Sie sind auf Seite 1von 9

QUESTION:

Standard Chartered Bank is a successful international company with 120 branches in 70 countries
around the world. Each branch has its own set of databases. The company has a data warehouse as
well. The President of Standard Chartered Bank has asked you to provide an analysis of the company’s
deposits per transaction type per branch for the third quarter. From this statement one can argue that
databases, data warehouses and data mining are crucial to the success of business.
Demonstrate and justify this statement
Introduction

The aim of any business is to expand, establish in new territories and generate more profits.
However, as the enterprise grows so do the complexities involved in the planning, analysis
and decision making for the management of the business (Joseph, 2013). Technology plays a
vital role and is continuously contributing to business development by providing various tools
and techniques to enhance decision making process. Enterprise-wide decisions are best done
using what is known as Business Intelligence. Business Intelligence (BI) refers to the use of
technology to collect and effectively use information to improve business effectiveness and it
incoporates among other technologies database management, data warehousing and data
mining (Laudon & Laudon, 2018; Nadeem & Jaffri, 2004). In this paper, database
management, data mining and data warehousing shall be referred to collectively as Business
Intelligence (BI). The major aim of this paper is to demonstrate and justify to the President
of Standard Chartered Bank the need for and applicability of the BI tools listed above. Commented [fc1]: Nice introduction

The paper shall start by defining the three technologies that will be used. The first of the BI
subset to be defined is a database. Singh (2017) defines a database as an integrated collection
of logically related files (data), along with details of interpretation of the data contained
therein. In a database, data are integrated and related so that one set of software programs
provides access to all the data, minimising data redundancy, data isolation, and data
inconsistency. At the same time increasing data sharing, data security, data integrity and
data/application independence. Databases are at the beginning of the whole BI process and
act as the anchor without which the whole process up to data mining would be just become
more complex. As the data in databases grows or as an enterprise grows and establishes
other offices and branches, the use distributed databases comes into play. As a result data
warehouses become are then the next logical step. A data warehouse can be simply defined
as “a repository of collected information from multiple sources stored under a unified
schema residing at a single site” (Anadakumar & Padmavathy, 2013). Pathak, Singh, &
Oberei (2013) view a data warehouse as an enterprise-wide integrated database of
summarised, historical information, whose data comes from multiple, incompatible sources.

The final step in the BI process is data mining. Data mining is a process of extracting
previously unknown, valid and actionable information from large sets of data and then using
the information to make crucial business decisions. Kayaalp & Basarslan (2018) called data
mining as “...the task of obtaining a ‘valuable’ information among vast amounts of data”.

1
The role of Databases

Standard Chartered Bank just like any business in the 21 st century utilises a database for its
transactions. Databases belong in the category of online transaction and query processing
(OLTP) systems whose major is to perform the day to day transactions of the bank. The type
of database is a relational database management system. In order to fulfil the request by the
President for the analysis of the deposits by transaction type of transaction it would mean that
a query is run against the database system of the bank. Figure 1 provides a simplified Entity
Relationship diagram for a transaction from the bank.

Figure 1 : Entity relationship diagram for a transaction


A query connecting different tables in order to get transaction type, customer details, branch
code, time, transaction details would have to be run. An example query using Structred Query
Language is provided.

USE transaction_database;
SELECT t.transactions_id, t.transactions_type, t.balance, c.id, c.name
FROM transaction t, customers c
JOIN reports r
WHERE r.quarter = 3;

The above is a very simplified query example. However, this query will have to be run
against every database and every branch in order to come up 120 different reports which will Commented [fc2]: In other words you do not have to do it with
databases. You are data warehouse OLAP operations.

2
then need to be analysed differently and then integrated to provide a singular report which Commented [fc3]: How much time do you need?

can be presented to the President. Such a task would be cumbersome, inefficient, ineffective,
error-prone, involving teams of programmers and analysts thereby needing a big investment.
Such difficulty levels limit the effective use of complex corporate data, which usually
represents a valuable resource of an organisation. A database has intense read/write
operations and running major or complex queries increases the overhead for the database
which in turn degrades performance of the whole banking system (Bhedi, Deshpande, &
Lanjewar, 2014; Ubiparipovi & Durkovic, 2011). Commented [fc4]: You should have said the importance of the
databases

The role of Data Warehouses

Figure 2: The data warehouse process. (Source: http://www.aspiretss.com/services/data-


warehousing-solution)

This is where a data warehouse comes into play. A data warehouse is a subject-oriented,
integrated, time-variant, and non-volatile data store which provides a clear picture of current
and historical operations of organisations (Joseph, 2013). Each unit of data is non-volatile
and relevant to some moment in time (Pathak et al., 2013). Using data warehouses instead of
databases minimises the impact of reporting and complex query processing on operational
systems. Data warehouses contain denormalised and cleaned data from various sources where
queries are run. Figure 2 provides a simplified view of how data warehouses work.

In a data warehouses there are two main types of tables; “Dimension tables” which provide
the descriptive context - attributes with the “who, what, when, why, or how” and the “Fact
tables” which contain the numeric, quantitative data or the measures (Bhedi et al., 2014). The
tables should always include friendly names & descriptions is that one important entity that in

3
a business model that influences business intelligence in some way. In the case of Standard
Chartered Bank the data warehouse would have a transactions fact table containing
transaction ID, amount, balance, customer ID (as a foreign key) and a Customer dimension
table containing all the information for the customer, a time dimension table (with the year,
month, quarter, week, day), branch dimension and the country dimension tables. This
example is simplified too.

Running a query against such a data warehouse would be easy as all the bank’s transactions
will be in the data warehouse and one query will therefore bring out all the information. An
example query will be to look for all transactions in the transactions fact table against the 3rd
quarter of the current year. This is visually presented as a data cube, in an example of slicing
a data warehouse to get location (branch) and time (quarter) from a data warehouse.(Figure
3). Using query tools available in data warehousing it is easier than getting the data from 120
databases.

Figure 3: Data warehouse (data cube example) source:


http://www.tutorialspoint.com/dwh/dwh_olap.html

Having presented the ways in which data can be stored using databases and data warehouses,
the next question would be; Does the bank need both databases and data warehouses? The
answer would yes and this paper shall briefly explain why before going on to provide explain
about the third pillar of BI which is data mining.

4
By using data warehouses the bank can preserve operational data for reuse after that data has
been purged from the operational systems. Transactions in operational systems are usually
removed from the system and moved to backup periodically.

Databases are operational systems and by default and design they contain current data. The
data can be used to answer simple questions such as how many transactions happened today.
A data warehouse on the other hand is time variant (it manages the data based on time).

A database provides multiple views of data depending on the access levels and rights
accorded a user. A data warehouse on the other hand unifies the data within a common
business definition, offering one version of reality from which the management can make Commented [fc5]: Is this crucial for the business and why? This
where you should have shown how the report could be generated.
informed decisions. Data warehouse has many benefits which makes it important for
business.

Data warehouses are subject-oriented because they hinge on enterprise-specific concepts,


such as customers, products, sales, and orders. On the contrary, operational databases hinge
on many different enterprise-specific applications.

Data mining
After capturing transactions data in the databases and transforming and cleaning it for the
data warehouses, the final part of the BI will be done by using data mining. The overall goal
of the data mining process is to extract information from a data set and transform it into an
understandable structure for further use. The key characteristics of data mining are;
automatic discovery of patterns, prediction of likely outcomes, creation of actionable
information and automated discovery of previously unknown patterns and future oriented.

Figure 4 provides a pictorial view of the Data Mining process.

5
Figure 4: The data mining process (Source: Kayaalp & Basarslan, 2018)

Because of the sheer volume of data from across 70 countries and 120 branches, it is near
impossible for system analysts or data analysts to come up with interesting information (or
patterns) that will help in the decision making process Kumar (2015). Data mining can help
the bank discover new information in the following areas:

Risk Management: Managing and measurement of risk is at the core of every financial
institution. Based on the advent of advanced database and data mining technology. Other
types of risks also available in the banking and finance sector i.e., liquidity risk, operational
risk, or concentration risk (Nadeem & Jaffri, 2004).

Performance: With new digital technology, banks needs to be more smart data aware and
analyse data from various sources. Data warehouses will help getting the insights into such Commented [fc6]: I thought you said this section is now for
data mining.
volatile environment and help bank leadership to make timely decisions.

Customer Insight: Data mining provides a 360 degree view of the customer. This enables
banks to create predictive models to single segment customer’s needs (Hv & Varadarajan,
2011). Data warehouses in banks help continuously optimising its cost model while striving
to increase the number and profitability of its customers—and enhance the bottom line. Commented [fc7]: Does it make it crucial?

Compliance: Properly designed data warehouses will help assessing risk reporting and
analytics, including risk and compliance assurance, independent model validation,
quantitative assessment and back testing, risk and finance data governance and integration,
and regulatory compliance monitoring and assessment. Commented [fc8]: What about mining?

Customer Relationship Management: Chitra & Subashini (2013), state that data mining
technique also helps to identify profitable customers from non-profitable ones. Banks have
many and huge databases containing transactional and other details of its customers.

Fraud detection: In the financial industry, fraud can involve using stolen credit cards, forging
cheques, exaggerated insurance claims and most recently card cloning. Data mining helps by
providing predictive analysis and recognising customer patterns so that these instances of
fraud can be stopped and/or contained before the bank loses money. Fraud detection and
mitigation is sometimes included under Risk Management.

As competition in the banking sector is quite intense globally, banks have to fight more
creatively and proactively to expand and maintain market shares. Data Mining techniques are

6
very useful to the banking sector for better targeting and acquiring new customers, most
valuable customer retention, automatic credit approval which is used for fraud prevention,
fraud detection in real time, providing segment based products, analysis of the customers,
transaction patterns over time for better retention and relationship, risk management and
marketing. Commented [fc9]: Do all this make it crucial?

Conclusion
Rapid developments in information technology have resulted in the construction of many
business application systems in numerous areas. Within these systems, databases often play
an essential role. Data has become a critical resource in many organisations, and therefore,
efficient access to the data, sharing the data, extracting information from the data, and
making use of the information stored, has become an urgent need. As a result, there have
been many efforts on firstly integrating the various data sources (e.g. databases) scattered
across different sites to build a corporate data warehouse, and then extracting information
from the warehouse in the form of patterns and trends. This is where data mining comes in.
With the 3 processes (pillars) any enterprise will manage to successfully provide BI for the
enhancement of the company. Databases are the ones which start the chain by collecting
customer data and transactions, data warehouses incorporate data from various sources, clean
and transform it so that it provides a single point of view over a long time. After that data
mining comes in to discover hitherto unknown patterns or behaviours which will provide the
much needed BI to the organisation’s top management who will then be able to make
informed decisions. Commented [fc10]: And thus make them crucial?

20/25

7
References

Anadakumar, K., & Padmavathy, V. (2013). A Survey on Preprocessing in Text Mining.


International Journal of Advanced Research in Computer Science, 4(10), 21–30.
Aspire Tech. (n.d.). Data Warehousing Solution. Retrieved August 20, 2018, from
http://www.aspiretss.com/services/data-warehousing-solution
Bhedi, V. R., Deshpande, S. P., & Lanjewar, U. A. (2014). Data Warehouse Architecture for
Financial Institutes to Become Robust Integrated Core Financial System using BUID.
International Journal of Advanced Research in Computer and Communication
Engineeringpredominantly, 3(3), 5652–5656. Retrieved from www.ijarcce.com
Chitra K., & Subashini, B.(2013).Data Mining Techniques and its Applications in Banking
Sector.International Journal of Emerging Technology and Advanced Engineering, l3(8),
566-581
Hv, S., & Varadarajan, S. (2011). Customer Segmentation of Bank based on Data Mining –
Security Value based Heuristic Approach as a Replacement to K-means Segmentation.
International Journal of Computer Applications, 19(8), 13–18.
Joseph, M. V. (2013). Significance of Data Warehousing and Data Mining in Business
Applications. International Journal of Soft Computing and Engineering, 3(1), 329–333.
Kayaalp, F., & Basarslan, M. S. (2018). Open Source Data Mining Programs: A Case Study
on R. Düzce University Journal of Science & Technology, 6, 455–468.
Kumar, C.S.(2015). Data Mining Techniques for Banking Applications. International
Journal of Research Studies in Computer Science and Engineering (IJRSCSE), l2(4),76-
88.
Laudon, K. C., & Laudon, J. P. (2018). Management information systems (15th ed.). New
York, N.Y: Pearson.
Nadeem, M., & Jaffri, S. A. H. (2004). Application of Business Intelligence In Banks
(Pakistan). Arxiv Preprint Cs0406004, 6. Retrieved from http://arxiv.org/abs/cs/0406004
Pathak, M., Singh, S., & Oberei, S. S. (2013). Impact of Data Warehousing and Data Mining
in Decision Making. International Journal of Computer Science and Information
Technologies, 4(6), 995–999.
Singh, G. (2017). Role of Relational Database Management System in Management
Information System. International Journal of Current Engineering and Technology,
7(6), 2109–2111.
Ubiparipovi, B., & Durkovic, E. (2011). Application of Business Intelligence in the Banking
Industry. Management Information Systems,6, 23–30. Retrieved from
http://www.ef.uns.ac.rs/mis/archive-pdf/2011 - No4/MIS2011_4_4.pdf

Das könnte Ihnen auch gefallen