Mining Colossal Prevalent Spatial Co-location Patterns Based On Pattern Fusion Method

The evolution of location based service and technology of dealing spatial information is generating lots of rich spatial data sets.These large amounts of data with spatial contexts have been collected and stored in spatial database,which makes the study of data mining extend to spatial database mining form relational and transactional database.As one of the most important spatial data mining task,spatial co-location pattern mining has been popularly studied for discovering the spatial dependency of objects.Spatial co-location pattern represents a set of spatial features frequently observed together in a spatial proximity.However,since the frame of traditional co-location pattern mining is based on the like-Apriori algorithm,so the existing frequent co-location pattern mining algorithm encounter challenges at mining rather large patterns,called colossal frequent co-location patterns,in the presence of an explosive number of mid-sized frequent co-location patterns.This will affect the performance of algorithm and have low resource utilization.In order to solve this problem,we investigate a novel mining approach based on pattern fusion method to efficiently find a good approximation to the colossal co-location pattern.With pattern-fusion method,a colossal co-location pattern is found by fusing its small core patterns in one step,whereas the incremental pattern-growth mining strategies,such as those adopted in like-Apriori and FP-growth,have to examine a large number of mid-sized patterns.This property distinguishes the algorithm presented by us from all the incremental pattern growth mining algorithm.Since the pattern fusion is aimed at getting a good approximation to the complete set of the colossal co-location pattern,so we introduce a quality evaluation model to evaluate the result set.And the experiment result shows that the proposed algorithm is effective.