Closed sennierer closed 1 year ago
I think the issue is that we don't have an ontology/definitions for roles.
In BS data we have:
<http://www.intavia.eu/bs/personproxy/3304> bioc:has_person_relation <http://www.intavia.eu/bs/personrelation/1/3304> .
<http://www.intavia.eu/bs/personrelation/1/3304> a <http://www.intavia.eu/familyreltype/parent> .
What we miss is this triple:
<http://www.intavia.eu/familyreltype/parent> rdfs:subClassOf bioc:Person_Relationship_Role
How are other datasets modeling the parent
role? Do we use a shared URI for the role (if data would be perfectly harmonized, we should!)? Should we create e.g. roles.ttl
that defines the roles (i.e. states rdf:type, rdfs:subClassOf and rdfs:label for them)?
hmm...to fulfill the shacl validation I added APIS_role rdfs:subClassOf bioc:Person_Relationship_Role
triple. But you are obviously right, we should have a common class that we inherit from.
And I see we already created something like that here https://github.com/InTaVia/idm-rdf/blob/main/idm-OWL/intavia_idm_classes_1.ttl ;) Should we add the triples there?
Yes, I think that https://github.com/InTaVia/idm-rdf/blob/main/idm-OWL/intavia_idm_classes_1.ttl would be good for adding the role classes.
@ptleskin will add the role class definitions in the file https://github.com/InTaVia/idm-rdf/blob/main/idm-OWL/intavia_idm_classes_1.ttl with a SPARQL CONSTRUCT based on used classes in IKG.
Moved the new new BS file to the data-source repo, created a PR and the automatic shacl validation shows that the new ttl is missing the
bioc:Person_Relationship_Role
class for the roles. Can you have a look at the report, see if that makes sense and change accordingly if it does? Btw.: improvements to the shacl constraints are obviously very welcome.