Inferring Demographics and Social Networks of Mobile Device Users on Campus From AP-Trajectories

Abstract

Exploring demographics and social networks of Internet users are widely used for many applications such as recommendation systems. The popularity of mobile devices (e.g., smartphones) and location-based Internet services (e.g., Google Maps) facilitates the collection of users' locations over time. Despite recent efforts to predict users' attributes (e.g., age and gender) and social networks based on utilizing the rich location context knowledge (e.g., name, type, and description) of places of interest (e.g., restaurants and hotels) they checked-in on location-based online social networks such as Foursqure and Gowalla, little attention has been given to inferring attributes and social networks of mobile device users based on their spatiotemporal trajectories with less/no location context knowledge. In this paper we collect logs of thousands of mobile devices' network connections to wireless access points (APs) of two campuses, and investigate whether one can infer mobile device users' demographic attributes and social networks solely from their spatiotemporal AP-trajectories. We develop a tensor factorization based method Dinfer to infer mobile device users' demographic attributes from their AP-trajectories by leveraging prior knowledge, such as users' social networks. We also propose a novel method Sinfer to learn social networks between mobile device users by exploring patterns of their AP-trajectories, such as fine-grained co-occurrence events (e.g., co-coming, co-leaving, and co-presenting duration). Experimental results on real-word datasets demonstrate the effectiveness of our methods.