Machine Learning

, Volume 37, Issue 2, pp 143–181

Effective and Efficient Knowledge Base Refinement

  • Leonardo Carbonara
  • Derek Sleeman

DOI: 10.1023/A:1007661823108

Cite this article as:
Carbonara, L. & Sleeman, D. Machine Learning (1999) 37: 143. doi:10.1023/A:1007661823108


This paper presents the STALKER knowledge base refinement system. Like its predecessor KRUST, STALKER proposes many alternative refinements to correct the classification of each wrongly classified example in the training set. However, there are two principal differences between KRUST and STALKER. Firstly, the range of misclassified examples handled by KRUST has been augmented by the introduction of inductive refinement operators. Secondly, STALKER's testing phase has been greatly speeded up by using a Truth Maintenance System (TMS). The resulting system is more effective than other refinement systems because it generates many alternative refinements. At the same time, STALKER is very efficient since KRUST's computationally expensive implementation and testing of refined knowledge bases has been replaced by a TMS-based simulator.

knowledge base refinement theory revision knowledge acquisition truth maintenance dependency networks expert systems 

Copyright information

© Kluwer Academic Publishers 1999

Authors and Affiliations

  • Leonardo Carbonara
    • 1
  • Derek Sleeman
    • 2
  1. 1.BT UK Markets, British TelecomLondonUK
  2. 2.Department of Computing Science, King's CollegeUniversity of AberdeenAberdeenUK

Personalised recommendations