@misc{SvenssonKasprzik2019, author = {Svensson, Lars G. and Kasprzik, Anna}, title = {Extracting Additional Semantics from the GND Using OWL Reasoning: A Possible Approach}, series = {TK 3: Content kuratieren / GND: Arbeit mit Normdaten (20.03.2019, 14:00 - 17:30 Uhr, Saal 4)}, journal = {TK 3: Content kuratieren / GND: Arbeit mit Normdaten (20.03.2019, 14:00 - 17:30 Uhr, Saal 4)}, year = {2019}, abstract = {The libraries in the German-speaking part of the world jointly maintain the Integrated Authority File (Gemeinsame Normdatei, GND), a highly interlinked collection of authority records covering people, families, corporate bodies, events, topical terms, etc. The RDF representation of the GND is accompanied by the GND Ontology (GNDO), an RDF element set declaring OWL classes and properties specifically developed to match the entity types and relation codes used in the GND. The GNDO, however, only mirrors the semantics that is already contained explicitly in the GND data model and - despite being modelled using OWL - does not make use of the full range of possibilities offered by a more formal, axiomatized representation of the semantic structure of the GND. Preliminary results have shown that an axiomatization of the GNDO that goes beyond simple subclass and subproperty relations and that adds stronger semantic constraints than the mere specification of domain and range can provide added value both by finding quality issues such as logical inconsistencies in the GND and by adding otherwise implicit axioms to the overall data model, thus creating a more complete knowledge graph. This presentation contributes - a proposal for a stronger axiomatization of the GNDO, and, resulting from this - examples of implicit knowledge to be gained from such an enhancement - logical inconsistencies in the underlying data that can be found by an OWL reasoner.}, language = {de} }