摘要

With the increasing number of vehicles equipped with GPS enabled wireless communication devices, locations and trajectories of vehicles can be collected every time. Finding similar trajectories in massive trajectory data can benefit emerging novel mobile applications, such as carpooling, friend recommendation and traffic analysis. This paper proposes a novel spatio-temporal based trajectory similarity evaluation method. In this method, the significance of each point on the query trajectory can be assigned according to personal preference. Speed factor is also considered in the evaluation approach. Furthermore, most of existing trajectory similarity evaluation methods hold the precondition that the compared trajectories should have the same space and time length. This is not a compulsive condition in our method.

全文