Discovering the Impact of Urban Traffic Interventions Using Contrast Mining on Vehicle Trajectory Data

摘要

There is growing interest in using trajectory data of moving vehicles to analyze urban traffic and improve city planning. This paper presents a framework to assess the impact of traffic intervention measures, such as road closures, on the traffic network. Connected road segments with significantly different traffic levels before and after the intervention are discovered by computing the growth rate. Frequent sub-networks of the overall traffic network are then discovered to reveal the region that is most affected. The effectiveness and robustness of this framework are shown by three experiments using real taxi trajectories and traffic simulations in two different cities.