An Empirical Study On Clustering For Isolating Bugs In Fault Localization

Debugging is the most labor-intensive and expensive task of the software development. And fault localization is one of the most time-consuming and money cost activities. A large amount of automatic fault localization technologies have appeared to help software testers to find faults and repair them as soon as possible. The core idea of automatic fault localization technology is to estimate the proper locations where the faults are by analyzing test results and software behaviors. Spectrum-based Fault Localization is one of the most popular automatic fault localization technologies.Spectrum-based Fault Localization (SBFL) techniques use risk evaluation formulas to calculate each statement’s likelihood of having a bug based on test results. SBFL can not only be used in statement level, but also can be used with other program entities such as branches, functions and so on. Most previous studies have been conducted under the assumption of a single bug. However, software always contains multi-bugs in practice. A natural idea of debugging is to isolate bugs and then use SBFL techniques to locate one bug for each group.In this paper, we conduct an empirical study on clustering for isolating bugs in fault localization. We analyze the effects of more than thirty fault localization techniques and popular cluster algorithms. The experiments conducted on the three different source programs which are written in C language, and these programs include tens of thousands line of codes.The main observations are:(1) ER5 (Wong1) achieves the best results of fault localization with clustering; (2) K-means outperforms hierarchical clustering for isolating bugs in fault localization.