TY - GEN
T1 - An automated bug triage approach
T2 - 8th International Conference on Intelligent Computing Technology, ICIC 2012
AU - Zhang, Tao
AU - Lee, Byungjeong
PY - 2012
Y1 - 2012
N2 - Generally speaking, the larger-scale open source development projects support both developers and users to report bugs in an open bug repository. Each report that appears in this repository must be triaged for fixing it. However, with huge amount of bugs are reported every day, the workload of developers is so high. In addition, most of bug reports were not assigned to correct developers for fixing so that these bugs need to be re-assigned to another developer. If the number of re-assignments to developers is large, the bug fixing time is increased. So "who are appropriate developers for fixing bug?" is an important question for bug triage. In this paper, we propose an automated developer recommendation approach for bug triage. The major contribution of our paper is to build the concept profile(CP) for extracting the bug concepts with topic terms from the documents produced by related bug reports, and we find the important developers with the high probability of fixing the given bug by using social network(SN). As a result, we get a ranked list of appropriate developers for bug fixing according to their expertise and fixing cost. The evaluation results show that our approach outperforms other developer recommendation methods.
AB - Generally speaking, the larger-scale open source development projects support both developers and users to report bugs in an open bug repository. Each report that appears in this repository must be triaged for fixing it. However, with huge amount of bugs are reported every day, the workload of developers is so high. In addition, most of bug reports were not assigned to correct developers for fixing so that these bugs need to be re-assigned to another developer. If the number of re-assignments to developers is large, the bug fixing time is increased. So "who are appropriate developers for fixing bug?" is an important question for bug triage. In this paper, we propose an automated developer recommendation approach for bug triage. The major contribution of our paper is to build the concept profile(CP) for extracting the bug concepts with topic terms from the documents produced by related bug reports, and we find the important developers with the high probability of fixing the given bug by using social network(SN). As a result, we get a ranked list of appropriate developers for bug fixing according to their expertise and fixing cost. The evaluation results show that our approach outperforms other developer recommendation methods.
KW - bug triage
KW - concept profile
KW - developer recommendation
KW - fixing cost
KW - reassignment
KW - social network
UR - http://www.scopus.com/inward/record.url?scp=84865294078&partnerID=8YFLogxK
U2 - 10.1007/978-3-642-31588-6_65
DO - 10.1007/978-3-642-31588-6_65
M3 - Conference contribution
AN - SCOPUS:84865294078
SN - 9783642315879
T3 - Lecture Notes in Computer Science (including subseries Lecture Notes in Artificial Intelligence and Lecture Notes in Bioinformatics)
SP - 505
EP - 512
BT - Intelligent Computing Technology - 8th International Conference, ICIC 2012, Proceedings
Y2 - 25 July 2012 through 29 July 2012
ER -