A Tableau Algorithm for Handling Inconsistency in OWL

Abstract

In Semantic Web, the knowledge sources usually contain inconsistency because they are constantly changing and from different view points. As is well known, as based on the description logic, OWL is lack of the ability of tolerating inconsistent or incomplete data. Recently, the research in handling inconsistency in OWL becomes more and more important. In this paper, we present a paraconsistent OWL called quasi-classical OWL to handle inconsistency with holding important inference rules such as modus tollens, modus ponens, and disjunctive syllogism. We propose a terminable, sound and complete tableau algorithm to implement paraconsistent reasoning in quasi-classical OWL. In comparison with other approaches to handle inconsistency in OWL, our approach enhances the ability of reasoning by integrating paraconsistent reasoning with important classical inference rules.

Robust and scalable Linked Data reasoning incorporating provenance and trust annotations

“…There are a variety of other approaches for handling inconsistencies in OWL ontologies, including works on paraconsistent reasoning using multi-valued, probabalistic, or possibilistic appraoches, by, e.g., Ma et al [52], Zhang et al [67], Huang et al [39], Qi et al [54], etc.…”

Scalable OWL 2 Reasoning for Linked Data

“…Given that consistency cannot be expected on the Web, we wish to avoid the arbitrary entailment of all possible triples from our knowledge-base. Along these lines, a number of paraconsistent reasoning approaches have been defined in the literature (see, e.g., [46,60,96,61,56]) typically relying upon four-valued logic [9]however, again, these approaches have yet to demonstrate the sort of performance required for our scenario.…”

Robust and Scalable Linked Data Reasoning Incorporating Provenance and Trust Annotations

“…There are a variety of other approaches for handling inconsistencies in OWL ontologies, including works on paraconsistent reasoning using multi-valued, probabalistic, or possibilistic appraoches, by, e.g., Ma et al [52], Zhang et al [67], Huang et al [39], Qi et al [54], etc. 52 However, all such approaches are again based on Description Logics formalisms, and only demonstrate evaluation over one (or few) ontologies containing in the order of thousands, tens of thousands, up to hundreds of thousands of axioms.…”