Publication

Abstract : Recently, Bayesian Belief Network (BBN) becomes one of the most popular choices for uncertainty modeling and has been widely used in software engineering such as defect prediction, reliability and quality prediction, testing effort prediction and software risk assessment. The Node Probability Tables (NPT) play a vital role in BBN. Failure data is not available in the early phases (i.e., phases which occur before testing phase) of software development life cycle (SDLC). However, metrics of early phases of SDLC can be assessed qualitatively. Therefore, an intelligent selection of software metrics also plays a vital role in developing a software defect prediction model using BBN. In this paper, a technique has been proposed to develop the NPT of a BBN using the qualitative value of software metrics.