Table 3 Qualitative comparison between GalenOWL and GalenDrools

Structured knowledge representation

Yes, ontology based.

Partial, relational DB.

Medical knowledge integration and reusability

Hierarchical class relationships (ICD10, UNII, ATC) and definition of Conditions are expressed using OWL expressivity. They can be utilized by any OWL reasoner.

ATC, UNII, ICD10 entities relationships and Conditions are materialized inside rule expressions. Materialization is specific to the rule language used.

Knowledge sharing

Ontology can be published and accessed through SW technologies, e.g. as a SPARQL endpoint.

Queries to DB have to follow the DB schema.

Rule expression

Rules for drug recommendations directly express pharmaceutical knowledge and can be immediately loaded to a reasoner.

Rules express pharmaceutical knowledge but have to be post processed, in order to materialize entities relationships before loading them to the rule engine.