Person Movement and Stay Classification Based on Base Station Network Data

Ke Wang, Shaoyang Zhang, Kun Yu, Ying Li, Weike Fan, Zibin Bu, Qinghua Ren, Fei Gao

Abstract: The base station network data contains the real-time location information of the user, and accurately extracting the user's travel trajectory information from the Base station network data is of great significance for traffic regulation. In order to overcome the uncertainty of stop point identification in existing research, this paper proposes a person movement/stay classification method based on the improved Fast Newman algorithm. The method utilizes the base station handover relationship and the base station dwell time in the user trajectory data, abstracts the user's trajectory into a weighted undirected graph with a self-loop, and classifies the user state using the community partitioning result of the improved Fast Newman algorithm and speed. By comparing with the GPS data collected by the APP, the correct rate of this experiment is 92.33%.