Research in the group currently focuses on the security and other quality aspects of mobile apps, distributed systems, and multi-language software, primarily through developing and applying scalable and cost-effective program analysis techniques in synergy with data-driven (especially machine/deep learning) approaches. We sometimes also conduct empirical studies to understand the problem space better before setting out to develop technical solutions.
Scalable and cost-effective program analysis techniques for industry-scale distributed software systems and their applications to the maintenance, evolution, and security analysis of those systems and, more broadly, autonomic information security defense in adversarial and dynamic environments (ASE’16, TOSEM’20, TSE’21, USENIX Security’21)
Data quality problems in learning deep code models for software vulnerability analysis, e.g. leveraging generative models for automated, massive generation of realistic data samples to support training of powerful, generalizable deep code representations and scientific assessments of existing analysis techniques (FSE’22, TSE’22, ICSE’23)
Program analysis and run-time support for testing and validation of multi-language software systems, and their applications to discovering and diagnosing correctness bugs and security vulnerabilities across (programming) language boundaries (USENIX Security’22, FSE’22, USENIX Security’23, ICSE’23, CCS’23)
Mobile software engineering with a focus on Android app and system security including longitudinal characterizations as well as effective and sustainable malware analysis using learning-based approaches (TIFS’19, ISSTA’19, TSE’20, TOSEM’20)
Our research has been enabled by grants offered by several funding agencies. We gratefully acknowledge their great support!