Speaker：Dr Zhepeng (Lionel) Li(Schulich School of Business, York University)

Time：15:30-17:00 April 27,2017(Thursday)

Site：B226

Abstract：Adoption and social contagion studies have been focused on using homogeneous social networks. We distinguish two types of social entities: social actors and social foci by using two-mode social network. Based on relevant theories, we identify and operationalize factors that drive social contagion. Interdependences between social actors and social foci are captured by a mutual reinforcement model. We formulate the problem of forecasting contagious social foci and propose to address it using a gradient-based mutual reinforcement learning (GMRL) method. The optimal parameters for the predictive model are obtained via supervised learning from data. Empirical evaluations are extensively conducted using real-world data sets from popular location-based social networking platforms.