For cybersecurity, we are working at malware modeling, detection, classification and generation with the focus on Javascript malware, desktop malware and Android malware. We are developing tools for vulnerability modeling and detection using machine learning and (both static and dynamic) program analysis on binary code. In our Securify research project (2015 - 2020), we are performing formal verification on security system from hardware, hypervisor, programs to security protocol using different verification approaches. Recently, we embark on the research on Automotive Security and autonomous vehicle Security in their security design, runtime security monitoring and response, and also the security testing and certification.

For software engineering, we are working on the topics related to program specification learning and model learning, performance analysis, Android energy analysis, reliability analysis, code clone analysis, program debugging, program testing, automatic loop analysis, testing and validating deep learning algorithms using techniques like model checking, symbolic execution and machine learning. We are building tools related to these aspects. For Android system, we have been working on security analysis on App malware detection & classification, generation and data analytic, App vulnerability analysis, App testing, Android OS testing and fuzzing, and Automatic UI generation.

For multi-agent systems, we are working on the topics related to formal modeling of various multi-agent systems, particularly trust management systems and their analysis in correctness, security and robustness.

For big data, we are promoting the concept called event analytic based on behavior learning and analysis, and their applications in sports and finance systems.