Multimedia Exploratory Data Analysis for Geospatial Data Mining: The Case for Augmented Seriation

The amount of spatial data collected from satellites, aerial photography, and land-based stations continues to grow at astounding rates. In this article, the role of exploratory data analysis (EDA) for spatial data mining is reviewed and a case study addressing environmental risk assessments in New York State is presented to illustrate the feasibility and usability of augmenting seriation for spatial data analysis. For this project, seriation, a univariate EDA technique, is augmented with a class of multimedia tools including iconic matrices, choropleth mapping, graphic interactions, and sound to exploit spatial datasets to better understand the relationships among spatial, temporal, and human variables. Additional software enhancements such as three-dimensional matrices, multivariate choropleth mapping, and tonal sonification are proposed to further improve these user-based tools for spatial data analysis.