The impact of context metrics on just-in-time defect prediction
The impact of context metrics on just-in-time defect prediction
复制标题
DOI：
10.1007/s10664-019-09736-3
复制
发表时间：
2019-08
影响因子：
4.1
通讯作者：
Masanari Kondo;D. Germán;O. Mizuno;Eun-Hye Choi
中科院分区：
文献类型：
--
作者：
Masanari Kondo;D. Germán;O. Mizuno;Eun-Hye Choi
传统的及时缺陷预测方法一直使用软件的变更行来预测软件开发中的缺陷变更。然而，他们忽略了变化线周围的信息。我们的主要假设是，这些信息对变更有缺陷的可能性有影响。为了在缺陷预测中利用这些信息，我们考虑在变更线（我们称之为上下文线）之前和之后的线（n= 1,2，…），并提出度量它们的量度，我们称之为“上下文量度”。具体来说，这些上下文度量被定义为上下文行中单词/关键字的数量。在使用六个开源软件项目的大规模实证研究中，我们比较了使用我们的上下文度量、传统的代码流失度量（例如，修改的子系统的数量）、我们的扩展的上下文度量（不仅度量上下文线，而且度量变化的线）以及在预测模型中使用两个扩展的上下文度量的组合度量来进行缺陷预测的性能。结果表明，在统计测试中，考虑添加行的上下文线的上下文度量在所有情况下都获得了最佳中值。此外，使用较少的上下文行数适用于考虑单词的上下文度量，使用较多的上下文行数适用于考虑关键字的上下文度量。最后，在所有研究项目中，两个扩展情境指标的组合指标显著优于所有研究指标，其中包括接收者操作特征曲线下面积（AUC）和马修斯相关系数（MCC）。
Traditional just-in-time defect prediction approaches have been using changed lines of software to predict defective-changes in software development. However, they disregard information around the changed lines. Our main hypothesis is that such information has an impact on the likelihood that the change is defective. To take advantage of this information in defect prediction, we considern-lines (n= 1,2,…) that precede and follow the changed lines (which we callcontext lines), and propose metrics that measure them, which we call “Context Metrics.” Specifically, these context metrics are defined as the number of words/keywords in the context lines. In a large-scale empirical study using six open source software projects, we compare the performance of using our context metrics, traditional code churn metrics (e.g., the number of modified subsystems), ourextended context metricswhich measure not only context lines but also changed lines, andcombination metricsthat use two extended context metrics at a prediction model for defect prediction. The results show that context metrics that consider the context lines of added-lines achieve the best median value in all cases in terms of a statistical test. Moreover, using few number of context lines is suitable for context metric that considers words, and using more number of context lines is suitable for context metric that considers keywords. Finally, the combination metrics of two extended context metrics significantly outperform all studied metrics in all studied projects w. r. t. the area under the receiver operation characteristic curve (AUC) and Matthews correlation coefficient (MCC).