We are honored to have our paper [A Bayesian Framework for Automated Debugging] accepted to the 32nd International Symposium on Software Testing and Analysis. As the title implies, the paper proposes a theoretical framework for automated debugging through the lens of Bayesian statistics. We demonstrate that existing theoretical results about fault localization can be re-derived using our Bayesian framework. In addition, we apply our framework to analyze existing automated program repair and unified debugging techniques to verify the wide potential use of our framework. To see if our framework can make useful predictions as well, we use it to directly derive a patch prioritization technique, BAPP, that utilizes program values. Our results reveal that BAPP can indeed reduce the number of required patch validations by 68%, and also improve the precision of fault localization.

If you are interested, you can find many more details in our preprint.