
Spatio-temporal Event Detection, Forecasting, and Modeling in Social Media
Speaker: Liang Zhao, Graduate Research Assistant at Virginia Polytechnic Institute and State University
Friday, March 25, 2016 – 11:00am – 12:00pm – Engineering Building, Room 3507
Abstract:
Nowadays, social media has become an important social sensor for significant social issues such as influenza epidemics, social movements, and public sentiment. Spatiotemporal social events detection, forecasting, and modeling in social media are important and promising problems. These open problems still suffer from a series of challenges including: 1) Dynamics of social media streams, 2) Heterogeneity of spatial and social network, and 3) Noisy and sparse nature of social media content. In this talk, I will describe new models that can address these challenges and effectively capture the underlying predictive patterns for social events. The performance of these models will be demonstrated on important applications of detection and forecasting for significant societal events, such as civil unrest and disease outbreaks. Finally, some of my future research works will also be discussed.
Bio:
Liang Zhao is a PhD candidate in the Department of Computer Science at Virginia Tech. His research interests include data mining and machine learning, with particular emphasis on social media modeling, feature selection, and text mining. He has led the papers in prestigious conferences and journals including ACM SIGKDD, IEEE ICDM, SIAM Data Mining, PLoS One, and IEEE BigData, and served as the reviewer for leading conferences and journals such as ACM SIGKDD, ACM TKDD, IEEE ICDM, SIAM Data Mining, ACM TIST, ACM SIGSPATIAL, and Geoinformatica. He also owns two US IP discloses on social media mining.