A visual interface technique for exploring OLAP data with coordinated dimension hierarchies

  • Authors:

  • Mark Sifer

  • Affiliations:

  • Sydney University, Australia

  • Venue:
  • CIKM '03 Proceedings of the twelfth international conference on Information and knowledge management
  • Year:
  • 2003

Quantified Score

Hi-index 0.00

Visualization

Abstract

Multi-dimensional data occurs in many domains while a wide variety of text based and visual interfaces for querying such data exists. But many of these interfaces are not applicable to OLAP, as they do not support use of dimension hierarchies for selection and aggregation. We introduce an interface technique which supports visual querying of OLAP data, that has been implemented in the SGViewer tool. It is based on a data graph rather than a data cube representation of the data. Our interface presents each dimension hierarchy in a zoomable panel which supports selection and aggregation at multiple levels. Users explore data and query it by making selections in several dimension views. Three view coordinations are identified; progressive, global and result only. Our main contribution, the progressive view coordination provides better support for query refinement than existing interfaces, by helping users decide the next query step with intermediate result overviews, and by helping users change a previous selection decision with retained selection context views. Our interface technique is demonstrated with a web log dataset of visits organised into time, download, visitor and referrer address dimensions.