Using cluster analysis for characteristics detection in software defect reports

Abstract

Computing and predicting testing metrics are highly important software defect management tasks. Before testing metrics are used, one needs to understand and evaluate the full array of software defects found during testing. There can be thousands of software defects found during testing, and it is difficult to process all of them en masse. Cluster analysis solves this problem, as it can compress the data by grouping a set of objects into a cluster. Moreover, clustering helps understand the nature of defects, point out weak functional areas of the software, and improve the testing strategy. This paper introduces a technique used for software defect reports clustering.