Semantic document indexing in ontology-driven organizational memories

Abstract

Effective document retrieval using domain knowledge and semantics is one of the major challenges in Information Retrieval. Over the last years, there has been a growing interest in ontologies as an artifact for human knowledge representation and a critical component in Knowledge Management, Semantic Web, and Business-to-Business applications. We have found that it is not easy to represent certain types of knowledge (skills or procedures) or to transform certain types of knowledge representation (knowledge contained in diagrams) into an appropriate ontological format. To overcome this problem, our proposal is to connect knowledge sources to the domain ontology associated with an Organizational Memory without forcing any transformation in the structure of the source itself. This connection will allow the semantic classification of knowledge sources so that when a user performs a query it is possible to recover the documents that have a higher probability of containing the answer.