Previous
Next

Abstract 2 of Volume 0 n. 0 - 2003

Visualization of Large Data Sets:
The Zoom Star Solution

M. Noirhomme-Fraiture

Abstract

Progress in Knowledge Discovery in Large Databases has raised the necessity of visualizing sets of data on aggregated form. These sets can be given as native data or as results of preliminary analysis. Symbolic data analysis has been developed to solve the problem of the analysis of data known on an aggregated form, i.e. where quantitative variables are given by intervals and where categorical variables are given by histograms. These data are called symbolic objects. In this paper, we present several methods and software dedicated to the visualization of symbolic objects. They use radial graphs. They can be represented in 2D (Zoom Star, Simple Star) or in 3D (3DZoom Star, Temporal Star). Each picture concerns only one set of individuals on aggregated form, but several objects can be compared when representing them side by side or superimposed. It is also possible to visualize the evolution of an object varying with time. Sound has been used to represent particular characteristics. The visualization modules have been developed thanks to different European IST projects.
 
Previous
Next
 
Download the full paper: PDF version PPT presentation
 
Table of Contents of the Volume 0 - n.0 - July 2002