Visual guidance in the exploration of large databases

  • Authors:

  • Jiang Du;Ian Spence;Michael J. McGuffin

  • Affiliations:

  • University of Toronto and École de Technologie Supérieure (ÉTS);University of Toronto and École de Technologie Supérieure (ÉTS);University of Toronto and École de Technologie Supérieure (ÉTS)

  • Venue:
  • Proceedings of the 2010 Conference of the Center for Advanced Studies on Collaborative Research
  • Year:
  • 2010

Quantified Score

Hi-index 0.00

Visualization

Abstract

Software tools for visualizing very large multidimensional databases have become increasingly important to discover interesting relationships among variables. While current tools implement operations such as drilling down, rolling up, and slicing data tables to help users notice interesting features of the data, the onus is on the user to choose the dimensions for drill down, or other operations. Expert knowledge is required to do this effectively and, since many users are novices, incorrect choices often lead to dead-ends, backtracking, confusion, and frustration. We suggest a novel approach to the selection of dimensions that relies on the interactive presentation of small multiples of thumbnail visualizations, before performing drill down or roll up operations. These previews of distributions, relationships, and associations, before variable selection, compel visual comparisons of change and difference, thus highlighting the options that are most likely to lead to productive paths.