Software Life Cycle-based Defects Prediction and Diagnosis Technique Research
A model based on Bayesian network is put forward to predict and diagnose software defects before project. In the model causes and effects inference is used to predict defects, and the Bayesian formula is introduced to analyze the prediction result that helps to find the root reason of bring defects. The model considers every phase of software life cycle, such as requirement, design, development and testing, maintenance. The model computes predict result through variational weight of affect-factor. The computation results of specifically affect-factor by model compared with practical defect in same condition which to indict that the model is validity. The model can predict and find defects early and effectively, it can control the software quality and the cost of development.
software life cycle software defects bayesian network causes and effects inference bayesian formula
Cuifang Zheng Jiaju Wu Fang Peng Zhijie Wu
Institute of Computer Application, China Academy of Engineering Physics, Mianyang Sichuan P.R.China Institute of Computer Application, China Academy of Engineering Physics,Mianyang Sichuan P.R.China
国际会议
太原
英文
192-195
2010-10-22（万方平台首次上网日期，不代表论文的发表时间）