Mining Urban Trajectories

点击翻译按钮获取中文摘要

Advances in modern sensing technologies and large scale computing infrastructures have generated a huge amount of trajectory data. These urban trajectory data contain rich information of moving objects including people, vehicle, etc., and foster a wide variety of applications, such as road network reconstruction, intelligent traffic management, and air quality control. While governments, industrial companies, and organizations have realized the tremendous potential of these data, their complex nature presents unique challenges that cannot be easily solved.The objective of the proposed research program is to address the challenges arising from mining urban trajectory data, and provides answers to the following three research questions: How to understand objects' mobility patterns from their trajectories? How to fuse the heterogeneous trajectories, and solve the data sparsity problem? How to detect anomalies from urban trajectories?The results of this research will advance the state-of-the-art technique in trajectory data mining, providing new insights on understanding people's moving behaviors from various perspectives. Besides, it will bring significant benefit to different parties, including governments, businesses, and end users, through a series of new applications and services enabled by the research results.