Friday, 14 June 2024, 13:30, HN 1.33/1.37, Hanna Neumann Building
Abstract
How do we test software when detailed specifications are unavailable? Effective testing requires both input and behavioural specifications. However, often in the industry, one cannot rely on the availability or reliability of either. In this talk, I will demonstrate how to extract detailed specifications from given programs when the provided specifications are unreliable. I will explain how to leverage detailed error feedback to construct input and behavioural models of the program. These models can be used to verify program behaviour, identifying, and eliminating software bugs before they are deployed in production.
Biography
Rahul Gopinath is a lecturer at the University of Sydney, where he specializes in the broad fields of Software Engineering and Cybersecurity. Previously, he was a postdoctoral scholar at CISPA Helmholtz Center for Information Security in Germany. His research focuses on inferring program input specifications and behavioural oracles under whitebox and blackbox conditions, which can then be leveraged for fuzzing.