HV: A Feature Based Method for Trajectory Dataset Profiling

摘要

The pervasiveness of location-acquisition and mobile computing techniques have generated massive spatial trajectory data, which has brought great challenges to the management and analysis of such a big data. In this paper, we focus on the trajectory dataset profiling problem, and aim to extract the representative trajectories from the raw trajectory as a subset, called profile, which can best describe the whole dataset. This problem is very challenging subject to finding the most representative trajectories set by trading off the profile size and quality. To tackle this problem, we model the features of the whole dataset from the aspects of density, speed and the directional tendency. Meanwhile we present our two kinds of methods to select the representative trajectories by the global heuristic voting (HV) function based on the feature model. We evaluate our methods based on extensive experiments by using a real-world trajectory dataset generated by over 12,000 taxicabs in Beijing. The results demonstrate the efficiency and effectiveness of our methods in different applications.