Incremental maintenance of discovered mobile user maximal moving sequential patterns

In the context of mobile computing, a special sequential pattern, moving sequential pattern that reflects the moving behavior of mobile users attracted researchers' interests recently. While there have been a number of efficient moving sequential pattern mining algorithms reported, this paper concentrates on the maintenance of mined maximal moving sequential patterns. In particular, we developed an incremental approach, where maximal moving sequential patterns are stored in prefix trees, and new moving sequences can be easily combined with the existing patterns. A performance study indicated that the proposed approach performs significantly faster than straightforward approaches that mine from the whole updated database.