Yang Xu, Shih-Lung Shaw, Jiaoli Chen, Qingquan Li, Zhixiang Fang, and Yuguang Li
“Global Positioning System (GPS) based vehicle tracking data have been used to derive useful traffic data such as computing travel speed or congestion level (e.g., Herrera et al., 2010; Mohan, 2008) or measuring urban dynamics (e.g., Calabrese et al., 2011; Reades et al., 2007). Vehicle tracking data also have been used to analyze travel activities (e.g., Li et al., 2011; Liu et al., 2010). This study, on the other hand, focuses on identifying repeated spatio-temporal patterns embedded in a GPS-based taxi tracking dataset collected in Wuhan, China. Although taxi trajectories may appear to be chaotic at first glance, there could be important repeated spatio-temporal patterns embedded in taxi tracking data. For example, certain taxi drivers may have preferences of waiting for passengers at particular locations such as airports, train stations, etc. In many developing countries, it also is common to have two work shifts of drivers for one taxi. Identifying repeated spatio-temporal patterns embedded in vehicle tracking data thus can shed light on important travel and activity behavioral patterns.
“This study uses a taxi tracking dataset collected in Wuhan, China as a case study to identify repeated spatio-temporal behavioral patterns among the taxi drivers. The main objective of this study is to develop a systematic method that can facilitate uncovering repeated behavioral patterns in a large tracking dataset. This method can be adapted for studies using other types of tracking data such as cell phone tracking data of individual trajectories or online tracking data of individual web browsing histories.”
- Read the paper [PDF]