Neural Attribution for Semantic Bug-Localization in Student Programs

# 6






Abstract

Providing feedback is an integral part of teaching. Most open online courses on programming make use of automated grading systems to support programming assignments and give real-time feedback. These systems usually rely on test results to quantify the programs’ functional correctness. They return failing tests to students as feedback. However, the students may find it difficult to debug their programs if they receive no hints about where the bug is and how to fix it. In this work, we present NeuralBugLocator, a deep learning based technique, that can localize the bugs in a faulty program with respect to a failing test, without even running the program. At the heart of our technique is a novel tree convolutional neural network which is trained to predict whether a program passes or fails a given test. To localize the bugs, we analyze the trained network using a state-of-the-art neural prediction attribution technique and see which lines of the programs make it predict the test outcomes. Our experiments show that NeuralBugLocator is generally more accurate than two state-of-the-art program-spectrum based and one syntactic difference based bug-localization baselines

Rahul Gupta

Rahul Gupta is a Ph.D. candidate at the Department of Computer Science and Automation, IISc Bangalore. He is a member of the Software Engineering and Analysis lab, led by Prof. Aditya Kanade. His research interests lie broadly in developing deep learning techniques to solve problems in software engineering.