Spatial High Utility Co-location Pattern Mining Based on Fuzzy Feature Clusters
Sets of spatial features whose instances frequently appear together in nearby areas are regarded as spatial co-location patterns.Spatial high utility co-location patterns is an extended research of spatial co-location pattern mining,which better reflects the high-quality aggregation phenomenon among spatial features.Most of the existing methods use the adaptive utility participation index(UPI)as a metric parameter for mining high utility co-locations.However,UPI does not take into account the influence of fuzziness and overlap of proximity relationships on the utility of co-location patterns.Furthermore,UPI does not satisfy the downward closure property,the mining efficiency is still not satisfactory.Combining fuzzy set theory,we establish fuzzy neighbor relationships and propose a new method for calculating the utility of co-location patterns.At the same time,we generate the fuzzy high utility feature clusters by the Fuzzy Chameleon Clustering algorithm,and extract spatial high utility co-location patterns from these high utility clusters.Finally,a large number of experiments are conducted on three synthetic datasets and two real datasets,which prove the rationality and effectiveness of the proposed algorithms.
spatial data miningspatial co-location patternhigh utilityfuzzy clusteringfuzzy feature cluster
万方数据
维普
