Coupling Co-location Patterns On Spatial Data Sets And Its Mining Methods

There is a variety of interesting knowledge in spatial data sets.Spatial co-location pattern mining can discover sets of different features that are co-located.The guiding ideology of co-location pattern mining is “the closer the location of spatial objects,the more similar the properties”.From this point of view,look for the features that appear together in the data set,call them patterns,and then use appropriate criteria to select the patterns that users need.However,in the real world,the closer the location of objects not necessarily mean that they have similar properties.Tobler's first law of geography states that “all attribute values on a geographic surface are related to each other,but closer values are more strongly related than are more distant ones”.We believe the relation comes from the mutual influence and interaction between spatial objects due to the close distance.Based on this idea,we propose a new pattern based on the co-location pattern,namely the coupling co-location pattern.Unlike the co-location pattern only lists the features that appear together.New pattern that takes into account the coupling not only considers the combination of different features,but also considers the appear number of same features.On the other hand,this paper proposes a new pattern extraction and measurement.In this paper we propose a novel spatial pattern called coupling co-location patterns.First,we discuss the properties of the coupling phenomenon between spatial features,and then the concept of coupling co-location patterns is defined formally.Second,the measurement of support and mining framework for coupling co-location patterns are proposed.Since the mining process needs to search for maximal cliques,this paper quotes a novel and efficient algorithm for maximal clique enumeration.And based on the idea of the algorithm,a new maximal clique algorithm is proposed.Finally,we conduct experiments on both real and synthetic data sets,and the results verify the practical significance of coupling co-location patterns.