ISPLOM: Interactive with Scatterplot Matrix for Exploring Multidimensional Data

Abstract

The scatterplot matrix is one of the most common methods for multidimensional data visualization. The scatterplot matrix is usually used to display all pairwise of data dimensions, that is organized a matrix. In this paper, we propose an interactive technique with scatterplot matrix to explore multidimensional data. The multidimensional data is projected into all pairwise orthogonal sections and display with scatterplot matrix. A user can select a subset of data set that separates from the rest of data set. A subset of the data set organized as a hierarchy cluster structure. The hierarchy cluster structure is display as a radial tree cluster. The user can select a cluster in the radial tree and all data points in this cluster are display on scatterplot matrix. The user is repeated this process to identify clusters. One of the most useful of our method can be identify the structure of multidimensional data set in an intuition fashion.