Intelligent Decision Support (IDS) in Software Risk Management Based on Data Mining, Rough Sets and Decision Theory

  • Mutheu, Rose, Wasike, Jotham
Keywords: Software Risks, Risk Decision-Making, Data Mining, Rough Set Theory, Decision Theory, Intelligent Decision Support.


Risks are intrinsic to any project and risk-taking is a necessary component of any process of decision making. High risks from software projects threaten healthy development of any Nation because of the complex nature of projects. For sustainable development, we should focus on risk assessment and risk decision. Assessment of risks in most of the software projects has been done qualitatively ignoring the risk decision. Attempts at risk decisions have been based on individuals’ rational opinions hence subjective. Previous reports have shown limited evidence on successful use of DSSs in practice.  To address this anomaly, this study proposes intelligent decision support that provides more objective, repeatable, and observable decision – making support for software risk management. Software risk managers will be supported in gathering and analyzing evidence, identifying and diagnosing problems, proposing possible courses of action and evaluating such proposed actions.  IDS is based on data mining, rough sets and decision theories which improve decision making in uncertain conditions. Risks will be looked at as identifiable and quantifiable possible events or factors from which negative or positive consequences may occur. The main sources of data will be a set of secondary data collected over time and knowledge of domain expert (s). The techniques used in this paper will provide efficient algorithms for finding hidden patterns in software risks and generate sets of decision rules to support decisions in software risks management.