Research in Social Computing Detects Meaningful Patterns in Messy Data

Jundong Li joined UVA Engineering to enhance the benefits and uses of big data. Li joined the Charles L. Brown Department of Electrical and Computer Engineering as an assistant professor in the fall 2019 term, with a joint appointment in the Department of Computer Science and School of Data Science. Li’s dissertation research focused on algorithms to make predictive models more accurate and productive, for example by gleaning actionable patterns from high-dimensional or networked data and tackling practical problems such as detecting anomalies that point to malicious activity.

The diversity of research underway in electrical and computer engineering aids Li’s efforts to detect meaningful patterns in messy data—data that is connected, incomplete and dynamic—and to fuse data from different sources. Leveraging the department’s strengths in signal processing, information theory, security of cyber-physical systems and bioinformatics, Li can generate a robust understanding of how these patterns emerge and to what effect.  

“We are at the stage of developing first principles,” Li said. “In addition to complementary engineering disciplines, we also need to involve domain experts in public policy, psychology, biology and physics.” This synergy supports Li’s research on computational health. Li is applying machine learning to the problem of cyberbullying to help parents keep tabs on their kids’ mental and emotional wellbeing with data about their activity online.  

Li received his Ph.D. degree in computer science at Arizona State University in 2019. He also holds a Master of Science degree in computer science from the University of Alberta, Canada, and a Bachelor of Engineering degree in software engineering from Zhejiang University in China. Li has published more than 50 papers, with more than 1,000 citations. He also leads the development of an open-source feature selection repository available on GitHub, a free, online platform where 40 million developers host and review code, manage projects and build software together across more than 100 million repositories.