Inconsistent Ontology Handling by Translating Description Logics into Defeasible Logic Programming

SERGIO GÓMEZ; CARLOS IVÁN CHESÑEVAR; GUILLERMO SIMARI,

INTELIGENCIA ARTIFICIAL. IBERO-AMERICAN JOURNAL OF ARTIFICIAL INTELLIGENCE

AEPIA

Lugar: Madrid; Año: 2007 p. 11 - 22

1137-3601

The emph{Semantic Web} is a future vision of the webwhere stored information has exact meaning, thus enabling computers tounderstand and reason on the basis of such information.Assigning semantics to web resources is addressed by means of emph{ontology definitions} which are meant to be writtenin an emph{ontology description language} such as OWL-DL that is based on so-called Description Logics (DL).Although ontology definitions expressed  in DL can be processed with existing DL reasoners, such DL reasoners are incapable of dealing with emph{inconsistent} ontology definitions. Previous research has determined that a subset of DL canbe effectively translated into an equivalent subset of logic programming.We propose a method for dealing with inconsistent ontology definitions in the Semantic Web. Our proposal involves mapping DL ontologies into DeLP programs. That is, given an OWL-DL ontology Oa_{Owl}, an equivalent DL ontology Oa_{DL} can be obtained. Provided Oa_{DL} satisfies certain restrictions, it can be translated into a DeLP program Oa_{DeLP}.Therefore, given aquery Q w.r.t Oa_{Owl}, a dialectical process will be performedto determine if Q is warranted w.r.t. Oa_{DeLP}.