Accurate developer recommendation for bug resolution

Xin Xia, David Lo, Xinyu Wang, Bo Zhou

Research output: Chapter in Book/Report/Conference proceedingConference PaperResearchpeer-review

91 Citations (Scopus)

Abstract

Bug resolution refers to the activity that developers perform to diagnose, fix, test, and document bugs during software development and maintenance. It is a collaborative activity among developers who contribute their knowledge, ideas, and expertise to resolve bugs. Given a bug report, we would like to recommend the set of bug resolvers that could potentially contribute their knowledge to fix it. We refer to this problem as developer recommendation for bug resolution. In this paper, we propose a new and accurate method named DevRec for the developer recommendation problem. DevRec is a composite method which performs two kinds of analysis: bug reports based analysis (BR-Based analysis), and developer based analysis (D-Based analysis). In the BR-Based analysis, we characterize a new bug report based on past bug reports that are similar to it. Appropriate developers of the new bug report are found by investigating the developers of similar bug reports appearing in the past. In the D-Based analysis, we compute the affinity of each developer to a bug report based on the characteristics of bug reports that have been fixed by the developer before. This affinity is then used to find a set of developers that are 'close' to a new bug report. We evaluate our solution on 5 large bug report datasets including GCC, OpenOffice, Mozilla, Netbeans, and Eclipse containing a total of 107,875 bug reports. We show that DevRec could achieve recall@5 and recall@10 scores of 0.4826-0.7989, and 0.6063-0.8924, respectively. We also compare DevRec with other state-of-art methods, such as Bugzie and DREX. The results show that DevRec on average improves recall@5 and recall@10 scores of Bugzie by 57.55% and 39.39% respectively. DevRec also outperforms DREX by improving the average recall@5 and recall@10 scores by 165.38% and 89.36%, respectively.

Original languageEnglish
Title of host publicationProceedings - 20th Working Conference on Reverse Engineering, WCRE 2013
PublisherIEEE, Institute of Electrical and Electronics Engineers
Pages72-81
Number of pages10
ISBN (Print)9781479929313
DOIs
Publication statusPublished - 1 Dec 2013
Externally publishedYes
Event20th Working Conference on Reverse Engineering, WCRE 2013 - Koblenz, Germany
Duration: 14 Oct 201317 Oct 2013

Publication series

NameProceedings - Working Conference on Reverse Engineering, WCRE
ISSN (Print)1095-1350

Conference

Conference20th Working Conference on Reverse Engineering, WCRE 2013
CountryGermany
CityKoblenz
Period14/10/1317/10/13

Keywords

  • Composite Method
  • Developer Recommendation
  • Multi-label Learning
  • Topic Model

Cite this