A Dynamic Fault Localization Technique with Noise Reduction for Java Programs

Existing fault localization techniques combine various program features and similarity coefficients with the aim of precisely assessing the similarities among the dynamic spectra of these program features to predict the locations of faults. Many such techniques estimate the probability of a particular program feature causing the observed failures. They ignore the noise introduced by the other features on the same set of executions that may lead to the observed failures. In this paper, we propose both the use of chains of key basic blocks as program features and an innovative similarity coefficient that has noise reduction effect. We have implemented our proposal in a technique known as MKBC. We have empirically evaluated MKBC using three real-life medium-sized programs with real faults. The results show that MKBC outperforms Tarantula, Jaccard, SBI, and Ochiai significantly.