Capturing Knowledge Representations Using Semantic Relationships An Ontology-based Approach

Knowledge representations in the scope of this work are a way to formalize the content of documents using dependent metadata i.e. words in document. One of the challenges relates to limited information that is presented in the document. While past research has made use of external dictionaries and topic hierarchies to augment the information, there is still considerable room for improvement. This work explores the use of complex relationships (otherwise known as Semantic Associations) available in ontologies with the addition of information presented in documents. In this paper we introduce a conceptual framework and its current implementation to support the representation of knowledge sources, where every knowledge source is represented through a vector (named Semantic Vector SV). The novelty of this work addresses the enrichment of such knowledge representations, using the classical vector space model concept extended with ontological support, which means to use ontological concepts and their relations to enrich each SV. Our approach takes into account three different but complementary processes using the following inputs: (1) the statistical relevance of keywords, (2) the ontological concepts, and (3) the ontological relations. Keywords-Information Retrieval; Ontology Engineering; Knowledge Representation