Incremental Transformation of Lattices: A Key to Effective Knowledge Discovery

  • Authors:
  • John L. Pfaltz

  • Affiliations:
  • -

  • Venue:
  • ICGT '02 Proceedings of the First International Conference on Graph Transformation
  • Year:
  • 2002

Quantified Score

Hi-index 0.00

Visualization

Abstract

We show that a formal concept lattice L, with explicit generators, constitutes a viable medium for discrete, deterministic, data mining. All implications of interest can be found and extracted from L independent of the frequency of their occurance.More importantly, we show that these lattices can be grown from a binary relation R one row, or observation, at a time using a transformation that is based on the mathematical properties of the generators and faces of closed sets. Incremental growth is orders of magnitude faster than existing methods which rely on global closure operators.