面向软件缺陷报告的提取方法.PDFVIP

面向软件缺陷报告的提取方法 1 1 1 1 2 林涛 , 高建华 , 伏雪 ， 马燕 ， 林艳 （上海师范大学计算机科学与工程系，上海，中国，200234 ）1 (奥克兰大学 信息系统系，奥克兰，新西兰，92019)2 摘要 软件工程中的软件缺陷报告在快速增长，开发者们越来越困惑于大量的缺陷报告。因此，为了缺陷 修复和软件复用等目的，有必要研究软件缺陷报告的提取方法。该文提出一种提取方法，该方法首先合并 缺陷报告中的同义词，然后建立空间向量模型，使用词频反文档频率以及信息增益等文本挖掘的方法来收 集软件缺陷报告中单词的特征，同时设计算法确定句子复杂度以选择长句。最后将贝叶斯分类器引入该领 域。该方法可以提高缺陷报告提取的命中率，降低虚警率。实验证明基于文本挖掘和贝叶斯分类器的软件 缺陷报告提取方法在接受者工作特征曲线面积 （0.71）、F-score （0.80）和Kappa 值（0.75）方面有良好效 果。 关键词 软件缺陷报告管理、文本挖掘、贝叶斯分类器、软件缺陷报告特征、空间向量模型、句子复杂度 中图法分类号 TP311.5 文献标识码 A Extraction Approach for Software Bug Report 1 1 1 1 2 LIN Tao , GAO Jian-hua , FU Xue , MA Yan , LIN Yan （Department of Computer Science and Technology, Shanghai Normal University, Shanghai , 200234, China ）1 (Department of Information Systems, The University of Auckland, Auckland, 92019, New Zealand) 2 Abstract: There are increasingly bug reports in software engineering and developers bewildered by the rapid reports accumulation. Therefore, it is necessary to extract bug report for the task of bug fixing and software reuse, etc. The paper proposes a novel extraction approach. Synonyms were merged into one specific word firstly in the approach. Then, it sets up a vector space model, and some text mining methods, such as TF-IDF and information gain, are presented to collect word features for bug reports specifically, but also there is an algorithm for determining sentence complexity, so as to choose the sentence in long length. This work introduced Bayes classifier into bug report extraction. TPR is increased and FPR is decreased in this approach. The experiment proves that bug report extraction by using data mining and Bayes classifier is competiti