Merging Locally Correct Knowledge Bases: A Preliminary Report

Abstract

Belief integration methods are often aimed at deriving a single and consistent knowledge base that retains as much as possible of the knowledge bases to integrate. The rationale behind this approach is the minimal change principle: the result of the integration process should differ as less as possible from the knowledge bases to integrate. We show that this principle can be reformulated in terms of a more general model of belief revision, based on the assumption that inconsistency is due to the mistakes the knowledge bases contain. Current belief revision strategies are based on a specific kind of mistakes, which however does not include all possible ones. Some alternative possibilities are discussed.