1996.
[8] Intrusion Detection System by using K-Means
intrusion detection fashions. future paintings Clustering, C 4.5, FNN, SVM Classifier Akshay
Takke1, Ravikumar Gujjul2, Mikhil Ghag3 ,
consist of analyzing with other records mining Vivek Pawar4, Vivek Pandey5 Page no:-636
algorithms to categories attack categories and how [9] International Journal of Advanced Research in
it could hit upon on different real-time Computer Science REVIEW PAPER Available
environment dataset. Online at www.ijarcs.info© 2015-19, IJARCS
All Rights Reserved 356 ISSN No. 0976-
We implement this project as more feasible as the
5697INTRUSION DETECTION SYSTEMS: A
basis of K-means algorithm.
REVIEW D. Ashok Kumar,S. R. Venugopalan
Page no:- [356-357]
V. REFERENCES
[10] Patcha, A. and Park, J. M. An overview of
anomaly detection techniques: Existing
[1] Adaptive Network Security: Solutions for solutions and latest technological
Managing Risk in an Interconnected World, trends.Computer Networks, 51(12);2007; 3448–
AberdeenGroup, Vol. 11, No. 5, January 1998. 3470.
[2] Hacker Stoppers? -- Companies Bought $65 [11] Forrest, S., Hofmeyr, S. A. , Somayaji, A. and
Million Worth Of Network-Intrusion Tools Longstaff, T. A. A Sense of Self for Unix
Last Year, But Capabilities Still Lag What's Processes, IEEE Symposium onResearch in
Promised, Information Week, April 1998 Security and Privacy, Oakland, CA, USA,
http://www.techweb.com/se/directlink.cgi?IW 1996;120--128.
K19980420S0066 [12] Duan, Z., Chen, P., Sanchez, F., Dong, Y.,
[3] Adaptive Network Security Management: Stephenson, M. and J. M. Barker, M. (2012).
Intrusion Detection and Security Assessment Detecting spam zombies bymonitoring
Come of Age, The Yankee Group Data outgoing messages, IEEE Trans. Dependable
Communications Report, Vol. 13, No, 10, June and Secure Computing, Apr 2012; 9(2):198–210
1998. [13] Gu, G., Porras, P., Yegneswaran V., Fong, M.
[4] Mark Crosbie and Gene Spafford. Active and Lee, W. BotHunter: detecting malware
defense of a computer system using infection through IDS-drivenialog correlation,
autonomous agents. Technical Report 95-008, Proc. of 16th USENIX Security Symp. (SS ’07),
COAST Group, Departmentof Computer Aug. 2007; 12:1–12:16.
Sciences, Purdue University,West Lafayette, IN [14] Gu, G., Zhang, J. and Lee, W. (2008).
47907-1398, Feb 1995. BotSniffer: detecting botnet command and
[5] Kumar S, Spafford EH (1994) An Application
ofPattern Matching in Intrusion Detection. control channels in network traffic, Proc.
TechnicalReport CSD-TR-94-013. Purdue Of15th Ann. Network and Distributed Sytem
University.
Security Symp. (NDSS ’08),

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


543
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

[15] Goyal, A. and Kumar, C. .GA-NIDS: A Genetic for detecting network intrusion and misuse.
Algorithm based Network Intrusion Detection Computers and Security, 12(3):235-248, May
System, Electrical Engineering and Computer 1993.
Science, North West University, Technical [23] S. R. Snapp, J. Brentano, G. V. Dias, T. L. Goan,
Report;2008.Feb. 2008. L. T. Heberlein, C. Ho, K. N. Levitt, B.
[16] Abdullah, B., Abd-algafar I., Salama G. I. and Mukherjee, S. E. Smaha, T. Grance, D. M. Teal,
Abd-alhafez A. Performance Evaluation of a and D. Mansur. DIDS (Distributed Intrusion
Genetic Algorithm BasedApproach to Network Detection System) - Motivation, Architecture,
Intrusion Detection System, Proceedings of and an early Prototype. In Proceedings of the
13th International Conference on Aerospace 14th National Computer Security Conference,
Sciences andAviation Technology (ASAT-13), pages 167-176, October 1991.
Military Technical College, Cairo, Egypt, [24] Biswanath Mukherjee, Todd L. Heberlein, and
2009;1-5. Karl N. Levitt. Network intrusion detection.
[17] Ojugo, A. A., Eboka, A. O., Okanta, O. E., Yora, IEEE Network, 8(3):26-41, May/June 1994.
R. E. and Aghware, F. O.Genetic Algorithm [25] Thomas H. Ptacek and Timothy N. Newsham.
Rule-Based IntrusionDetection System Insertion, evasion, and denial of service:
(GAIDS), Journal of Emerging Trends in Eluding network intrusion detection. Technical
Computing and Information Sciences, report, Secure Networks, Inc., January 1998.
3(8);2012; 1182 – 1194. [26] S. Staniford-Chen, S. Cheung, R. Crawford, M.
[18] Roshani Gaidhane, Vaidya, C. and Dilger, J. Frank, J. Hoagland, K. Levitt, C. Wee,
Raghuwanshi, M. Survey.Learning Techniques R. Yip, and D. Zerkle. GrIDS: A graph based
for Intrusion Detection System intrusion detection system for large networks.
(IDS),International Journal of Advance In Proceedings of the 19th National
Foundation and Research in Computer Information Systems Security Conference,
(IJAFRC) Feb 2014. ISSN 2348 – 4853, volume 1, pages 361-370. National Institute of
2014;1(2). Standards and Technology, October 1996.
[19] Gaikwad, Sonali Jagtap, D.P. Kunal Thakare [27] Gregory B. White, Eric A. Fisch, and Udo W.
and Vaishali Budhawant. Anomaly Based Pooch. Cooperating security managers: A peer
Intrusion Detection System UsingArtificial based intrusion detection system. IEEE
Neural Network and fuzzy clustering., Network, pages 20-23, January/February 1996.
International Journal of Engineering Research [28] Review Paper on Shallow Learning and Deep
& Technology (IJERT), ISSN:2278-0181, Learning Methods for Network security Afzal
November- 2012; 1(9). Ahmad1*, Mohammad Asif2, Shaikh Rohan Ali
[20] L. Heberlein, G. Dias, K. Levitt, B. Mukherjee, 3 (page: -52) 2018
J. Wood, and D. Wolber. A Network [29] Youssef Ahmed and Ahmed Emam, “Network
SecurityMonitor. In Proceedings of the IEEE Intrusion Detection Using Data Mining and
Symposium on Research in Security and Network Behavior Analysis”, International
Privacy, May 1990. Journal of Computer Science & Information
[21] R. Heady, G. Luger, A. Maccabe, and M. Technology (IJCSIT) Vol 3, No 6, Dec 2011.
Servilla. The Architecture of a Network Level [30] X. Wu, V.Kumar, Ross Quinlan, Joydeep
Intrusion Detection System. Technical report, Ghosh, Qiang Yang, Hiroshi Motoda, Geoffrey
University of New Mexico, Department of J. McLachlan, Angus Ng, Bing Liu, Philip S. Yu,
Computer Science, August 1990. Zhi-Hua Zhou, Michael Steinbach, David J.
[22] Judith Hochberg, Kathleen Jackson, Cathy Hand, and Dan Steinberg, “Top 10 algorithms
Stallings, J. F. McClary, David DuBois, and in data mining”, Survey Paper(2008).
Josephine Ford. NADIR: An automated system

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


544
Mohammad Asif et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 535-545

[31] Intrusion Detection Based On Clustering Cite this article as :


Algorithm Nadya El MOUSSAID 1, Ahmed
TOUMANARI 2, Maryam ELAZHARI 3 Page- Mohammad Asif, Pratap M. Mohite, Prof. P. D. Satya,
1062 "Machine learning algorithm for Cyber Security - A
[32] P. S. Rath, M. Hohanty, S. Acharya and M. Review", International Journal of Scientific Research
Aich, “Optimization of IDS Algorithms Using
in Computer Science, Engineering and Information
Data Mining Technique”, Proceeding of 53rd
Technology (IJSRCSEIT), ISSN : 2456-3307, Volume
IRF International Conference, Pune,
India,ISBN 978-93-86083-01-2, 2016. 5 Issue 1, pp. 535-545, January-February 2019.
[33] L.S. Parihar and A. Tiwari, “Survey on Journal URL : http://ijsrcseit.com/CSEIT1951141
Intrusion Detection Usingn Data Mining
Methods”,IJSART, , Volume-2 Issue-1 ISSN
(online: 2395-1052) January-2016.

Authors Profile

Mohammad Asif received his B.E. degree


from NMU JALGAON INDIA.
Currently pursuing M.Tech.in Computer
Engineering from S.Y.C.E.T. Aurangabad
Affiliated to BATU Lonere.

Prof. Pratap Mohite B.Tech CSE


M.E.(Software Engineerng), Assistant
Professor Department of computer
Science and Engineering Shreeyash
College of Engineering and Technology,
Satara Parisar,Beed Bypass Road
Aurangabad.

Prof. Satya P.D. B.Tech. CSE, M.Tech.


and P.hD.(Pursuing), Assistant Professor
Department of computer Science and
Engineering Shreeyash College of
Engineering and Technology, Satara
Parisar,Beed Bypass Road Aurangabad.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


545
International Journal of Scientific Research in Computer Science, Engineering and Information Technology
© 2019 IJSRCSEIT | Volume 5 | Issue 1 | ISSN : 2456-3307
DOI : https://doi.org/10.32628/CSEIT1951100
Analysis and Importance of Deep Learning for Video Aesthetic Assessments
Manish Bendale, Madhura V. Phatak, Dr. Nitin N. Pise
Department of Computer Engineering, MIT, Pune, Maharashtra, India
ABSTRACT

Deep Learning is one of the active analysis topic obtaining a great deal of analysis attention recently. This
increase in analysis interest is driven by several area as that are being worked on like machine-based reality
finding, good over-seeing, sensory activity recognition, online learning, world of advertisement, text analysis
and so on. Videos have specific characteristics that make their method unique. Visual aesthetic typically:
Remember what they see, understand and learn rather than what they hear. This paper principally emphasizes
deep learning on basics of automatic video aesthetic assessments.
Keywords : Video Aesthetics, Preprocessing Techniques, Feature Extraction, Deep Learning, Applications.

I. INTRODUCTION It is to be noted that ''Deep Learning'' does not


mingle with the standard of computers emotions. It
Computer technology are playing a very important would be imperative to first begin with human
and necessary role in our daily life. Mainly brain structures, to get a basic grasp on them and
computer involved already in many tasks of thus to outline the relation with computers. Then
gathering information using calculation, logic, we based on human psychology, the computers state of
can say computers are very well predict complex being are going to be put into use under the
events in the future like the real condition of thought of Deep Learning and its completely
weather, can able to manage complex systems and different points of view.
architecture like nuclear and atonic power plants, II. DEEP LEARNING
but still we do not take into account the state they
2.1 How video aesthetics are related to Deep Neural
are in the form of ''living''. Computers are very
Network?
different from humans from the perspective of
“thinking of mind”. As for the person, feeling
Deep learning is a part of machine learning
behind beauty could be a special kind of having
technology which has ability to learn and mimic the
thoughts; its Target is special outlines not forever
working principal of human brain. To mimic the
in agreement like “Value”. Because the underlying
brain working structure, deep neural network
values become completely different, from our
required large amount of datasets [2] to learn. Deep
psychological thinking amendment. There’s no
learning is a subset and the new research area of
great purpose or quality not like between different
machine learning and its original goal is: Artificial
science like calculation or physics. So, to bridge
Intelligence. Deep learning has ability to learn from
over the limit of computers, the new research
the multiple levels of representation and the
concept of ''Deep Learning'' is introduced by
abstraction, these layers are helps in processing the
Walter Pitts and Warren McCulloch in 1943 [1].
amount of data like, images, videos, text, sound, etc.

CSEIT1951100 | Received : 20 Feb 2019 | Accepted : 28 Feb 2019 | January-February -2019 [ 5 (1) : 546-554 ]
546
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

Deep learning is a part of machine learning Aesthetics assessment is a subjective field [4] because
technology which has ability to learn and mimic the individual preferences differ according to personal
working principal of human brain. To mimic the taste too, thus what may be pleasing to one person
brain working structure the deep learning technology may not be pleasing to the other. Therefore different
required more number of datasets [3]. communities define aesthetics differently, based on
psychological and emotional aspects [8].
There is a different types of media aesthetics visions,
video aesthetics is one of the aesthetics, due to the
increasing in the popularization of high quality smart
phones, the videos and image aesthetics are in great
demands. In the current market the videos are
playing important role in various fields like,
advertisement, video conferencing, industrial
representation etc. Previous works done using
machine learning approaches, also called as
handcrafted method in which, feature extraction 2.3 Comparisons of Deep Learning with previous
were a common ways and methods for aesthetic method about Video Aesthetic Assessment
analysis. CNN is recently used method to solve
critical and complex problems as state-of-arts All the present system is based on the handcrafted
approaches [4]. methodologies as we find in literature. Hand-crafted
feature for images and videos aesthetics is good
feature extractors based on user preferences
2.2 Which are the features that can be identified?
algorithms, to design this features required a
considerable amount of engineering skill and domain
Many other specific photographic rules can be sued expertise. The main limitation is the excessive time
to discover particular aesthetic of videos by directly consumption and the accuracies of aesthetic
measuring physiological facts. This understanding assessments. There are many researches done on the
and captured information is used to processed and image aesthetics using hand crafted as well as deep
extract meaningful information from the gathered learning methods [16].
data. In the earlier research, aesthetics was defined
by Handcrafted and learned features [6]. Video aesthetics using deep learning, there exists a
scope for future research as this field is less explored
[19]. The table (table 1) shows the difference
The aesthetic quality of images and videos are judged
between the hand crafted and deep learning
by their low, middle and high level properties. The
technology:
low level properties of an image are that of color,
texture, edges and intensity [7]. The middle level
Deep Learning
property is the object in the image or video. And the Handcrafted Method
Method
high level properties are the photographic rules,
Handcrafted features Deep learning
mainly comprising of the Rules of Thirds (RoT),
judge by commonly method work on
Visual Balance (VB), Diagonal Dominance (DD),
established highly and accurate
Simplicity and the Depth of Field (DoF) [7].

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


547
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

photographic rules, feature extractions 2.4 Areas of Deep Learning related to video aesthetic
mainly low and middle by using hidden assessments
level features. layer concept. • Detecting and recognizing aesthetic information
This is sequential step- This is parallel Aesthetics is a study that relates to the relationship
by-step method and method using hidden between the mind and emotions in assessing beauty
time consuming to layers concept and [8]. The aim of aesthetics analysis is to define the
extract and classification required less amount science in assessing the aspects of arts and beauty
of features. of time. sought in images and videos. Human brain thinking
This method require This method structure begins with gathering information from
engineering skills and depends on the different environment which take fact about users
human experience and amount of dataset to psychology and understanding with behavior. The
do not need large train the network for fact gathered are of the form of features which are
dataset to train network. accuracy. contains in a video and this features used to perceive
Backpropagation is not Backpropagation is people understanding behind videos beauty. For
allowed to correct the main advantage of example, a camera might take pictures of the any
final results. CNN. object; Then human mind start its working and
Classification done at Allow to use more identify the object position, color contrast between
once only. fully connected layer foreground and background, Depth of Field, and etc.
for classification. • Psychology in machine
Another area within video aesthetic computing is
Table 1. Difference between the handcrafted and that the design of procedure systems to exhibit either
deep learning method. innate robust mind psychological powers or that is in
a position of as definitely acting the part of
There are the many researcher working on the deep aesthetics. An additional helpful move involving
learning to improve more accurate result in different computers based mostly on current technology based
field also. In the today life, the video aesthetic is the powers, is that the simulation of aesthetic in taking
important multimedia application. Many field use the pleasure in communication agents to boost
video for improver the selling quantity of the interactivity between human like and machine like
products. So, the video aesthetic is self-increasing communications. Thinking ability in machines could
area of aesthetic and in the future, it will go on the be connected with define states connected with
great demand. The deep learning provides a good forward development (or exist without of forward
opportunity for the video aesthetic assessment to development) of a pc primarily based learning
improve the aesthetic time and the accuracy of video system supported rules [8].
aesthetic editing. In this method, we deploy the
2.5 Another applications related to video aesthetic
multiple method of photography on the neural
network and extract the good feature using the Consider online-learning applications based on
hidden layer of neural network. computers wherein computing the users feelings can
be used to adjust the conference learning, content
can be more polished and tasteful form may be used
while teaching when a learner is uninterested, or
paused. Human Psychological being healthy,

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


548
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

suggestions behind aesthetic, can help to the users, 2.6 Different Algorithms Employed by Researchers
based on current emotion behind video beauty by In the work on regression and classifying video
computing applications when working out a aesthetic, the most frequently used classifier
particular user’s state of feelings [9]. algorithms are Artificial Neural Networks (ANN),
Deep Neural Network, Hidden Markov Models
Visual aesthetic has possible applications in human
(HMMs), Gaussian Mixture Model (GMM), Support
computer interaction, now a day human computer
Vector Machines (SVM), k-nearest neighbor (k-NN),
interaction is varying area in different fields. Many
and Decision Tree Algorithms. Numerous studies
researchers are working on gesture technology, in
have decisively showed that choosing the acceptable
which camera capture the video of moving body
classifier will meaningfully improve the performance
gesture and according to movement of gesture the
of the system.
machine will understand, what action required on
the basis of gesture. Virtual reality is also a greater Below gives a brief description about algorithm:
area in human computer interaction, which shows
• K-NN – Classification prediction by defining the
the 3D video and create environment according to
object in the feature space, and comparing this
human psychology and emotion at the time of
object with “k” nearest neighbors. In which the
interactions.
majority vote predict the classification of
gathered data.
Video aesthetics in Advertisements [10]: Now a day,
• GMM – it is a probabilistic based model used for
the increasing in the population the requirement of
signifying and classification of the presence of
the peoples goes on high demand. Due to renewed
sub part within the whole pats. Every sub part is
and changing consumer demands and the rapidly
outlined exploitation the mixture distribution
developing technological factors, companies need to
which permits for classification of explanations
search of new strategies to make a difference in
into the sub-parts [27].
their products. As competition increased marketers
• SVM – is used to binary linear classification and
started to focus on new approaches for product to
it is also used for regression, which selects in the
attract consumers' perception and attention. One of
two (or more) different classes, each input may
the most effective ways of differentiating is using
fall into this class [28].
video aesthetics.
• ANN – is a mathematical model, which has
In gamming area a mistake that’s unfortunately only ability to mimic brain structure of working
too common within the gaming business, is to lump process, it is enthused by biological neural
video graphics and video aesthetics together as if they networks that can work and extract feature
were the same factor. This, unnecessary to say, is similar to human mind [29].
false. The distinction becomes obvious once one • Decision Tree algorithms – are supported
compares, as an example, a heavily artificial, cartoon succeeding a decision tree throughout that the
game that aims for realism specially else. The leaves of the tree represent the classification
former’s video aesthetic, despite its presumptively result, and branches represent the conjunction
smaller budget and scope, can possibly hold up longer of consecutive features that lead to the
than the latter’s. classification [30].
Video aesthetic is also being applied to the growth • HMMs – An applied mathematics markov model
of communicative technologies for use by people during which the conditions and state
[10]. transitions are unobtainable for the observation.

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


549
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

In its place, the series of outputs dependent on recognize and classify by the deep learning method
the states are visible [31]. to predict accurate results.

Other applications are like advertisement [20] and


2.7 Artificial Intelligence and Deep Learning online learning, is great in demands, and video
In ‘Streaming Media magazine by Nadine Krefetz aesthetic has ability to make perfect communication
[33]’, defined Video AI (artificial intelligence) has the between the industries and their client. And for the
capability to resolve variety of time-consuming, child, learning environment creation according to
video-related issues with automation. However that child psychology is great aspects of AL and deep
doesn’t mean its supernatural powers which will learning in video assessment areas.
exclude human brain thinking behind beauty control. III. FUTURE SCOPE
To supply a way of where video AI is in early 2018,
what follows are variety of real-life examples during 3.1 Deep learning and Market of video aesthetic
which AI helps to feature structure to the application
unstructured world of video aesthetic assessment. It is likely that the Deep Learning market will
revenue in the Video Aesthetic Advertising segment
At current years, there are many hype of applied
amounts to US$27,799m in 2018. Revenue is
science and especially within the field of artificial
expected to show an annual growth rate (CAGR
intelligence. Inspiring the machines to find out from
2018-2023) of 13.7%, resulting in a market volume of
the information which are gathered from different
US$52,760m by 2023. The average revenue per
source and build accurate decisions within the areas
Internet user currently amounts to US$9.34 [34].
like in the field of education, in medical and
healthcare industries, finance and etc. There are Deep Learning is about to revolutionize the approach
multiple applications which are the mix of each organizations, particularly across sectors like retail,
artificial intelligence and Deep Learning. healthcare, government & defense, and academe,
gather, organize, collaborate, and convey data. This is
often a large rising within the world IT market and
holds an awfully strong growth potential.

This currently innovative technology permits


devices or machines to assemble, predict, resolve
complexity, scrutinize, and decide future outcomes
and be sort of a sophisticated and intelligent system,
like human brain working structure. The system
captures all dimensions of sensory vertex and
Fig. 6 Surveillance with Deep learning and AI different physiological and psychological changes in
Surveillance and security are the major concern of subjects/people beneath observation. Deep Learning
daily life, now a day, and every field contain video mechanism relies on the component of the human
camera, with the intelligence in which when any brain that plays a large role in higher psychological
unwanted event happened then camera or mental process by examining the behavior of a
automatically operate and collect the large amount of human’s thinking and taking choices and decision
data continuously. Internally this complex data are on the concept of it slow series analysis [22].The
Deep Learning systems incorporate technologies as

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


550
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

well as contact primarily based and non-contact


technologies and captures human brain calculation
through software system ways like neural analytics,
gesture recognition, speech recognition, natural
language processing and different physiological
recognition features, to capture, archive, and discuss
with existing knowledge [25].

3.2 Comparison between The Gartner Hype Cycle

"The Gartner hype Cycle for emerging Skills” as


shown in figure 4 is that the broadest hype Cycle,
which incorporates that includes technologies that Fig 7. Gartner Hype Cycle 2015
are the stress of attention due to significantly nice
levels of interest, and people that Gartner believes
have the potential for important impression. Betsy
Burton distinguished analyst at Gartner says, "We
encourage CIOs and alternative IT front-runners to
devote time and energy focused on innovation,
instead of simply business progression, whereas
additionally gaining inspiration by scanning on the
far side the bounds of their trade."

“Deep Learning” isn’t on the graph for the first time


since 2015”, on that time people do not motivated
about the Deep learning technology(see figure 7). In
the 2018 Garter hype Cycle for rising Technologies Fig 8. Gartner Hype Cycle 2018
(see figure 8) the location of “Deep Learning” within
the transition section is on the “Peak” whereby a lot IV. CONCLUSION
of analysis focus is driving expectations from the
sector on top of ever before and within the next Aesthetic assessment of videos and the images is a
decade it's expected to peak because the technology challenging field. Many researchers are working on
matures and analysis outcomes are reported from the classification techniques and they find that
varied analysis teams [32]. there is no foolproof technique for the classification
and prediction [2]. The field of aesthetic assessment
is subjective [17] and thus deep learning provides a
multi-level learning method to get accurate
classification [22]. Here deep learning provides the
automatic aesthetic assessment for videos and
images [10]. The image features provided for the
video motion feature are extraction, which can
provide to be broadly applicable for video aesthetic

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


551
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

assessments. The frame conversion and the shots Detection Method for Art Images”, IEEE
detection is help to understand the characteristics Photonics Journal ( Volume: 9, Issue: 6, Dec.
of video motion to good shot quality of videos. In 2017
the future, the feature extraction and the [7]. Roberto Gallea, Edoardo Ardizzone, and
classification both using the deep learning to take Roberto Pirrone, “Automatic Aesthetic Photo
more accurate result on aesthetics. Many editor are Composition”, Image Analysis and Processing –
working on the animation work in the videos, using ICIAP 2013: 17th International Conference,
deep learning we can modify and get the better Naples, Italy, September 9, 2013.
result over that videos. The video transmission is [8]. Stephen Groening, “Introduction: The
based on the video size, so we can use the deep Aesthetics of Online Videos”, The Aesthetics of
learning base aesthetic video transmission with less Online Videos, June 2016,
size and high aesthetic assessments. DOI: http://dx.doi.org/10.3998/fc.13761232.004
0.201
[9]. christos tzelepis, eftichia mavridaki, vasileios
V. REFERENCES
mezaris, ioannis patras, “Video aesthetic quality
assessment using kernel support vector
[1]. Dhiraj Joshi, Ritendra Datta, Elena
machine with isotropic gaussian sample
Fedorovskaya, Quang-Tuan Luong, James Z.
uncertainty (ksvm-igsu)”, IEEE Int. Conf. On
Wang, Jia Li, and Jiebo Luo, “Aesthetics and
Image Processing (Icip 2016), Phoenix, Az, Usa.
Emotions in Images”, IEEE Signal Processing
[10]. Jun-Ho Choi, Jong-Seok Lee, “Automated
Magazine, Volume: 28, Issue: 5, Sept. 2011
Video Editing for Aesthetic Quality
[2]. Xin Lu, Zhe Lin, Hailin Jin, Jianchao Yang, and
Improvement”, 23rd ACM international
James. Z. Wang, “Rating Image Aesthetics using
conference on Multimedia, Brisbane, Australia
Deep Learning ”, IEEE Transactions on
— October 26 - 30, 2015
Multimedia, Volume: 17, Issue: 11, Nov. 2015
[11]. Anselm Brachmann, Christoph Redies,
[3]. Yubin Deng, Chen Change Loy, and Xiaoou
“Computational and Experimental Approaches
Tang, “Image Aesthetic Assessment”, IEEE
to Visual Aesthetics”, Front. Comput.
Signal Processing Magazine, Digital Object
Neurosci., 14 November 2017 |
Identifier 10.1109/MSP.2017.2696576, 11 July
https://doi.org/10.3389/fncom.2017.00102.
2017.
[12]. Aarushi Gaur, “Ranking Images Based on
[4]. Yubin Deng, Chen Change Loy, and Xiaoou
Aesthetic Qualities”, Submitted for the Degree
Tang, “Image Aesthetic Assessment: An
of Doctor of Philosophy from the University of
Experimental Survey”, arXiv:1610.00838v2
Surrey, April 2015.
[cs.CV] 20 Apr 2017.
[13]. Loris Nanni, Stefano Ghidoni, Sheryl Brahna,
[5]. Ritendra Datta Dhiraj Joshi Jia Li James Z.
“Handcrafted vs. non-handcrafted features for
Wang, “Studying Aesthetics in Photographic
computer vision classification”, Elsevier Journal
Images Using a Computational Approach”,
on Pattern Recognition 71 (2017) 158–172,
Computer Science, vol 3953. Springer, Berlin,
2017.
Heidelberg, ECCV 2006
[14]. Hsin-Ho Yeh, Chun-Yu Yang, Ming-Sui Lee,
[6]. Chongke Bi ,Ye Yuan, Ronghui Zhang , Yiqing
and Chu-Song Chen, “Video Aesthetic Quality
Xiang,Yuehuan Wang, and Jiawan Zhang, “A
Assessment by Temporal Integration of Photo-
Dynamic Mode Decomposition Based Edge

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


552
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

and Motion-Based Features”, IEEE Transactions classification,” in IEEE Conference on


on Multimedia, Vol. 15, No. 8, December 2013. Computer Vision and Pattern Recognition
[15]. Katharina Schwarz, Patrick Wieschollek, (CVPR), pp. 3642–3649, 2012.
Hendrik P. A. Lensch, “Will People Like Your [23]. Y. Sun, X. Wang, and X. Tang, “Hybrid deep
Image? Learning the Aesthetic Space”, learning for face verification,” in The IEEE
arXiv:1611.05203v2 [cs.CV] 4 Dec 2017 International Conference on Computer Vision
[16]. Yanran Wang, Qi Dai, Rui Feng, Yu-Gang (ICCV), 2013.
Jiang, “Beauty is Here: Evaluating Aesthetics in [24]. Jingwei Xu, Li Song, Rong Xie, "Shot boundary
Videos Using Multimodal Features and Free detection using convolutional neural
Training Data”, ACM international conference networks", Visual Communications and Image
on Multimedia, Pages 369-372, Barcelona, Processing (VCIP) 2016, pp. 1-4, 2016.
Spain — October 21 - 25, 2013 [25]. Simone Bianco, Luigi Celona, Paolo
[17]. Yuzhen Niu, Feng Liu, “What Makes a Napoletano, and Raimondo Schettini,
Professional Video? A Computational “Predicting Image Aesthetics with Deep
Aesthetics Approach”, IEEE Transactions on Learning” ,Springer International Journal on
Circuits and Systems for Video Technology, Advanced Concepts for Intelligent Vision
Volume: 22, Issue: 7, July 2012 Systems, 21 October 2016, volume: 10016.
[18]. Subhabrata Bhattacharya, Behnaz Springer, Cham.
Nojavanasghari, Tao Chen, “Towards a [26]. C. D. Manning and H. Schütze, “Foundations of
Comprehensive Computational Model for Natural Language Processing,” Reading, pp.
Aesthetic Assessment of Videos”, ACM 978-1- 678, 2000.
4503-2404-5, 2013, Barcelona, Spain, [27]. P. K. Abhishek Kumar Chauhan, “Moving
http://dx.doi.org/10.1145/2502081.2508119. Object Tracking using Gaussian Mixture Model
[19]. Chun-Yu Yang, Hsin-Ho Yeh, and Chu-Song and Optical Flow,” vol. 3, no. 4, pp. 243–246,
Chen, “Video Aesthetic Quality Assessment By 2013.
Combining Semantically Independent And [28]. D. K. Srivastava, L. Bhambhu, and B. Cet, “Data
Dependent Features”, Acoustics, Speech and classification using support vector machine,”
Signal Processing (ICASSP), IEEE International Journal of Theoretical and Applied Information
Conference, 2011, Prague, Czech Republic. Technology, 2010.
[20]. Yanxiang Chen, Yuxing Hu, Luming Zhang, [29]. M. S. and W. P, “Research Paper on Basic of
Ping Li, Chao Zhang, “Engineering Deep Artificial Neural Network,” International
Representations for Modeling Aesthetic Journal of Recent Innovation Trends Computer
Perception”, IEEE Transactions on Communication, vol. 2, no. 1, pp. 96–100, 2014.
Cybernetics, Volume: PP, Issue: 99 , 11 [30]. P. Sharma and A. P. R. Bhartiya,
December 2017. “Implementation of Decision Tree Algorithm to
[21]. A. Krizhevsky, I. Sutskever, and G. E. Hinton, Analysis the Performance,” Int. J. Adv. Res.
“Imagenet classification with deep Computer Communication. Eng., vol. 1, no. 10,
convolutional neural networks,” in Advances in pp. 861–864, 2012.
Neural Information Processing Systems (NIPS), [31]. F. Park, “The Hierarchical Hidden Markov
pp. 1106–1114, 2012. Model : Analysis and Applications,” vol. 62, pp.
[22]. D. Ciresan, U. Meier, and J. Schmidhuber, 41–62, 1998.
“Multi-column deep neural networks for image [32]. Gartner Hype Cycle:

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


553
Manish Bendale et al Int J Sci Res CSE & IT. January-February-2019 ; 5(1) : 546-554

[33]. https://medium.com/machine-learning-in-
practice/deep-learnings-permanent-peak-on-
gartner-s-hype-cycle-96157a1736e
[34]. NadineKrefetz , streaming media
https://www.streamingmedia.com/Authors/683
7-Nadine-Krefetz.htm

Cite this article as :

Manish Bendale, Madhura V. Phatak, Dr. Nitin N.


Pise, "Analysis and Importance of Deep Learning for
Video Aesthetic Assessments", International Journal
of Scientific Research in Computer Science,
Engineering and Information Technology
(IJSRCSEIT), ISSN : 2456-3307, Volume 5 Issue 1, pp.
546-554, January-February 2019. Available at doi :
https://doi.org/10.32628/CSEIT1951100
Journal URL : http://ijsrcseit.com/CSEIT1951100

Volume 5, Issue 1, January-February-2019 | http:// ijsrcseit.com


554

Das könnte Ihnen auch gefallen