Publications

Abstract : Defect estimation and severity analysis of software artifacts are very helpful for developing reliable software product. In order to achieve the reliable software within time and costs every software organization want to know how many software defect can be exit in developing software and which phase of software development life cycle is more severe. In this paper phase wise software defect severity analysis method is proposed using Bayesian networks. In the proposed method, severity of software defect in each phase of SDLC is predicted using top ranked reliability relevant software metrics. Bayesian belief network (BBN) and the linguistic values of software metrics related to requirement analysis, design, coding and testing phase have been considered to develop the proposed model. To validate the proposed model, 20 real software project data sets have been used. The predictive accuracy of the proposed model is validated and compared with existing work.