This project investigates how explainable AI techniques can help software engineers in identifying the root causes of bugs and understanding why suggested fixes are effective. Current automated bug detection and fixing tools do not provide developers with informative explanations, requiring security expertise to interpret these tools. During this project, developer-centric explanations tailored to varying expertise levels will be generated from software execution. Through studies with software developers, this research will evaluate and refine explanation quality, contributing methodologies for interpretable root-cause analysis and practical insights into developers’ explanatory needs.
People
Publications
2026
How Humans, Bots, and Agents Communicate About Vulnerabilities in Pull Requests
Pien Rooijendijk, Christoph Treude, Mairieli WesselICSME
Explainable AI for Software Vulnerabilities
Pien RooijendijkCHASE DECS
Who Said CVE? How Vulnerability Identifiers Are Mentioned by Humans, Bots, and Agents in Pull Requests
Pien Rooijendijk, Christoph Treude, Mairieli WesselMSR Mining Challenge