Enhancing Ontology by Integrating Facts from Unstructured Data and Mapping with Linguistic Knowledge

Abstract

Ontology plays a crucial role in the growth of semantic web by providing a domain- knowledge resource comprehensible to both humans as well as computers. Knowledge resources can be of two types-taxonomical knowledge supporting description logic in the form of ontology representing a particular domain and linguistic knowledge. An approach is presented in this paper for constructing a third type of knowledge resource(Hybrid) that accommodates variations in the structure as per RDF triple extracted from unstructured corpus and subsequently semantically associating ontological concepts with their linguistic counter-part. The purpose of this knowledge base is to add linguistic expressiveness to conceptual knowledge and to accommodate the dynamic nature of knowledge. Semantic similarity and term coverage are computed between terms(concepts). A case study i.e enhancing a dengue ontology by incorporating RDF triple extracted from patients’ case sheets and semantically associating concepts with WordNet by using similarity and relatedness measures is presented.