Keeping Researchers Updated by Automatically Enriching an Ontology in the Medical Field.

Ontologies provide a common and standard dictionary of terms in some domain for researchers to easily exchange data. Forming an ontology requires several years of work performed by human experts, and an ontology for a given domain is thought to be stable for many years. Nevertheless, as scientific articles are continuously published to gather knowledge on recent findings, existing ontologies risk becoming stale, or require further human effort. Moreover, searching new articles without referring to an ontology can be very time consuming and confusing, especially for novice researchers. We propose an approach for automatically relating newly available published articles to existing ontologies. By automatically selecting relevant scientific articles and making them appear besides other data in an ontology, we aim at supporting experienced and novice researchers. Therefore, as knowledge grows and articles are available, the ontology used by researchers will also be automatically connected, allowing them to readily discover new findings. To validate the effectiveness of the proposed approach, we have enriched OBIB, an ontology for biobanking, with a selection of articles extracted from PubMed. The approach is general enough and can be applied to other ontologies or publishers.