Interactive Construction of Classification Rules

  • Authors:
  • Jianchao Han;Nick Cercone

  • Affiliations:
  • -;-

  • Venue:
  • PAKDD '02 Proceedings of the 6th Pacific-Asia Conference on Advances in Knowledge Discovery and Data Mining
  • Year:
  • 2002

Quantified Score

Hi-index 0.01

Visualization

Abstract

We introduce an interactive classifier construction system, CVizT, in which the entire process is visualized based on a multidimensional visualization technique, Table Lens. The CVizT system is a fully interactive approach. The appropriate visualization-based interaction capabilities are provided for the user to include human perception into the construction process. Our experiments with data sets from the UCI repository demonstrates that the CVizT system is straightforward and easily learned. The user's preference and domain knowledge can also be integrated into the construction process.