The EL is a tractable family of lightweight description logics that underlay the OWL EL profile. It guarantees the tractability of the reasoning process, especially for concept classification. In particular, such a fragment is widely used for medical applications. This paper investigates the evolution of EL ontologies when a new piece of information that can be conflicting or attached with a confidence level reflecting its credibility or priority is available. To encode such knowledge, we propose an extension of EL description logic within the possibility theory, which provides a natural way to deal with ordinal scale reflecting ranking between pieces of knowledge. We then show how such a ranking between axioms is induced from the ontology with the presence of new information and study the evolution process at the semantic level. Finally, we propose a polynomial syntactic counterpart of the evolution process while preserving the consistency of the ontology.
In this paper, we place ourselves in the context of inconsistency-tolerant query answering over lightweight ontologies, which aims to query a set of conflicting facts using an ontology that represents generic knowledge about a particular domain. Existing inconsistency-tolerant semantics typically consist in selecting some of (maximal) consistent subsets of facts, called repairs. We explore a novel strategy to select the most relevant repairs based on the stratification of the assertional base into priority levels that we automatically induce from the ontology. We propose a method that exploits conflict statistical regularities between facts to induce an embedding, in which each fact is represented by a vector. Based on Euclidean distances between facts, we classify the assertions from the most reliable to the least important ones. We then use these distances to define relevant repairs. Interestingly enough, we show that the obtained repair is done in polynomial time.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.