摘要
针对如何提高代码评审效率的问题,本文提出了BPR-CR2评审者推荐模型.该模型中结合了评审者与代码Pull请求的专业关联性、与Pull请求提出者的社交关联性、与Pull请求的代码路径相关性以及评审者的积极性因素,基于贝叶斯个性化排序的思想学习每个评审者在进行Pull请求选择时的权重偏好,从而能够对每个Pull请求推荐评审者.在Github平台的5个流行项目的数据集上进行了测试,与目前5个典型算法相比,BPR-CR2的性能优于其他算法.
In order to improve the efficiency of code review,we propose a Bayesian Personalized Ranking based Code Review er Recommendation model(BPR-CR2).BPR-CR2 considers multiple factors that include the know ledge level relevance betw een review ers and the pull request,social connections betw een reviews and submitters of the pull request,the path relevance betw een review and the pull request and the activeness of review ers.Besides,BPR-CR2 learns the preferences of review ers based on Bayesian Personalized Ranking(BPR)so that multiple review ers can be recommended to a pull request.The model is tested on the five popular projects collected from Github platform.Comparing with five typical algorithms,the performance of BPR-CR2 is better than other algorithms.
作者
李敏
赵海燕
陈庆奎
曹健
LI Min;ZHAO Hai-yan;CHEN Qing-kui;CAO Jian(Shanghai Key Lab of Modern Optical System,and Engineering Research Center of Optical Instrument and System,Ministry of Education,University of Shanghai for Science and Technology,Shanghai 200093,China;Department,University,Department of Computer Science and Technology,Shanghai Jiaotong University,Shanghai 200030,China)
出处
《小型微型计算机系统》
CSCD
北大核心
2021年第1期27-33,共7页
Journal of Chinese Computer Systems
基金
国家重点研发计划项目(2018YFB1003800)资助。
关键词
代码评审
审阅者推荐
开源社区
贝叶斯
code review
reviewer recommendation
open source community
Bayesian
