Moving sequential pattern mining based on Spatial Constraints in Mobile Environment
Zou Yong-gui, Hong Zhi Yu · 2010
Moving sequential pattern mining is important in the trajectory analysis of moving objects and mobility prediction in mobile environment. The traditional sequential pattern mining method-PrefixSpan, is not applicable to the spatial constrained applications and it will produce a large number of duplicated projected databases in mining data sets. In order to overcome these drawbacks, a new algorithm is proposed, named Sequential Mining of Moving Patterns Based on Spatial Constraints in Mobile Environment (named SMPM). This algorithm has a property of spatial constraints, records the first position of the suffix in the sequence, and can avoid mining the same projected database and avoid physical projection according to the suffix's position and spatial constraints.