ESTATE: Strategy for Exploring Labeled Spatial Datasets Using Association Analysis

We propose an association analysis-based strategy for exploration of multi-attribute spatial datasets possessing naturally arising classification. Proposed strategy, ESTATE (

E

xploring

S

patial da

T

a

A

ssociation pat

TE

rns), inverts such classification by interpreting different classes found in the dataset in terms of sets of discriminative patterns of its attributes. It consists of several core steps 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. Detailed description of the ESTATE framework is followed by its application to the domain of ecology using a dataset that fuses the information on geographical distribution of biodiversity of bird species across the contiguous United States with distributions of 32 environmental variables across the same area.