Tensor Decomposition for Spatiotemporal Mobility Pattern Learning with Mobile Phone Data

Detecting urban mobility patterns is crucial for policymakers in urban planning. Mobile phone data have been increasingly deployed to measure the spatiotemporal variations in human motion. This work applied non-negative Tucker decomposition (NTD) to mobile-phone-based OD (Origin-Destination) matrices to explore mobility patterns' hidden spatial and temporal relationships in the Province of Liège, Belgium. Four 310 x 310 x 24 traffic tensors have been built for one regular weekday, one regular weekend day, one holiday weekday, and one holiday weekend day, respectively. The proposed method inferred twenty-six spatial clusters and three temporal patterns. Besides, the study focused on interpreting the correlation between spatial clusters and temporal patterns through geographical visualization. As a result, we found the symmetry for the trips of the temporal patterns with PM peak and AM peak on the weekday. Moreover, we investigated the attraction of different spatial clusters over three temporal patterns on weekdays and
weekends by combing the land use map. Finally, the differences in spatial and temporal interactions on four data sets have been addressed in detail.