An Algorithm for Knowledge Base Extraction

Many approaches have been proposed for reasoning based on conflicting information in general and in particular on stratified knowledge bases, i.e. bases in which all pieces of information are assigned a rank. In this paper, we want to reflect on a particular family of algorithmic approaches known as Adjustments, which have been suggested for extracting a consistent knowledge base from a possibly inconsistent stratified one. We will point out counter-intuitive results provided by these approaches and develop an algorithm we call Refined Disjunctive Maxi-Adjustment which does not have these drawbacks.

• Type: Conference
• Year: 2004
• Where: ECAI
• Authors: Alexander Nittka
• Keywords: Artificial Intelligence, ECAI 2004, Knowledge Base, Many Approaches, Stratified Knowledge Bases