From metadata application profiles to semantic profiling: ontology refinement and profiling to strengthen inference-based queries on the semantic web

Ontologies on the Semantic Web form a basis for representing human-conceivable knowledge in a machine-understandable manner. Ontology development for a specific knowledge domain is, however, a difficult task, because the representation produced has to be adequately detailed and broad enough at the same time. The CIDOC-CRM is such an ontology, pertaining to cultural heritage, which we align to the Semantic Web environment: first, transforming it to OWL and then profiling it not in the usual flat metadata sense, but by refining and extending its conceptual structures, taking advantage of OWL semantics. This kind of profiling maintains applicability of the model, while enabling more expressive reasoning tasks. To this end, we construct a mechanism for acquiring implied and web-distributed information that is used to conduct and present a series of experimental inferences on the CRM-profiled form.