Data Mining is process of discovering new techniques, patterns and trends by dredging into mining large amounts of data stored in warehouses via the use of artificial intelligence, statistical and mathematical techniques.(Chowdry et al.,) Data Mining can also be defined as an indespensible toool in decision support system which plays a major role in market segmentation, fraud detection, customer services, credit and behaviour scoring in banking.([Paolo, 2001] and [Thomas, 2000]). . There are other terms used in referring to Data Mining; such as Knowledge discovery in databases (KDD), knowledge extraction, data or pattern analysis. The importance of collecting data that reflect business or scientific activities to achieve competitive advantage is widely recognised now powerful systems for collecting data and managing it in large databases are in place in all large and mid-range companies. However, the hindrance of turning this data into business application is the difficulty of extracting knowledge about the system studied from the collected data. DSS is th earea of information systems that is focused on supporting and improving top management decisions.( D. Arnott, and G. Pervan, 2005).
This is inherently prescription which enhances decision making in some way in business most especially the banking where enormous data is used and store in a large data warehouse and this requires pattern to solve it. DSS are closely related to the concept of rationality which states that the tendency to act in a reasonable way to make good decision. (Chowdury, et al.,)
This research based on two factors which include:
Data Mining
Decision Support System
As reiterated earlier, decision support system is a dependent factor on Data Mining which is a Data Warehouse that stores all the banks data from different departments and sections within its business activities the bank operates, and this research will show, how patterns are and can be created to help the retail banking make decisions where and when needed. This would be done by using the CRISP-DM which states different steps needed to carry out necessary tasks on the generated data, also using object oriented programming technologies with additional software like SPSS, Rapid Miner to aid the decision generated from patterns created from mined data in to different sets of data and also make management decision about a product or department.
Aims and Objectives
Research Aims
The main aim of the research project is to review and analyze the use of Data Mining and Decision Support Systems in retail banking. The relationship between Data Mining and Decision Support System making impact in the banking activities. This would be done by understanding the business activities in banks and also identify the areas where this services is required and most especially why Data Mining is an essential guide for the deployment of Decision Support System.
The research will then explores, design and develop Data Mining scenarios and Decision Support System for business requirement in banking and evaluate the likely impact that such technology will have on these banks.
The study aims to answering the following questions:
How does DM and DSS work together to enhance business activities in banking
To what extent do banks use data mining and decision support systems? What types of tools is/are required?
Is there any impact due as a result of the combined technology in improvements to the banking operations?
What types of strategic decisions that already enhanced or improved by using Data Mining and Decision Support System?
The main objectives of the research project are to discover:
How DM and DSS work for retail banking.
Identify areas where DM and DSS are utilized and can be utilized in retail banking.
Analyze different patterns in the retail banking.
Identify all business activities that require the application.
Demonstrate how business activities can utilize the application.
Methodology
Conduct an analytical study in the business understanding, data sets in order to learn the extent of integration of data mining and decision support system in banks. A questionnaire will be used as data collection tool in order gather information about the usage of data mining in banks, and at the same time the decision support it gives the users to manage the banking operations to serve its customers.
The questionnaire will also be used to gather other information relevant to the entities of the topic. Where some tools would be used to carry out the methods of integrating the business undersatnding with data intergration.
The process includes downloading and installing open source software for datamining which are: Weka Software and Rapid Miner, this two softwares performs same funtions in data mining but, would want to have thw two should one of it fails testing, also the use use of Ms-Excel application to input data and import it to the repository platform created in the softwares, Ms-Office visio will be used to plot the linkage of different data tables depending on the data available, SPSS SPSS statistical tool to analyse the gathered data through the questionnaire. IBM-SPSS will be the main statistical analysis tool to be used in this study, correlations, T-test, A-Test, Chi-Square are expected to be implemented in this case and Ms-Word would be used for reporting the whole Data Mining process.
Project Plan
I will be adopting the CRISP-DM methodology to tackle the development process of this topic because CRISP-DM framework shows the cycle in which data mining works and how the mining flows in the cycle thereby reteirating what each of the six stages requires for the implementation .
Crisp -DM methodology framework for this project includes:
Business Understanding :Understanding the business objectives, by assessing the situation and also determining the data mining goals.
Data Understanding: Gathering the initial data, decribing, exploring data and verifying the data quality.
Data Preparation: Select data after verifying the data quality, clean, construct and integrate the selected data and format the data.
Modelling: Select modelling technique, generate test design and build a model to assess the mining model.
Evaluation: Evaluate result, review the evaluation process and determine the nextstep.
Deployment: Plan deployement, plan monitoring and assistance, produce final report and review report.
Creating a large data set, pre-processing of data, filtering or clearing, data transformation, identifying dimensionally and useful feature. It also involves the classification, association, regression, clustering and summary.
Choosing the mining algorithm is the most important parameter for the process. The final stage includes pattern evaluation which means visualisation, transformation, removing redundant pattern etc, use of discovery knowledge of the process.
The DSS aspect will involve the use of programming technologies i.e. OO programming to develop a GUI to display the decision of differ data patterns as to what the client and user require of the data at a particular time.
Figure.1: CRISP DM Diagram
Source: www.crip-dm.org/CRISPWP-0800.pdf
OVERVIEW
The overview of this study is looking at the banking operations and how Data Mining is functional in the retail banking environment, also as a decision support system to display output and aid business activities.
To achieve this, the research will explore a number of areas through review of extensive literature which includes:
What is Data Mining?
Why Organizations need Data Mining?
What is the relation between Data Mining and Decision Support System?
What is the current use of Data Mining in banks?
What is the current use of Data Mining and Decision Support System in Banks?
REFERENCES:
Arnott, D.and Pervan, G. "A critical analysis of Decision Support Systems research",
Journal of Information Technology, 20, 2, June, 2005, pp67-87.
Usama F., Paul S., (November 1997) . Data mining and KDD: Promise and challenges
Future Generation Computer Systems, Volume 13, Issues 2-3, , Pages 99-115
S. Sharma, K.-M. Osei-Bryson (2009) .Framework for formal implementation of the business understanding phase of data mining projects/ Expert Systems with Applications 36 (2009) 4114-4124
Y. I-Cheng, L. Che-hui (2009) . The comparisons of data mining techniques for the predictive accuracy of probability of default of credit card clients Expert Systems with Applications 36 2473-2480
G. Paolo, (2001) Bayesian data mining with application to benchmarking and credit scoring, Applied Stochastic Models in Business and Society 17 (2001), pp. 69-81.
L.C. Thomas,(2000) A survey of credit and behavioral scoring: Forecasting financial risk of lending to consumers, International Journal of Forecasting 16 (2000), pp. 149-172.
M.A.Chowdury., et al (2004). Data warehousing & Data mining; Prediction Decision Support
System for Renewal energy in Bangladesh. In proceedings of the conference on "Energy For Sustainable Development: Technology Advances & Environmental Issues"
M. Kamber and J. Han, (2006). Data Mining:Concept and Techniques.
E. Turban, et al,.(2005). Decision Support Systems and Intelligent Systems And Opportunities. 2005.
Intech Open Beta. intechopen.com. [Online] [Cited: 3 November 2010.] http://www.intechopen.com/books/show/title/data_mining_and_knowledge_discovery_in_real_life_applications.
www.crip-dm.org/CRISPWP-0800.pdf [Assesed on 03/10/2010]
