期刊名称：International Journal of Innovative Research in Science, Engineering and Technology
印刷版ISSN：2347-6710
电子版ISSN：2319-8753
出版年度：2017
卷号：6
期号：2
页码：2124
DOI：10.15680/IJIRSET.2017.0602116
出版社：S&S Publications
摘要：Arrangement bugs are one of the overwhelming reasons for programming disappointments. Past reviewsdemonstrate that a design bug could bring about enormous money related misfortunes in a product framework. Thesignificance of arrangement bugs has pulled in different research thinks about, e.g., to identify, analyze, and settle setupbugs. Given a bug report, an approach that can distinguish whether the bug is an arrangement bug could help designersdiminish investigating exertion. We allude to this issue as arrangement bug reports forecast. To address this issue, webuild up another mechanized structure that applies content mining advancements on the natural language portrayal ofbug reports to prepare a factual model on chronicled bug reports with known names (i.e., design or non-arrangement),and the measurable model is then used to anticipate a name for another bug report. Designers could apply our model tonaturally foresee names of bug reports to enhance their profitability. Our instrument first applies include choicemethods (e.g., data pick up and Chi-square) to preprocess the literary data in bug reports, and afterward appliesdifferent content mining systems (e.g., innocent Bayes, SVM, guileless Bayes multinomial) to construct factual models.We assess our answer on 5 bug report datasets including accumulo, activemq, camel, flume, and wicket. Wedemonstrate that credulous Bayes multinomial with data picks up accomplishes the best execution. By and large overthe 5 extends, its precision, arrangement F-measure and non-setup F-measure are 0.811, 0.450, and 0.880, separately.We likewise contrast our answer and the technique proposed by Arshad et al.. The outcomes demonstrate that ourproposed approach that utilizations credulous Bayes multinomial with data pick up all things considered enhancesprecision, design F-measure and non-setup F-measure scores of Arshad et al's. technique by 8.34%, 103.7%, and4.24%, separately.
关键词：bug Configuration; Bug Detection and Prediction; Selection of Features; Mining Principles.
