Discovering Local Social Groups using Mobility Data

Abstract

The ever increasing popularity of location aware mobile devices, has facilitated the collection of large amounts of human mobility data. This data usually contains where a person was and when, which can give us an insight into the complex social behavior of people. In this paper we explore the use of co-location as a means to estimate the strength of relationship between people. First we modify existing co-location algorithms to take the significance of the location of co-occurrence into account in predicting relationship strength. Next, we propose a system to identify social groups that exists in geographic areas called Local Social Groups by using co-location information. Finally, we run extensive experiments on our sample data to test the efficiency of our approach.