The Research On The Method Of Assessing Risk On Software Project Based On Bayesian Networks

This paper presents that the method of assessing risk on software project based on Bayesian Networks can make the results of risk assessment more intuitionistic, accurate and objective after discussing the theory and technique of the software project risk management and Bayesian Networks. Along with the increase of the requirements on software system, more and more questions which will result in software project failure appear, such as budget excess and schedule delay. So, how to identify and analyze risk accurately is becoming more important to software organizations.This paper firstly introduces the basic theory about risk management, especially the stage of risk identification and analysis. After analyzing several typical risk management models, an advanced risk assessment model is proposed into which some advantages of these typical models are integrated. In the most important part of risk analysis, the model adopts Bayesian Networks. Bayesian Networks is a graphic analytic tool which is based on the Bayesian theorem and can handle both of qualitative and quantitative information synchronously. The method of risk analysis based on BN can greatly enhance the effectiveness of the risk assessment.This paper also discusses the design of the risk historical database which can help managing the risk in advance and constructing the Bayesian Networks automatically. As an application instance, the detailed design of the whole model and the implementation of the part key modules are given in this paper. The results of the test on the system show that the method of assessing risk on software project based on Bayesian Networks is effective and practical.