Empirical Study of Fault-Prone Method's Name and Implementation: Analysis on Three Prefixes-Get, Set and Be
Empirical Study of Fault-Prone Method's Name and Implementation: Analysis on Three Prefixes-Get, Set and Be
复制标题
易错方法命名及实现的实证研究——Get、Set、Be三个前缀分析
DOI：
10.1109/acit-csii-bcd.2017.20
复制
发表时间：
2017
期刊：
影响因子：
--
通讯作者：
Minoru Kawahara
中科院分区：
文献类型：
--
作者：
Sho Suzuki;Hirohisa Aman;Minoru Kawahara
本文重点研究Java方法的第一个单词（前缀）与其实现之间的关系，以预测容易出错的方法。作为聚焦分析方法的试点研究，本文分析了方法名称的三大前缀：“get”、“set”和“be”。本文的实证研究收集了大量方法名称和实现特征的数据，分析了方法名称和实现特征之间关系的异常情况。然后，该研究检查了八个开源软件产品中方法的异常与其故障倾向之间的关系。实证结果证明了关注方法名称与其实现特征之间的差距在预测容易出错的方法中的有用性。
This paper focuses on the relationship between Java method's first word (prefix) and its implementation, for predicting fault-prone method. As a pilot study of the focused way of analysis, this paper analyzes three major prefixes of methods' names: "get", "set" and "be." The empirical study in this paper collects many data of methods' names and implementation features and analyzes abnormality of relationship between methods' names and their implementation features. Then, the study examines the relationship between a method's abnormality and its fault-proneness in eight open source software products. The empirical results prove the usefulness of focusing on the gap between methods' names and their implementation features in predicting fault-prone methods.