Reality mining: Digging the impact of friendship and location on crowd behavior

Abstract

Crowd behavior of human deserves to be studied since it is common that people are influenced and change their behavior when being in a group. In pervasive computing research, an amount of work has been directed towards discovering human movement patterns based on wireless networks, mainly focusing on movements of individuals. It is surprising that social interaction among individuals in a crowd is largely neglected. Mobile phones offer on-body tracking and they are already deployed on a large scale, allowing the characterization of user behav- ior through large amounts of wireless information collected by mobile phones. In this paper, we observe and analyze the impact of friendship and location attributes on crowd behavior, using location-based wire- less mobility information. This is a cornerstone for predicting crowd behavior, which can be used in a large number of applications such as crowdsourcing-based technology, traffic management, crowd safety, and infrastructure deployment.