Spatial and temporal mining method using GPS data

Xiaolei Li

School of Electronic and Information, NingboDahongyingUniversity, Ningbo, Zhejiang, 315175, China

Geographic information has spawned many novel Web applications where global positioning system (GPS) plays important roles in bridging the applications and end users. Learning knowledge from users’ raw GPS data can provide rich context information for both geographic and mobile applications. However, so far, raw GPS data are still used directly without much understanding. Spatial-temporal data analysis plays an important role in many applications, including transportation infrastructure, border security and inland security. To analyse the moving patterns of vehicles on a road network, a measure for determining the similarity of vehicle trajectories with respect to space and time has to be defined. Although previous research has addressed the trajectory similarity problem, most of the studies focus on Euclidian distance instead of network distance. This paper deals with the variations in applying a spatial-temporal similarity measure with given Points of Interest (POI) and Time of Interest (TOI), treating spatial similarity as a combination of structural and sequence similarities that is evaluated using the techniques of dynamic programming. The similarity set thus formed will be used by the remote database to broadcast trigger-based messages to participating vehicles in a neighbourhood for future route- and information-sharing activities. The performance of the scheme is evaluated using experiments on standard real-life data.