Locating Relevant Source Files for Bug Reports using Textual Analysis

Locating Relevant Source Files for Bug Reports using Textual Analysis

Bug reports are an important part of software project’s life cycle since they help improve the software’s quality. However, in well-known systems, the huge number of bug reports make it difficult for the developer team to efficiently locate the bug and then assign it to be fixed. To solve this issue, various bug localization techniques have been proposed to rank all the source files of a project with respect to how likely they are to contain a bug. This makes the source files’ search space smaller and helps developers to find relevant source files quicker. In this paper, we propose a three component bug localization approach which leverages different textual properties of bug reports and source files as well as the relations between previously fixed bug reports and a newly received one. Our approach uses information retrieval, textual matching, and multi-label classification to improve the performance of bug localization. We evaluate our approach on two open source software projects (i.e. SWT and ZXing) to examine its performance. Experimental results show that our approach can rank appropriate source files for more than 80% of bugs in top 10 for these projects and also improve the MRR and MAP values compared to two existing bug localization tools, BugLocator and BLUiR.