Information Studies Seminar Series: "Discovering Social Interaction Mechanisms and Diffusion Networks in Online Social Networks." Guest speaker Dr. W. Cheung

Join us for a McGill School of Information Studies (SIS) Seminar Series talk with Visiting Professor Dr. William Cheung on social interaction mechanisms and diffusion networks in online social networks.

ABSTRACT

The use of online social media for communication is now an irreversible trend. High volume of related activities captured via this new media has enabled human behaviors to be studied using the data-driven approach. Very often, those recorded interaction activities form large-scale complex networks. Recovering behavioral regularities from the networks is challenging as social relationship among people is often complex and hidden.

This talk consists of two parts. For the first part, I will present how we are inspired by the study of network motifs (in complex networks) and the theories of human communication (in social science), and propose the Stochastic Network Motif modeling approach to represent and discover social interaction mechanisms from human interactions observed in social networks. Other than characterizing different networks using the discovered motifs, we have also derived a motif-based link prediction algorithm which can increase the accuracy of followee recommendation based on a Weibo dataset by 5-10%. For the second part, I will present how we are inspired by notions like structure hole and social role in social network research and propose variants of independent cascade models for better modeling of information diffusion in social networks. One of our recent works that incorporates stochastic network motifs into diffusion network modeling (to be presented at the upcoming IJCAI’16) will also be covered.

BIO

Dr. William K. Cheung is currently the Associate Head and an Associate Professor in the Department of Computer Science, Hong Kong Baptist University. He received BSc and MPhil degrees in electronic engineering from The Chinese University of Hong Kong and a PhD degree in computer science from The Hong Kong University of Science and Technology. He has served as the co-chair and/or program committee member of a number of international conferences/workshops, as well as a guest editor of journals in areas including artificial intelligence, Web intelligence, data mining, Web services, e-commerce technologies, and health informatics. Also, he is currently the Managing Editor of the IEEE Intelligent Informatics Bulletin. His research interests include artificial intelligence and machine learning, as well as their applications to collaborative filtering, social network analysis, health data analytics, web/text/network data mining, distributed and privacy-preserving data mining, among others. Website: http://www.comp.hkbu.edu.hk/~william