Fault localization via behavioral models

Abstract

Software errors significantly impact software quality, and improvement of fault localization can reduce the expense of debugging. SBFL is promising approach and have received a lot of attention due to its simplicity and effectiveness. However, there are still some limits in this approach. In this paper we address these limits and present a technique to build a novel behavioral model, and propose a corresponding fault localization technique. Furthermore, we performed a study to empirically validate the effectiveness of our approach. And experimental results indicated that our method outperformed other representative SBFL techniques in locating faults in the benchmark. We show that the effectiveness of the technique comes from the guidance of hierarchical behavioral model.