Systematic Analysis of Learning-Based Software Fault Localization

This paper reviews the evolution of learning-based software fault localization methods, examining their benefits, challenges, and prospective developments in practical applications. We analyze the limitations of traditional methods and provide an in-depth discussion of learning-based fault localization approaches, covering the application of supervised, unsupervised, and reinforcement learning techniques in software fault localization. Furthermore, we explore key issues in current research and future research directions, and analyze threats that could impact the effectiveness of the research. In conclusion, we summarize the current status and future prospects of these methods, along with our perspective on the forthcoming advancements in this field.