Beruflich Dokumente
Kultur Dokumente
HTTPS://SITES.GOOGLE.COM/SITE/JOURNALOFCOMPUTING/
WWW.JOURNALOFCOMPUTING.ORG 68
Abstract— Opinion mining is a recent sub discipline of information retrieval which is not about the topic of a document, but with the
opinion it expresses. There are number of research activities that focused on determination of the semantic orientation at the word
level, since seed words play an important role as sentiment clues in indicating the sentiment of a sentence or a document. Web
users document their opinion in the form of reviews or opinionated texts at different opinionated sites, shopping sites, personal pag-
es etc., to express and share their opinion with other web users. The opinion expressed by web users may be on diverse topics
such as politics, sports, products, movies etc. These opinions will be very useful to others such as, leaders of political parties, selec-
tion committees of various sports, business analysts and other stake holders of products, directors and producers of movies as well
as to the other concerned web users. In this paper, we present an unsupervised sentence and document based approaches for find-
ing opinion of web users from opinionated texts and to classify web user’s opinion into positive or negative. Our approach performs
better when subjected to different data sets of nearly four hundred and forty five opinionated texts. The results of our approach are
good compared to other published results.
Index Terms— Opinion Detection, Product Analysis, Phrase Detection, Sentiment Analysis.
—————————— ——————————
1 INTRODUCTION
W
ith the rapid expansion of e-commerce, more and tions etc., to make them available globally and also gain
more products are sold on the Web, and more knowledge globally. The individual user, on the other
and more people are also buying products online. end, is provided with an opportunity to gain knowledge
In order to enhance customer satisfaction and shopping and to share knowledge. The web is the source of many
experience, it has become a common practice for online research activities and one interesting area is to mine user
merchants to enable their customers to review or to ex- opinion from web on diverse topic. The study of opinions
press opinions on the products that they have purchased. is useful for both producers and consumers of the topic.
With more and more common users becoming comforta- The producers can be manufacturers of digital products,
ble with the Web, an increasing number of people are automobile manufactures, movie producers, editor of
writing reviews. As a result, the number of reviews that a news article etc., very much interested to find opinion of a
product receives grows rapidly. Some popular products user. The consumers are individual user’s who document
can get hundreds of reviews at some large merchant sites. their opinion and want to share it with others about the
Furthermore, many reviews are long and have only a few topic [14].
sentences containing opinions on the product. This makes
it hard for a potential customer to read them to make an Many readers of online reviews say that these reviews or
informed decision on whether to purchase the product. If opinions influence their purchasing decision [4]. Today
customer reads only few reviews, he/she may get a bi- people of all ages and from all over the world use web for
ased view. The large number of reviews also makes it collecting opinions. There are many sites which allow
hard for product manufacturers to keep track of customer user’s to express their opinion such as Epinions, Amazon,
opinions of their products. For a product manufacturer, CNet[5][6]and [7] etc., some of these sites are supervised
there are additional difficulties because many merchant manually, therefore all opinions expressed by user’s may
sites may sell its products, and the manufacturer may not be published. The opinionated sites seem to often ex-
(almost always) produce many kinds of products.[2] hibit highly skewed rating distributions with a particular
bias towards positive reviews [8][9][10]and [11]. There are
The development of web and its related technologies cases in opinionated sites where the user’s have selected a
have fueled the popularity of the web with all sections of low rating while their opinion indicates a positive opinion
society. The web has been rightfully used by govern- [12].
ments, business houses, industries, educational institu-
Before the Web, when an individual needs to make a de-
————————————————
cision, he/she typically asks for opinions from friends
Anil Kumar.K.M is with Sri Jayachamarajendra College of Engineering, and families. When an organization needs to find opi-
Mysore, Karnataka, India.
Anil Kumar.P is with Sri Jayachamarajendra College of Engineering, nions of the general public about its products and servic-
Mysore,Karnataka, India. es, it conducts surveys and focused groups. With the
Suresha is with Dept of studies in Computer Science, University of My- Web, especially with the explosive growth of the user
sore.
generated content on the Web, the world has changed. Wang and Araki [35] proposed a variation of the Seman-
One can post reviews of products at merchant sites and tic Orientation-PMI algorithm for Japanese for mining
express views on almost anything in Internet forums, dis- opinion in weblogs. They applied Turney method to Jap-
cussion groups, and blogs, which are collectively called anese webpage and found results slanting heavily to-
the user generated content. Now if one wants to buy a wards positive opinion. They proposed balancing factor
product, it is no longer necessary to ask one’s friends and and neutral expression detection method and report a
families because there are plentiful of product reviews on well balanced result.
the Web which give the opinions of the existing users of
the product. For a company, it may no longer need to Kamps et al [16] have focused on the use of lexical rela-
conduct surveys, to organize focused groups or to employ tions, defined in Word Net. They defined a graph on the
external consultants in order to find consumer opinions adjectives contained in the intersection between the Tur-
or sentiments about its products and those of its competi- ney’s seed set and Word Net, adding a link between two
tors. Finding opinion sources and monitoring them on the adjectives whenever WordNet indicate the presence of a
Web, however, can still be a formidable task because a synonymy relation between them. The author’s defined a
large number of diverse sources exist on the Web and distance measure d (t1, t2) between terms t1 and t2, which
each source also contains a huge volume of information. amounts to the length of the shortest path that connects t1
and t2. The orientation of a term is then determined by its
In this paper we focus on detecting opinions expressed by relative distance from the seed terms good and bad.
web users using both Sentence and Document based ap-
proaches and provide a summary of users opinion. The Esuli and Sebastiani [17] proposed semi-supervised learn-
remainder of this paper is organized as follows: In Section ing method started from expanding an initial seed set
2 we give a brief description of related work. Then, in based on Turney and Littman’s seed set [18], by using
Section 3, we discuss our methodology. In Section 4, the WordNet. Their basic assumption is terms with similar
experimental results are discussed. We discuss applica- orientation tend to have similar glosses. They determined
tion in section 5. Conclusion is discussed in Section 6. the expanded seed term’s semantic orientation through
gloss classification by statistical technique.
2 RELATED WORK
Morinaga et al [19] presents a new frame work where a
Opinion mining is a recent sub discipline of information user inputs products names and the system collects
retrieval which is not about the topic of a document, but people’s opinion and attaches three labels to each, such as
with the opinion it expresses [28][29]. In literature opi- the name of the product, positive or negative nature of
nion mining is also known as sentiment analysis[30], sen- opinion and opinion likeliness. These labeled opinions are
timent classification [31], opinion extraction[32], affective then put into an opinion database. A user specifies a tar-
classification[29] and affective rating[33]. It has emerged get category for analysis using the value of labels. The
in the last few years as a research area, largely driven by system conducts text mining to extract statistically mea-
interests in developing applications such as mining opi- ningful information corresponding to the specified target.
nions in online corpora, or customer relationship man-
agement e.g., customer’s review analysis[29]. Pang et al [20] adopted a statistical technique-based ap-
proach, using supervised machine learning with words
There are number of research activities that focused on and n-grams as features to predict orientation at the doc-
determination of the semantic orientation at the word ument level.
level, since seed words play an important role as senti-
ment clues in indicating the sentiment of a sentence or a Kim and Hovy [21] presents orientation detection system
document. Hatzivassiloglou and McKeown [34] have at- that assigns to each term, a positive score and a negative
tempted to predict semantic orientation of adjectives by score, the terms may have both a positive and a negative
analyzing pairs of adjectives (i.e., adjective pair is adjec- correlation, with different degrees, and some terms may
tives conjoined by and, or, but, either-or, neither-nor) ex- carry a stronger positive or negative orientation than oth-
tracted from a large unlabelled document set. ers. Their system starts from a set of positive and negative
seed terms, and expands the positive and negative seed
Turney has obtained remarkable results on the sentiment set by adding to it the synonyms of positive and negative
classification of terms by considering the algebraic sum of seed terms and the antonyms of negative and positive
the orientations of terms as representative of the orienta- seed terms. The system classifies then a target term t into
tion of the document. Turney and Littman have boot- either positive or negative by means of two alternative
strapped from a seed set, containing seven positive and learning-free methods based on the probabilities that
seven negative words, and determined semantic orienta- synonyms of t also appear in the respective expanded
tion according to Point wise Mutual Information- seed sets.
Information Retrieval (PMI-IR) method.
JOURNAL OF COMPUTING, VOLUME 2, ISSUE 8, AUGUST 2010, ISSN 2151-9617
HTTPS://SITES.GOOGLE.COM/SITE/JOURNALOFCOMPUTING/
WWW.JOURNALOFCOMPUTING.ORG 70
Popescu and Etzioni [22] introduced OPINE. It is an un- features are extracted from noun or noun phrases by the
supervised information extraction system that outputs set association miner. They use adjectives as opinion words
of features which is accompanied by a list of opinions that and assign prior polarity of these by WordNet exploring
are ranked based on strength for a given product and method. The polarity of an opinion expression which is a
corresponding reviews. To find features, it first extracts sentence containing one or more feature terms and one or
the noun phrases from reviews and retains those with more opinion words is assigned a dominant orientation.
frequency greater than an experimentally set threshold. The extracted features are stored in a database in the form
OPINE’s Feature Assessor evaluates each noun phrase by of feature, number of positive expression and number of
computing the PMI scores between the phrase and mero- negative expression. The system shows the results in a
nymy discriminators associated with the product. It ex- graph format showing opinion of the product feature by
tracts opinion phrases, which are adjective, noun, verb or feature [27].
adverb phrases representing customer opinions and uses
relaxation labeling, unsupervised classification technique, Previous works on mining opinions can be divided into
for finding the semantic orientation of words. two directions: sentiment classification and sentiment
related information extraction. The former is a task of
Pang and Lee [23] report of work in progress on using identifying positive and negative sentiments from a text
simple statistics in an unsupervised fashion to re-rank which can be a passage, a sentence, a phrase and even a
search engine results for a review oriented query. They word. The latter focuses on extracting the elements com-
report that their proposed technique performs compara- posing a sentiment text. The elements include source of
bly to methods that rely on sophisticated pre-encoded opinions who expresses an opinion. Some researchers
linguistic knowledge. refer this information extraction task as opinion extraction
or opinion mining. Comparing with the former one, opi-
ReviewSeer is a tool that automates the work done by
nion mining usually produces richer information [24].
aggregation sites. It uses various methods such as meta-
data and statistical substitutions, linguistic substitutions,
language based modifications, n-gram and proximity for 3 METHODOLOGY
feature extraction. Naive Bayes classifier is used with pos- In this paper, we present unsupervised approaches to
itive and negative review sets for assigning a score to the detect opinion of web users from product reviews and to
extracted feature terms. The classifier performed well for classify web user’s opinion into positive or negative.
reviews collected from CNet and Amazon for training People use phrases to express their opinions to a consi-
and testing. The classifier did not perform well for web derable extent, based on these phrases, the opinions are
pages crawled from the result of a search engine. It dis- detected using two different approaches (viz. Sentence
plays attributes and score of the attribute along with re- based and Document based) and classify them as positive
view sentences [24]. or negative. The system will detect the opinion and con-
vey whether the opinion of public about the entity is posi-
WebFountain uses beginning definite Base Noun Phrase tive or negative. The system is tested against nearly four
(bBNP) heuristic for extracting product features. It ex- hundred and forty five files and the accuracy of the result
tracts the base noun phrases at the beginning of the sen- is encouraging and good.
tences followed by a verb phrases. To assign sentiments
to the features, reviews are parsed and traversed with 3.1 Sentence based Approach
two linguistic resources namely the sentiment lexicon and
the sentiment pattern database. The sentiment lexicon The Sentence based approach contains five steps. In the
defines the polarity of terms and sentiment pattern data- first step, the file containing product reviews is split into
base defines sentiment extraction patterns for a sentence sentences and then the review text is tagged using Monty
predicate [25]. tagger. The tagged text is used for pattern detection for
user defined patterns. The phrases are extracted based on
Red Opal is a tool that assumes online shoppers are high-
the patterns and each phrase is checked to obtain the
ly task driven and have certain goal in mind and that they
polarity. Based on the window size, the polarity of each
are looking for product with features that are consistent
with that goal. This system enables users to find products phrase is calculated authenticating the presence in seed
based on features. It scores each product based on fea- lists and total score of the review is calculated. The flow
tures from the customer reviews. It uses frequent nouns chart for the sentence based approach is shown in Fig.1.
and noun phrases for feature extraction and user ratings
are used to compute product score for features mentioned
in reviews. The results are shown in descending order for
each feature along with the URL [26].
Opinion observer is a sentiment analysis system for ana-
lyzing and comparing opinions on the web. The product
JOURNAL OF COMPUTING, VOLUME 2, ISSUE 8, AUGUST 2010, ISSN 2151-9617
HTTPS://SITES.GOOGLE.COM/SITE/JOURNALOFCOMPUTING/
WWW.JOURNALOFCOMPUTING.ORG 71
Split review into sentences Opinion Detection – Total score for the review file is
calculated by adding the scores obtained by pattern
detection step and check for polarity step for each
Apply Monty tagger sentence. The total score of the review is calculated as
the sum of scores of all the sentences in the review
Phrase Detection Check for file. We use here 0 as threshold. If the score of the re-
Polarity
view is greater than threshold T, the opinion is consi-
dered as positive and if the score is less than thre-
shold, then the opinion is considered as negative.
Opinion Detection
Fig.1 Flow chart of design process for sentence approach 3.2 Document based Approach
Check for Polarity – This step deals with checking Result-1: The output of the file for the input
the polarity of the features extracted in the previous (inp/04_nokia6610.txt) is written into a new file. The Score
step. The output from the features extraction is taken obtained is greater than threshold (T > 0) and hence opi-
as input. The phrases extracted are analyzed for their nion is classified as Positive.
polarity. The extracted phrases contain adjectives or
adverbs. The extracted phrases are searched for their Generated Report - Sentence Approach
presence in the positive list and negative list. If the -------------------
Inp/04_nokia6610.txt Score=6 Opinion : Positive
feature is present in the positive list, a score of +1 will
be assigned. Similarly, if the feature is present in the
negative list, then a score of -1 will be assigned. This Example – 2
process is applied for all the features extracted from
the document. Result -2: The output of the file for the input
(inp/16_Canon.txt ) is written into a new file as shown below.
Opinion Detection – Once the polarity of the doc- The Score obtained is greater than threshold (T > 0) and
uments is known, the final step deals with the analy- hence opinion is classified as Positive.
sis of the Statistics. These statistics are used to deter-
mine the quality of the product. First the documents Generated Report - Sentence Approach
which feel positive are passed through entire process -------------------
and result of that is calculated in terms of percentage
Inp/16_Canon.txt Score=5 Opinion : Positive
later negative set of documents is passed through the
process. The statistics of the entire document is taken.
The score of a document review is calculated by add-
Experiment - 2: Feature extraction using Document
ing the positive score and negative score. We use
based Approach
here 0 as threshold T. If the score of the review is
greater than threshold T, the opinion is considered as
Example – 1
positive and if the score is less than threshold, then
the opinion is considered as negative.
In this experiment, the tagged review file is given as in-
put. The output is the score of the review using Docu-
ment based Approach.
The product with a higher percentage of positive
results yields in good product and the products with Result -1: The output of the file for the input
higher percentage of negative results will end up with a (inp/04_nokia6610.txt) is written into a new file. The Score
bad product. The reviews help the manufacturers in obtained is greater than threshold (T > 0) and hence opi-
knowing the positive and negative features of their nion is classified as Positive.
product and thus further help them in developing better
Inp/04_nokia6610.txt – Document Approach
products. There is a huge need in the industry for such
Inp/04_nokia6610.txt Score=6 Opinion : Positive
services because every company wants to know how
consumers perceive their products and services and those
of their competitors.
JOURNAL OF COMPUTING, VOLUME 2, ISSUE 8, AUGUST 2010, ISSN 2151-9617
HTTPS://SITES.GOOGLE.COM/SITE/JOURNALOFCOMPUTING/
WWW.JOURNALOFCOMPUTING.ORG 73
Example – 2
TABLE: 3 CLASSIFICATION OF MULTIPLE REVIEWS USING
Result – 2: The output of the file for the input DOCUMENT BASED APPROACH ON DATA SET 1
(inp/16_Canon.txt) is written into a new file as shown below.
Results obtained using Document based Approach
The Score obtained is less than threshold (T < 0) and
hence opinion is classified as Negative. Sl Total Positive Negative Avg
Product
N No.of Accuracy Accuracy Score
inp/16_Canon.txt – Document Approach Name
o Files (%) (%) (%)
TABLE: 2 CLASSIFICATION OF MULTIPLE REVIEWS USING True positive represent number of opinionated texts clas-
SENTENCE BASED APPROACH ON DATA SET 1 sified correctly as positive, similarly true negative
represent number of opinionated texts classified correctly
Results obtained using Sentence based Approach
as negatives. The result obtained is good considering that
Total Positive Negative Avg we use only adjectives to find opinionated phrases. We
Sl Product
No.of Accuracy Accura- Score obtain an accuracy of 69% and 61% with Sentence based
No Name
Files (%) cy (%) (%)
approach for positive and negative data sets. An accuracy
1 NikonCoolPix 33 93.55 100.00 93.94 of 63% and 63% were obtained with Document based
approach for both positive and negative datasets.
2 Nokia6610 34 96.88 50.00 94.12
REFERENCES [12] Luis Cabral and Ali Hortacsu, “The dynamics of sel-
ler reputation: Theory and evidence from eBay”,
[1] Bo Pang and Lillian Lee (2008) "Opinion Mining and working paper, downloaded version revised in
Sentiment Analysis", Foundations and Trends® in In- March, 2006. URL http://pages.stern.nyu.
formation Retrieval: Vol. 2: No 1–2, pp 1-135.
[13] http://www.cs.uic.edu/~liub/FBS/opinion-
[2] Minqing Hu and Bing Liu, “Mining and Summariz- mining.pdf
ing Customer Reviews”, Proceedings of KDD’04, Au-
gust 22-25, 2004, Seattle, Washington, USA. [14] Anil Kumar K.M and Suresha, “Threshold Based Ap-
proach for Mining Users’ Opinion on Products from
[3] YuanbinWu, Qi Zhang, Xuanjing Huang, LideWu, Web” IICAI-2009
“Phrase Dependency Parsing for Opinion Mining”,
Proceedings of the 2009 Conference on Empirical Me- [15] Anil, Kumar, K.M., Suresha: Identifying Subjective
thods in Natural Language Processing, pages 1533– Phrases from Opinionated Texts Using Sentiment
1541, Singapore, 6-7 August 2009 Product Lexicon. International Journal of Advanced
Engineering & Applications. 2, 63-271 (2010)
[4] ComScore/the Kelsey group, “Online consumer-
generated reviews have significant impact on offline [16] Jaap Kamps, Maarten Marx, Robert J. Mokken and
purchase behavior”, Press Release, November 2007. Maarten De Rijke, “Using wordnet to measure se-
www.comscore.com/press/release.asp?press=1928. mantic orientation of adjectives”, Proceedings of 4th
International Conference on Language Resources and
[5] Peter D. Turney, Thumbs up or thumbs down? Se- Evaluation, pp. 1115-1118, Lisbon, Portugal, 2004.
mantic orientation applied to unsupervised classifica-
tion of reviews”, Proceedings of 40th Annual Meeting [17] Andrea Esuli and Fabrizio Sebastiani, “Determining
of the Association for Computational Linguistics, pp. the semantic orientation of terms through gloss clas-
417-424, Philadelphia, 2002. sification”, Proceedings of 14th ACM International
Conference on Information and Knowledge Man-
[6] Kushal Dave, Steve Lawrence and David M. Pennock, agement, pp. 617-624, Bremen, Germany, 2005.
“Mining the peanut gallery: Opinion extraction and
semantic classification of product reviews”, Proceed- [18] Peter D. Turney and Michael L. Littman, “Measuring
ings of 12th International World Wide Web Confe- praise and criticism: Inference of semantic orientation
rence, pp. 519-528, Budapest, Hungary, 2003. from association”, ACM Transactions on Information
Systems, pp. 315- 346, 2003.
[7] Bing Liu, Minqing Hu,“Mining and summarizing
customer reviews”, Proceedings of 10th ACM [19] Satoshi Morinaga, Kenji Yamanishi, Kenji Tateishi
SIGKDD International Conference on Knowledge and Toshikazu Fukushima, “Mining Product Reputa-
Discovery and Data Mining, pp. 168- 177, Washing- tions on the Web”, Proceedings of 8th ACM SIGKDD
ton, 2004. international conference on Knowledge discovery
and data mining SIGKDD’02, pp. 341-349, Edmonton,
[8] Judith A. Chevalier and Dina Mayzlin, “The effect of Alberto, Canada, 2002.
word of mouth on sales: Online book reviews”, Jour-
nal of Marketing Research, pp. 345-354, 2006. [20] Bo Pang, Lillian Lee and Shivakumar Vaithyanathan,
“Thumbs up? sentiment classification using machine
[9] Michael Dewally, “Internet investment advice: In- learning techniques”, Proceedings of 7th Conference
vesting with a rock of salt”, Financial Analysts Jour- on Empirical Methods in Natural Language
nal, pp. 65-77, 2003. Processing, pp. 79-86, Philadelphia, US, 2002.
[10] Daniel Houser and JohnWooders, “Reputation in [21] Soo-Min Kim and Eduard Hovy, “Determining the
auctions: Theory, and evidence from eBay”, Journal sentiment of opinions”, Proceedings of 20th Interna-
of Economics and Management Strategy, pp. 252-369, tional Conference on Computational Linguistics, pp.
2006. 1367-1373, Geneva, Switzerland, 2004.
[11] Nan Hu, Paul A. Pavlou and Jennifer Zhang, “Can [22] Popescu and Etzioni, “Extracting product features
online reviews reveal a products true quality?: empir- and opinions from reviews”, Proceedings of Confe-
ical findings and analytical modeling of online word- rence on Human Language Technology and Empiri-
of-mouth communication”, Proceedings of Electronic cal Methods in Natural Language Processing, pp.
Commerce, pp. 324-330, New York, USA, 2006. 339-346, Vancouver, British Columbia, Canada, 2005.
JOURNAL OF COMPUTING, VOLUME 2, ISSUE 8, AUGUST 2010, ISSN 2151-9617
HTTPS://SITES.GOOGLE.COM/SITE/JOURNALOFCOMPUTING/
WWW.JOURNALOFCOMPUTING.ORG 76
[23] Bo Pang and Lillian Lee, “Using very simple statistics logs, California, US, 2006.
for review search: An Exploration”, Coling 2008:
Companion volume Posters and Demonstrations, pp. [34] Vasileios Hatzivassiloglou and Kathleen R.
75-78, Manchester, UK, 2008. McKeown, “Predicting the semantic orientation of
adjectives”, Proceedings of 35th Annual Meeting of
[24] YuanbinWu, Qi Zhang, Xuanjing Huang, LideWu, the Association for Computational Linguistics, pp.
“Phrase Dependency Parsing for Opinion Mining”, 174-181, Madrid, Spain, 1997.
Proceedings of the 2009 Conference on Empirical Me-
thods in Natural Language Processing, pages 1533– [35] Wang and Araki, “Modifying SO-PMI for Japanese-
1541, Singapore, 6-7 August 2009 Weblog Opinion Mining by Using a Balancing Factor
and Detecting Neutral Expressions”, Proceedings of
[25] Yi and Niblack, “Sentiment Mining in Web Foun- NAACL HLT 2007, pp. 189-192, Rochester, New
tain”, Proceedings of 21st international Conference York, US, 2007
on Data Engineering, pp. 1073-1083, Washington DC,
2005. [36] Alistair, Kennedy, Diana, Inkpen: Sentiment Classifi-
cation of Movie and Product Reviews Using Contex-
[26] Christopher Scaffidi, Kevin Bierhoff, Eric Chang, tual Valence Shifters. In: Proceedings of FINEXIN
Mikhael Felker, Herman Ng and Chun Jin, “Red 2005, Workshop on the Analysis of Informal and
Opal: product-feature scoring from reviews”, Pro- Formal Information Exchange during Negotiations.
ceedings of 8th ACM Conference on Electronic Canada, 2005.
Commerce, pp. 182-191, New York, 2007.
Anil Kumar.K.M is a faculty member with Department of Computer
Science and Engineering, Sri Jayachamarajendra College of Engi-
[27] Bing Liu, Minqing Hu and Junsheng Cheng, “Opi- neering, Mysore. He is also a Research Scholar, Department of
nion observer: analyzing and comparing opinions on Studies in Computer Science, University of Mysore. He received his
theWeb”, Proceedings of 14th international Confe- B.E. degree in 1999 from University of Mysore and M.Tech degree in
2006 from Visvesvaraya Technological University, Karnataka. He is
rence onWorldWideWeb, pp. 342-351, Chiba, Japan, working towards doctoral work under the supervision of Dr. Suresha.
2005.
Anil Kumar.P is a Post Graduate scholar in the Department of
Computer Science and Engineering, Sri Jayachamarajendra College
[28] Andrea Esuli and Fabrizio Sebastiani, “Determining of Engineering, Mysore, Karnataka, India. He received his Bachelor’s
term subjectivity and term orientation for opinipon degree in Computer Science and Engineering in 2002 from Visves-
mining”, Proceedings of 11th Conference of the Eu- varaya Technological University, Karnataka, India.
ropean Chapter of the Association for Computational
Dr. Suresha is presently working as Reader, Department of Studies
Linguistics, Trento, Italy, 2006. in Computer Science, University of Mysore. He received his B.Sc.
degree in 1987 from University of Mysore, M.Sc degree in 1989 from
[29] Youngho Kim and Sung-Hyon Myaeng, “Opinion University of Mysore and M. Phil degree in 1991 from DAVV, Indore.
He received M.Tech degree from Indian Institute of Technology, Kha-
Analysis based on Lexical Clues and their Expan- ragpur in 1996 and Ph. D. from prestigious Indian Institute of
sion”, Proceedings of NTCIR-6 Workshop Meeting, Science, Bangalore in 2007. He was awarded second prize in IRISS-
pp. 308-315, Tokyo, Japan, 2007. 2002 competition, which is an all India level research student compe-
tition called “Inter Research Institute Student Seminar”. He has a
teaching experience of 20 years at post graduate level, has 12 publi-
[30] Bo Pang and Lillian Lee, “A sentimental education: cations to his credit and currently supervising four research scholars
Sentiment analysis using subjectivity summarization towards their doctoral work.
based on minimum cuts”, Proceedings of 42nd Meet-
ing of the Association for Computational Linguistics,
pp. 271-278, Barcelona, Spain, 2004.
[31] Hugo, “MontyLingua: An end-to-end natural lan-
guage processor with common sense”,2003.