Association Rule Mining on Spatio-Temporal Processes

Entities in the real world evolve along the time-line from the start time point to the future in multi-dimensions space. The spatio-temporal process composed of meta-spatio-temporal processes, is bought out and used to describe this change process. The traditional association rule mining is extended to spatio- temporal processes, generating spatio-temporal process association rule mining which is used to extract association knowledge among spatio-temporal process data. Change processes of the western Pacific Ocean warm pool and rainfall of southeast area of China are the typical spatio-temporal processes, and relation between them is a remote relation. Based on spatio- temporal process association rule mining, we obtain some interesting association rules between change processes of them. Finally, it is concluded that spatio-temporal association rule mining can extract valuable association knowledge from spatio- temporal processes, and change trend of one entity or phenomenon can be forecasted through varying trend of others based on those association rules.