Data Intensive Software Analytics (DISA)
We conduct research on the development of intelligent software systems using machine learning and natural language processing that focus on addressing critical problems within technical, social, and organizational contexts of software development teams. As such, our research work often lies at the intersection of software engineering, data science, human computer interaction, and social science.
AI4SE: AI for Software Documentation and Dependability
Combine Machine Learning (ML), Natural Language Processing (NLP), and Software Engineering (SE) techniques to automatically analyze and improve software dependability (e.g., software security vulnerability analysis and detection), reliability (e.g., program repair), and documentation (e.g., software review and insight summarization, produce better quality software library documentation).
SE4AI: Software Engineering for Machine Learning Application
Improving the design, engineering, quality control, continuous integration, and maintenance of Machine Learning Software Application (MLSA) based on an incorporation of MLSA specific attributes into traditional software development life cycles (SDLC).
Using statistical, machine learning, and natural language processing techniques to analyze, synthesize, and fuse vast amount of software repository data (e.g., crowd-sourced/internal software forums, code repositories, issue tracking and code reviews) to derive active and passive insights that could be useful for various software teams and stakeholders.