First steps in the computation of root justifications

Abstract

Description Logics (DLs) are widely accepted as an appropriate class of knowledge representation languages to represent and reason about ontologies. Tools for performing standard reasoning tasks such as satisfiability and consequence checking have grown increasingly powerful and sophisticated in the last decade. In section 2 we review the standard approach for the resolution of modelling errors encountered in ontologies and propose the first steps in a new method for resolving such errors in section 3. The method is based on the notion of root justifications which we define and discuss. The approach we describe is applicable to a wide class of DLs. We don’t provide a comprehensive formal introduction to DLs, but rather point the reader to the book by Baader et al.