Mining Group Periodic Moving Patterns from Spatio-Temporal Trajectories

Periodic behaviors is essential to understanding objects' movements. In real world situations, the movement of groups hides useful periodic patterns. Mining such periodic patterns can benefit many applications, such as urban planning, traffic management and public security. However, the previous work mainly focuses on detecting individual periodicities, and rarely studies group periodic behaviors. In this paper, we propose an algorithm for mining group periodic moving patterns, called GPMine, which adopts filter-refine paradigm to discover patterns. In the filter phase, GPMine filters the initial candidates generated by sub-patterns, and refines them to determine the final results in the refine phase. Furthermore, in order to improve the performance of pattern mining, spatial pruning is devised to filter out invalid candidates by spatial proximity. To reduce search space, an index structure is designed to support more efficient trajectory queries. Experiments on two real trajectory datasets have verified the effectiveness and efficiency of our proposed algorithms respectively.