Use of Data Mining in System Development Life Cycle - Data Mining: Theory, Methodology, Techniques, and Applications - page 105

Use of Data Mining in System Development Life Cycle
Richi Nayak 1 and Tian Qiu 2
1 School of Information Systems, QUT, Brisbane, QLD 4001, Australia
[e-mail]
2 EDS Credit Services, Adelaide, Australia
[e-mail]
Abstract. During the life cycle of a software development project, many
problems arise. Resolutions to these problems are time consuming and
expensive. This paper discusses the use of data mining in solving some of these
problems to improve the system development life cycle process. A case study of
applying data mining to the software Problem Report management data is also
presented. The empirical results demonstrate the capability and benefit of data
mining analysis in systems development life cycle.
1 Introduction
The System Development Life Cycle (SDLC) includes various phases during which
the defined software products are created or modified [22]. These phases include
planning, definition, requirement analysis, design, development, testing and
integration, implementation, operation and maintenance. During the SDLC process,
huge repositories for configuration management, risk management, project metric
report and problem report management are maintained in addition to source code.
These repositories are potential sources of useful information that can be used in
improving the SDLC process. Researchers have started using data mining (DM)
[5,9,10,11,16,18,21] techniques in this process. Some examples of DM usage are in (1)
software maintenance by summarizing and augmenting software changes, (2) software
development process by automatically generating test cases and checking their outputs,
(3) software reuse by predicting success or failure of components beforehand, and
matching and discovering reusable patterns, and (4) project planning and estimation by
identifying relationships between human resources and product types.
This paper discusses the capability and benefit of data mining analysis in systems
development life cycle. The paper is organised into two parts. The first part includes
the discussion on various DM applications in SDLC. The second part presents a case
study of applying DM to the software problem report management data.
2 Data Mining Applications in SDLC
Figure 1 illustrates a general framework of applying DM techniques to aid the SDLC
process. We have summarised the use of DM in SDLC into two areas: (1) software
development process that includes management rule generation, risk assessment and