基于机器学习的软件质量预测技术综述  

Software Quality Prediction Technology Based on Machine Learning

作　　者：姚日煌[1,2] 高海洋 周聪 洪智学 YAO Rihuang;GAO Haiyang;ZHOU Cong;HONG Zhixue(CEPREI,Guangzhou 511370,China;Shenzhen CEPTEI Industrial Technology Research Institute Co.,Ltd.,Shenzhen 518055,China)

机构地区：[1]工业和信息化部电子第五研究所,广东广州510610 [2]深圳赛宝工业技术研究院有限公司,广东深圳518055

出　　处：《电子产品可靠性与环境试验》2023年第6期16-22,共7页Electronic Product Reliability and Environmental Testing

摘　　要：如今,各个行业对软件的要求越来越高,而软件的检测是实现软件质量的关键,也是整个系统的一个关键环节。在开发软件之前,开发人员常常要花费很多时间对软件可能出现的故障进行预测分析,以确保其可靠度,从而设计出高质量和高可靠性的软件系统。以往的机器学习的自动化测试技术可以有效地帮助软件从业人员提高效率。随着技术发展,机器学习技术已经可以实现对软件系统故障过程的预测,且辅助软件质量管理机制和质量控制流程的构建,在减轻开发人员工作负担的同时提高软件的质量和可靠性。通过对近10年有关机器学习预测软件可靠性的文献进行了研究,总结了软件质量预测领域的相关技术。Nowaday,various industries are demanding more and more software,and software testing is the key to achieving software quality and a critical aspect of the overall system.Before developing software,developers often spend a lot of time predicting and analyzing possible software failures to ensure its reliability in order to design high-quality and highly reliable software systems.In the past,automated testing techniques with machine learning could effectively help software practitioners improve efficiency.With the development of technology,machine learning technology has been able to achieve the prediction of software system failure process and assist in the construction of software quality management mechanism and quality control process,it can improve the quality and reliability of software while reducing the workload of developers.By studying the literature on machine leaming to predict software reliability in the last decade,the relevant techniques in the field of software quality prediction are summarized.

关 键 词：软件工程 软件质量 机器学习 软件质量预测 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]