Rank and Classification of Bug Reports and Feature Evaluation Using SVM
Abstract: When a new bug report is received, developers usually need to procreate the bug and perform code reviews to find the origination, this process can be tedious and time consuming. A instrument for ranking all the source files with regards to how likely they are to contain the cause of the bug would enable developers to narrow down their search and better result. This paper introduce an adaptive ranking approach that leverages project acquaintance through functional decomposition of source code, API metaphors of library components, the bug-fixing record, the code change record, and the file dependency graph. Given a bug report, the ranking evaluation of each source file is computed as a weighted collection of an array of feature film, where the weights are trained mechanically on previously resolved bug reports using a learning-to-rank technique. We estimate the ranking system on six large scale open source Java projects, using the earlier-fix version of the project for every bug report. The explore results show the Support Vector Machine approach that outperforms in defect prediction.
Keywords: Bug Localization, Feature Evaluation, Information Retrieval, Ranking, SVM
How to Cite?: Gomathi S, Swathini K, Suvetha M, Deepa M, " Rank and Classification of Bug Reports and Feature Evaluation Using SVM", Volume 6 Issue 2, February 2017, International Journal of Science and Research (IJSR), Pages: 1288-1291, https://www.ijsr.net/getabstract.php?paperid=ART2017697, DOI: https://dx.doi.org/10.21275/ART2017697