Evaluate the owlready2 approach

@St3ff3nBusch has proposed an alternative approach to instantiate rdfs/owl classes in Python: https://git.rwth-aachen.de/nfdi4earth/softwaretoolsarchitecture/kh-populator/-/blob/8e33a569271bb8a9ccce15e19cf257b0529d6b67/kh_populator/pipelines/harvest_incubators.py

The approach loads the experimental OWL port of the schema.org (rdfs) ontology (available here) with owlready2. If we expand on this (combining with other vocabularies in use in the KH), it would basically replace the kh_populator_model module, so that we do not have to curate it manually and check that it is in sync with the official vocabulary and our KH data model/ schema.