基于Bayes网的软件需求预测技术探讨

Discussion of Software Requirement Prediction Technique Based on Bayes

软件需求贯穿于整个软件开发周期,是影响软件质量的重要因素。在项目开展之前控制住软件需求的质量,对提高软件质量、合理控制软件开发过程有着重要的意义。考虑到影响软件需求质量的因素多而不确定,而Bayes网把用概率分布表示的科学严格性与用专家经验表示的验前信息有机结合起来,能很好地处理不确定问题。于是建立了基于Bayes网的软件需求质量预测模型,并用贝叶斯公式对出现的预测结果进行了分析。提出预测模型结构清晰,采用贝叶斯网络的概率计算有较强理论基础,计算简洁、易于理解。

Software requirement impenetrate whole software development cycle,it is an important factor for software quality. It is important for improving software quality,controlling software development process rational before developing project to control software quality. Considering that the factor which affected software requirement quality are more and uncertainty, but Bayes newtwork combination with the science strict of probability distributing and the prior information which is expressed by expert experience,it can deal with undetermined problem well. So the software requirement prediction model based on Bayes network is proposed and the Bayes formula is introduced to analyze the prediction result. The prediction model of this paper is legible in framework. The probability computation based on Bayesian has strong theory basis and is concise to compute and easy to understand.