Computer Science Seminar
Faculty Candidate 2014
Unifying Manual and Mechanical Changes for Better Software Testing
When: Monday, April 7, 2014
Where: PGH 232
Time: 11:00 AM
Speaker: Dr. Lingming Zhang, University of Texas at Austin
Host: Prof. Zhigang Deng
Software testing is the most commonly used methodology for validating software quality. In practice, regression testing is a widely used approach to validate that manual code changes are not faulty. In research, mutation testing is a widely used approach to evaluate testing techniques based on injected mechanical code changes.
This talk presents the first unification of these two well-known testing approaches, which have been studied by researchers for decades but always applied separately. The unification is driven by the concept of change, which is central to each approach, albeit in different forms: (1) manual changes to code provide the basis for regression testing techniques; and (2) mechanically induced changes to code provide the basis of mutation testing. Novel unification techniques, which use the foundations of regression testing to enable more efficient mutation testing and the foundations of mutation testing to enable more effective debugging in regression testing, are introduced. Rigorous experimental evaluation using a variety of real-world Java programs is presented to validate the efficacy of the techniques. The talk concludes with an overview of new research ideas that are inspired by insights into change unification and aim to provide more scalable and effective techniques for testing and maintaining modern software systems.
Bio:
Lingming Zhang is a Ph.D. student in the Department of Electrical and Computer Engineering at The University of Texas at Austin. He works under the supervision of Dr. Sarfraz Khurshid. Lingming received the M.S. degree in Computer Science from Peking University, and the B.S. degree in Computer Science from Nanjing University in China. His research interests lie broadly in software engineering and programming languages, including automated software analysis, testing, debugging, and verification, as well as software evolution.