Debugging with Intelligence Via Probabilistic Inference

Zhaogui Xu,Shiqing Ma,Xiangyu Zhang,Shuofei Zhu,Baowen Xu

DOI: https://doi.org/10.1145/3180155.3180237

2018-01-01

Abstract:We aim to debug a single failing execution without the assistance from other passing/failing runs. In our context, debugging is a process with substantial uncertainty - lots of decisions have to be made such as what variables shall be inspected first. To deal with such uncertainty, we propose to equip machines with human-like intelligence. Specifically, we develop a highly automated debugging technique that aims to couple human-like reasoning (e.g., dealing with uncertainty and fusing knowledge) with program semantics based analysis, to achieve benefits from the two and mitigate their limitations. We model debugging as a probabilistic inference problem, in which the likelihood of each executed statement instance and variable being correct/faulty is modeled by a random variable. Human knowledge, human-like reasoning rules and program semantics are modeled as conditional probability distributions, also called probabilistic constraints. Solving these constraints identifies the most likely faulty statements. Our results show that the technique is highly effective. It can precisely identify root causes for a set of real-world bugs in a very small number of interactions with developers, much smaller than a recent proposal that does not encode human intelligence. Our user study also confirms that it substantially improves human productivity.