Review membership predicate (Collection in Dataset and Feature in Collection)

[ldesousa]

Prez has recently dropped the dcterms:isPartOf predicate as semantic conver of membership. It is now using rdfs:member, following the recent update to GeoSPARQL. Since this profile is meant to guide the creation of datasets usable by Prez (and earlier ogcldapi), a review/update is in order.

Considering how the profile is currently relying exclusively on DCTerms, I would rather leave it as it is and adapt the software instead. That would be a semantically sounder option in my view. But the important thing is to keep all these different assets in synchronicity.

[ldesousa]

@recalcitrantsupplant thank you for updating the profile. There is still the lingering issue with the semantics of rdfs:member, please recall the discussion yesterday, this predicate translates to "has member" in natural language. It is not intuitive and for sure it confounded me.

In principle, the rdfs:member predicate cannot be used to fulfill the requirements Requirement D-member and Requirement FC-member. In this case you might need to re-introduce dcterms:isPartOf.

In any event, with the requirements Requirement FC-memberof and Requirement F-memberof it should now be possible to get a compliant knowledge graph.