A Project on Software Defect Prevention at Commit-Time: A Success Story of University-Industry Research Collaboration

Abstract

In this talk, we describe a research collaboration project between Concordia University and Ubisoft. The project consists of investigating techniques for defect prevention at commit-time for increased software quality. The outcome of this project is a tool called CLEVER (Combining Levels of Bug Prevention and Resolution techniques) that uses machine learning to automatically detect coding defects as programmers write code. The main novelty of CLEVER is that it relies on code matching techniques to detect coding mistakes based on a database of historical code defects found in multiple related projects. The tool also proposes fixes based on known patterns.