A Data-Driven Model for Software Development Risk Analysis Using Bayesian Networks

In this paper, a data-driven model based on Bayesian Networks (BNs) is presented for the risk analysis of software development. The modeling process consists of three phases: BN initialization, conflict analysis, and risk monitoring and analysis. Using new project data obtained from the process of software development, the model can continually estimate risk probability, identify the sources of risk, and perform model revision. The significance of the work is that the model provides more objective and visible support for risk analysis in software development.