MoveMine: Moving Object Trajectory Clustering

Abstract

With the maturity of Geographical Positioning System (GPS), wireless, and Web technologies, increasing amounts of movement data collected from various moving objects, such as animals, vehicles, mobile devices, and climate radars, has become widely available. Analyzing such data has broad applications, such as, in ecological study, vehicle control, mobile communication management, and climatological forecast. MoveMine is designed for sophisticated moving object data mining by integrating several attractive functions including moving object pattern mining and trajectory mining. Trajectory clustering is one of the major functions in trajectory mining. Existing trajectory clustering algorithms group similar trajectories as a whole, thus discovering common trajectories. In this paper, moving object trajectory clustering is presented. In trajectory segmentation, the global optimum segmentation can be found by dynamic programming. We present a pattern based clustering algorithm that extends k-means algorithm for clustering moving object trajectory data. The system will be evaluated on elk, deer and cattle’s movement dataset which has been generated by the Starkey project for effectiveness of the system.