Software Defect Prediction Model Research for Network and Cloud Software Development

With the process of software development and application basing on network and cloud, the change of software development requires new software defect prediction method for these kinds of software development, which can solve the problems of the traditional software defect prediction method based on target project, such as the same predict background and higher cost of defect tagging. A new software defect prediction method based on multi source data oriented network and cloud development process is proposed. This method selects the predictive candidates from multisource projects which have similar characteristics as objective projects, and then guides the training data selection by the software modules, finishes the prediction based on Naive Bayesian algorithm. Finally through algorithm experiment this method is proved superior to the traditional WP prediction model.