Finding Time-dependent hot path from GPS trajectories

Abstract

Finding hot path is important in many scenarios like trip planning, traffic management and animal movement studies. However, in practice, the hotness of paths may change over time, e.g., for one path, it is hotter in morning rush hour than in the midnight. This paper studies how to find time-dependent hot path. Given two locations, a departure time and a travel time limit, our task is to get a hot path highly fitting the real-time physical world within a user-specified travel time limit. We first analyze the change of edge hotness in different time ranges by learning historical GPS trajectories. Then, with the time-dependent hotness information, we propose an effective algorithm to answer the hot path query mentioned above. Extensive experiments on a real dataset show that our methods outperform the baseline approaches in terms of both effectiveness and efficiency. © 2014 Springer International Publishing Switzerland.