Search

ORPHANET is a reference information portal on rare diseases and orphan drugs for healthcare professionals and general audiences. After ten years of evolution, the current ORPHANET tools and knowledge representation cannot represent appropriately the constantly evolving knowledge on rare diseases. ORPHANET knowledge base has evolved from a simple thesaurus to a multi-classification terminology over the years, without refactoring the knowledge organization from the top (using a metamodel or/and an ontology). We propose in this paper to review the knowledge organization of ORPHANET by introducing a core ontology for rare diseases that has the specificity to also model classifications. We are conducting research to build and use a rare diseases knowledge base in an ontology-based architecture that complies with the W3C standards of the semantic web : OWL, RDF, SPARQL and SKOS. Using a specific knowledge cycle, we propose new edition, validation and sharing processes for rare diseases knowledge in ORPHANET. We show in this paper that ontologies are designed to manage the generation of multi-classifications into complex knowledge organization systems (KOS). We also demonstrate that the introduction of formal knowledge representation systems (OWL) improved ORPHANET knowledge base quality. This experiment highlights a continuity in the use of different knowledge organization systems. Nevertheless, the complex knowledge curation of this domain involves a formalization that can be appropriately supported only by the use of rules and an ontology.