Meng Jiang
Meng Jiang

I am a postdoctoral researcher of Computer Science in University of Illinois at Urbana-Champaign working with Professor Jiawei Han. My research focuses on Data-Driven Behavioral Analytics with Networks. Problems I investigate vary from (1) mining behavior networks with social, spatiotemporal contexts (behavior prediction, recommendation and suspicious behavior detection), (2) attributed information network construction to (3) behavior and information network integration.

What's New

Latest Publications

Short Bio

Dr. Meng Jiang is a postdoctoral researcher in University of Illinois at Urbana-Champaign. He received his Ph.D. from the Department of Computer Science at Tsinghua University, Beijing in 2015. He received his bachelor from the same department in 2010. He visited Carnegie Mellon University in 2013 and University of Maryland, College Park in 2016. You can find more about him here:

His research lies in the field of data mining, focusing on user behavior modeling. He gives two tutorials in major conferences. His Ph.D. thesis won the Dissertation Award at Tsinghua University. He was the recipient of National Scholarship of China. His work on "Suspicious Behavior Detection" was selected as one of the Best Paper Finalists in KDD'14. His work on "Social Contextual Recommendation" has been deployed by the Tencent social network since 2012. The package of his work on "Automatic Attribute Discovery" is now transferring to U.S. Army Research Lab.

Topic of my recent talk: Data-Driven Behavioral Analytics with Networks.

Talk Abstract

We interact not just with each other but also with the world around us. Can we build intelligent and trustworthy user-oriented systems by analyzing massive behavior data of the interactions? In principle, data-driven approaches can extend the methodology of observation, representation and models beyond our eyes and hands. However, we lack means to fully exploit the structured contextual information and unstructured content information. In this talk, I will present three lines of work towards this goal: (1) Mining behavior network with social spatiotemporal context. (i) How can we recommend items in social network mechanism? What are the key factors of users' decisions? (ii) How can we catch suspicious behaviors made by smart bad actors? Instead of what they can do, what do they have to do? Two computational models were developed and deployed for social recommender systems and anti-fraud/spam systems; (2) Structuring behavioral content into attributed information network. Can we automatically discover attributes and facts from text corpus? A novel methodology, Meta Pattern mining, will be introduced for open-domain Information Extraction from the data-driven perspective; (3) Integrating behavior network with rich information network. I will uncover the potential of integrating structured and unstructured data for better understanding human behaviors, and conclude by sharing my thoughts for future directions.