Bug Reports Summarization: A Case Study

Abstract

Bug Report is one of the most important software artifacts which helps the developers during the software maintenance and evolution process. It is a conversation-based artifact which not only contains the insights on the resolution of a bug or an issue but also contains suggestions for the enhancements. Automatic Bug Report summarization is to automatically create the summaries for bug reports to help developers save time and effort which they spent on reading and understanding the report. There are many techniques like machine learning approaches, deep learning and unsupervised approaches which have been successfully implemented for the summarization of bug reports. This paper gives a review on various techniques and the works which have been done in this field. The paper systematically studies the papers related to bug report summarization and classifies the works according to the technique used. We have also analyzed the various evaluation techniques which have been used to measure the effectiveness of the approach. The paper also discusses the future direction for research in this field.

Keywords: Extractive summarization, Feature based extraction, MMR, Summarization, Unsupervised summarization.

View PDF