2016 · 15 citations · 13 references
ComputingComputer ScienceKnowledge RepresentationNatural Language ProcessingSemanticsConcept ModelingFormal OntologyOntology DesignOntology EngineeringOntological AlignmentOntologiesOntology AlignmentOntology Alignment AlgorithmsOntology MaintenanceOntology MatchingOntology VerificationSemantic Similarity
This paper presents a method to validate the insertion of a new concept in an ontology. This method is based on our previous works which add new concepts in a basic ontology using a general ontology (genaral ontology contains all the concepts of the basic ontology). To verify the semantic relevance of an ontology, we have proposed a method with three steps. First, we have found the neighborhood of the concept C in the basic ontology Ob and we store their semantic similarity values in a stack. The neighbourhood represents the concepts which are more similar to C in Ob. Secondly, we have assessed in the general ontology Og the semantic similarity between C and its neighbourhood found in the first step. Finally, we have evaluated the correlation between values found in the previous steps. We have considered the basic ontology as ontology with which we work and the general ontology as ontology used to align concepts with the basic ontology. The result obtained thanks to the method is a validated ontology after an update by adding a new concept. To illustrate our method, we have used the whole WordNet as the reference ontology and a branch of WordNet as basic ontology.
13
Verbs semantics and lexical selection
Zhibiao Wu, Martha Palmer · 1994 · 3.1K citations · Full text
Using Information Content to Evaluate Semantic Similarity in a Taxonomy
Philip Resnik · ArXiv.org · 1995 · 2.2K citations · Full text
Development and application of a metric on semantic nets
Roy Rada, Hafedh Mili, Ellen J. Bicknell et al. · IEEE Transactions on Systems Man and Cybernetics · 1989 · 1.8K citations