In this talk, we will share some interesting findings in a social media project about multimedia big data analytics. Since 4 years ago, researchers in HKUST-Social Media Lab. have been collecting over 10M+ images shared by users across different social media platforms from the West and East. We discovered a phenomenon that users with some related online connections are likely to share visually similar images than those are not related. By exploiting this phenomenon with latest deep learning technologies, unnoticeable user connections or interests could be potentially identified and predicted even without the access of user social graphs through Facebook or Twitter. We proved how these discovered user connections could be practical to recommendation and user profiling applications. Some interesting stories about this research and related ones will be shared to tell everyone how our Lab. is trying to bring our research impacts more than just the journals and patents.
James She is an assistant professor in the Department of Electronic and Computer Engineering at the Hong Kong University of Science and Technology (HKUST). He is the founding director of Asia’s first social media lab, HKUST-NIE Social Media Lab, and spearheads multidisciplinary research and innovation in multimedia big data analytics and systems, mobile/wearable-based interactive media systems, IoT-based interactive system and analytics. In addition, he is an associate editor for ACM Transaction on Multimedia Computing, Communications and Applications.
He holds a Ph.D. in Electrical and Computer Engineering from the University of Waterloo, and he is the recipient of multiple national, university and best paper awards, including the NSERC Postdoctoral Fellowship Award, NSERC Innovation Challenge Award, and Canada-UK Millennium Research Award. Check http://smedia.ust.hk/ for other details.
