A Novel Bug Report Extraction Approach

There are more and more bug reports in software. Software companies and developers invest a large number of resources into the dramatic accumulation of reports. We introduce Bayes classifier into bug reports compression, which is the first effort in the literature. For this purpose, the vector space model as well as some conventional text mining values, such as tf-idf and chi-squared test, are designed to collect features for bug reports. The experiment proves that bug reports extraction by using Bayes classifier is outperformance to the method based on SVM through the evaluation of ROC and F-score.

[1]  Shadi Banitaan,et al.  Bug Reports Prioritization: Which Features and Classifier to Use? , 2013, 2013 12th International Conference on Machine Learning and Applications.

[2]  Christian Kästner,et al.  Variability Mining: Consistent Semi-automatic Detection of Product-Line Features , 2014, IEEE Transactions on Software Engineering.

[3]  Gerard Salton,et al.  A vector space model for automatic indexing , 1975, CACM.

[4]  James C. Wetherbe,et al.  An Empirical Comparison of Four Text Mining Methods , 2010, 2010 43rd Hawaii International Conference on System Sciences.

[5]  Nikolaos G. Bourbakis,et al.  Graph-Based Methods for Natural Language Processing and Understanding—A Survey and Analysis , 2014, IEEE Transactions on Systems, Man, and Cybernetics: Systems.

[6]  T. Martin McGinnity,et al.  A Context-Based Word Indexing Model for Document Summarization , 2013, IEEE Transactions on Knowledge and Data Engineering.

[7]  Dane Bertram,et al.  Communication, collaboration, and bugs: the social nature of issue tracking in small, collocated teams , 2010, CSCW '10.

[8]  Gail C. Murphy,et al.  Automatic Summarization of Bug Reports , 2014, IEEE Transactions on Software Engineering.