A Bayesian Networks-Based Risk Identification Approach for Software Process Risk: The Context of Chinese Trustworthy Software

For all of the many advantages and enormous benefits that information technology has brought to us, it has subjected to increasing risks that need enough attention. Bayesian network (BN) is an important probabilistic inference approach to support reasoning under uncertainty. This paper describes how BN is applied to quantify the occurrence probability of software process risk factors and the influence strength among the process risk factors in the context of Chinese trustworthy software. The information of 52 factors was obtained through a questionnaire survey of 93 project managers in five high Capability Maturity Model Integration (CMMI) level software companies. The focus of this paper is to present the key risk checklist and good timing for process risk control to improve software process risk management. Special effort has been put on the description of the experimental study, which provides the top 20 key risk factors in software process and critical software sub-processes for process risk management. The findings can provide the key risk checklist to software risk manager for risk identification and decision-making in process risk management. This is a general approach and, as such, it can be applied to a certain software project or some software enterprises with updated data.