Personalized defect prediction in practice

点击翻译按钮获取中文摘要

Defect prediction techniques identify probable code or code changes that are buggy, which help improve software quality. While many existing defect prediction techniques take the author of the code into consideration, none of these techniques build a separate prediction model for each developer. Different developers have different coding styles, commit frequencies, and experience levels, causing different defect patterns. When the defects of different developers are combined, such differences are obscured, hurting prediction performance. Personalized defect prediction builds a separate prediction model for each developer to predict software defects, which improve the performance of defect prediction. This project applies the personalized defect prediction technique on code bases from the industry partner and addresses the relevant research challenges to improve software quality and reduce development cost.