Correspondence Clustering: An Approach to Cluster Multiple Related Spatial Datasets

Abstract

Abstract. Domain experts are frequently interested to analyze multiple related spatial datasets. This capability is important for change analysis and contrast mining. In this paper, a novel clustering approach called correspondence clustering is introduced that clusters two or more spatial datasets by maximizing cluster interestingness and correspondence between clusters derived from different datasets. A representative-based correspondence clustering framework and clustering algorithms are introduced. In addition, the paper proposes a novel cluster similarity assessment measure that relies on reclustering techniques and co-occurrence matrices. We conducted experiments in which two earthquake datasets had to be clustered by maximizing cluster interestingness and agreement between the spatial clusters obtained. The results show that correspondence clustering can reduce the variance inherent to representative-based clustering algorithms, which is important for reducing the likelihood of false positives in change analysis. Moreover, high agreements could be obtained by only slightly lowering cluster quality.

A polygon-based methodology for mining related spatial datasets

“…Then co-clustering partitions rows and columns of the data matrix and creates clusters which are subsets of the original matrix. Correspondence clustering [9] is introduced by Rinsurongkawong et al to cluster two or more spatial datasets by maximizing cluster interestingness and correspondence between clusters. Cluster interestingness and correspondence interestingness are captured in a plug-in fitness functions and prototype-based clustering algorithms are proposed that cluster multiple datasets in parallel.…”

“…In conclusion, coupled clustering [3] and co-clustering [4], [5] are not designed for spatial data and they cluster point objects using traditional clustering algorithms. The techniques introduced in correspondence clustering [9] are applicable to point objects in the spatial space whereas this paper focuses on clustering spatial clusters that originate from different, related datasets that are approximated using polygons.…”