Computer Science · Artificial Intelligence
Ontology-based Data Quality enhancement for Drug Databases
Olivier Curé, Jean‐Paul Giroud
See on the mapAbstract
This paper proposes a solution to enhance data quality features in drug databases. The method we are using is based on the integration of controlled terminologies for the drug domain. This approach requires several steps : (1) transforming the terminologies into a Semantic Web compliant Description Logics, namely OWL DL, (2) associating new axioms to concepts of these ontologies based on inductive reasoning on drug databases, and (3) assisting, using these OWL ontologies, our team of health care professionals in detecting and repairing data quality violations of selected features in drug databases. This last step aggregates drug products using ontology concepts and checks the characteristics of the products with the concept’s properties. We present a concrete example of this solution on a self-medication application using the European Pharmaceutical Market Research Association and Anatomical Therapeutical Chemical terminologies.