Moving Object Grouping Rule Mining Based on Accumulated Spatio-temporal Data

With the advance of mobile electronic devices and the development of positioning technology, a large volume of spatio-temopral data are collected in the form of desultorily data streams, which contain a lot of potential information. In this study, we focus on discovering the composition relationships between observation moving objects in a long period. Such research can be widely used in military and civilian areas, including recommendation systems, wildlife research, military monitoring and battlefield situation awareness. The composition relationships of moving objects can be called as moving object grouping rule. In this paper, we proposed an improved traveling companion discovery method based on Nearest neighbor of time to obtained the object transactions in short time and used the incremental association rule mining (ARM) method to discovering the grouping rules of moving objects in long-term.