Mining Individual Mobility Patterns Based on Location History
Xiaopeng Chen, Dianxi Shi, Banghui Zhao, Fan Liu · 2016
At present, it is a hot-spot in the research of data mining to mine mobility patterns on the basis of GPS trajectories. In order to solve the problems that data sampling frequency uncertainty, the value of experience as period parameter by human input, and spatiotemporal noises and outliers, we propose a framework called PMPM (Periodic Mobility Pattern Mining) to adaptively detect periodic parameters of GPS trajectories, and mine individual mobility patterns. First of all, some approach and algorithms are proposed to extract reference spot set from raw trajectories. Secondly, period of every reference spot is detected by a probabilistic model. Then individual mobility patterns are mined using frequent pattern mining algorithms. Finally, we use real collected data to perform functions for mining individual mobility patterns and demonstrate effectiveness of our framework.