Frequent spatiotemporal trajectory pattern mining based on pheromone concentration

With the development of positioning technologies (GPS, GSM networks, etc.), the real time data of mobile objects becomes increasingly available. It is leading to new opportunity of discovering behavior pattern and useful knowledge automatically in spatiotemporal database. We focus our study on frequent trajectory pattern mining for moving trajectory in this paper. In particular, we introduce a novel method which integrates stay time and visited frequency to detect interesting areas. Based on interesting areas, we transformed trajectory data into stay time sequence with respect to finite interesting areas. Finally, a spatiotemporal trajectory mining algorithm is proposed to discover frequent trajectory pattern. The approaches are then validated by a range of real and synthetic data sets to evaluate the usefulness and efficiency.