Resolution-Based Approximate Reasoning for OWL DL

Document Type

Conference Proceeding

Publication Date


Find this in a Library

Catalog Record


We propose a new technique for approximate ABox reasoning with OWL DL ontologies. Essentially, we obtain substantially improved reasoning performance by disregarding non-Horn features of OWL DL. Our approach comes as a side-product of recent research results concerning a new transformation of OWL DL ontologies into negation-free disjunctive datalog [1, 2, 3, 4], and rests on the idea of performing standard resolution over disjunctive rules by treating them as if they were non-disjunctive ones. We analyse our reasoning approach by means of non-monotonic reasoning techniques, and present an implementation, called SCREECH.


Presented at the 4th International Semantic Web Conference, Galway, Ireland, November 6-10, 2005.

Catalog Record