Exploring labeled spatial datasets using association analysis

Abstract

We use an association analysis-based strategy for exploration of multi-attribute spatial datasets possessing naturally arising classification. In this demonstration, we present a prototype system, ESTATE (Exploring Spatial daTa Association patTErns), inverting such classification by interpreting different classes found in the dataset in terms of sets of discriminative patterns of its attributes. The system consists of several core components including discriminative data mining, similarity between transactional patterns, and visualization. An algorithm for calculating similarity measure between patterns is the major original contribution that facilitates summarization of discovered information and makes the entire framework practical for real life applications. We demonstrate two applications of ESTATE in the domains of ecology and sociology. The ecology application is to discover the associations of between environmental factors and the spatial distribution of biodiversity across the contiguous United States, and the sociology application aims to discover different spatio-social motifs of support for Barack Obama in the 2008 presidential election.