Applying CodeBERT for Automated Program Repair of Java Simple Bugs
This addresses the time-consuming task of software debugging for developers, but it is incremental as it applies an existing pre-trained model to a specific domain.
The paper tackles automated program repair for Java simple bugs by fine-tuning CodeBERT on ManySStuBs4J datasets, achieving fix prediction accuracy of 19-72% in under a second per bug.
Software debugging, and program repair are among the most time-consuming and labor-intensive tasks in software engineering that would benefit a lot from automation. In this paper, we propose a novel automated program repair approach based on CodeBERT, which is a transformer-based neural architecture pre-trained on large corpus of source code. We fine-tune our model on the ManySStuBs4J small and large datasets to automatically generate the fix codes. The results show that our technique accurately predicts the fixed codes implemented by the developers in 19-72% of the cases, depending on the type of datasets, in less than a second per bug. We also observe that our method can generate varied-length fixes (short and long) and can fix different types of bugs, even if only a few instances of those types of bugs exist in the training dataset.