A Unifying Framework for Clustering with Plug-In Fitness Functions and Region Discovery

The goal of spatial data mining [SPH05] is to automate the extraction of interesting and useful patterns that are not explicitly represented in spatial datasets. Of particular interests to scientists are techniques capable of finding scientifically meaningful regions in spatial datasets as they have many immediate applications in medicine, geosciences, and environmental sciences, e.g., identification of earthquake hotspots, association of particular cancers with environmental pollution, and detection of crime zones with unusual activities. The ultimate goal of region discovery is to provide searchengine-style capabilities that enable scientists to find interesting places in geo-referenced data automatically and efficiently.