THUMP: Semantic Analysis on Trajectory Traces to Explore Human Movement Pattern

Shreya Ghosh, Soumya K. Ghosh

Abstract: Exploring human movement pattern from raw GPS traces is an interesting and challenging task. This paper aims at analysing a large volume of GPS data in spatio-temporal context, clustering trajectories using geographic and semantic location information and identifying different categories of people. It tries to exploit the fact that human moves with an intent. The proposed framework yields encouraging results using a large scale GPS dataset of Microsoft GeoLife.