Portal de Periódicos da CAPES

Automated Classification of Issue Reports from a Software Issue Tracker

2017; Springer Nature; Linguagem: Inglês

10.1007/978-981-10-3373-5_42

2194-5357

Nitish Pandey, Abir Hudait, Debarshi Kumar Sanyal, Amitava Sen,

Text and Document Classification Technologies

Software issue trackers are used by software users and developers to submit bug reports and various other change requests and track them till they are finally closed. However, it is common for submitters to misclassify an improvement request as a bug and vice versa. Hence, it is extremely useful to have an automated classification mechanism for the submitted reports. In this paper we explore how different classifiers might perform this task. We use datasets from the open-source projects HttpClient and Lucene. We apply naïve Bayes (NB), support vector machine (SVM), logistic regression (LR) and linear discriminant analysis (LDA) separately for classification and evaluate their relative performance in terms of precision, recall, F-measure and accuracy.