- From: aisaac via GitHub <sysbot+gh@w3.org>
- Date: Thu, 14 Feb 2019 00:29:49 +0000
- To: public-dxwg-wg@w3.org
@kcoyle it is easy to retrieve SKOS (transitive) B/N relationships, but it requires activating transitivity which may not be activated in every tool by default. That said, quite often schema inference (incl transitivity of rdfs:subPropertyOf) is not always activated either - for similar reasons. On the other hand PROF already relies on inference of transitivity, so I guess relying on it for the role hierarchy won't change the situation much. Extending by subproperties is harder than with SKOS, because rdfs:subPropertyOf has more precise (formal) semantics. So either it will be harder for implementers who want to do it carefully, or it won't be harder for implementers who don't care about misusing rdfs:subPropertyOf. I expect problems will especially occur when the classification of roles could become vague and roles could be classified under several categories. Multiple inheritance of properties is notoriously hard to handle in data modeling exercises - more than with classes I believe. Re your point on discussing use, I agree. But I can already anticipate that if we don't have much application scenarios right now for hierarchies, we won't get very far in terms of formal semantics and then opting for a solution that minimizes the semantic commitment for them (and thus opting for SKOS rather than RDFS/OWL) seems a wiser first choice. -- GitHub Notification of comment by aisaac Please view or discuss this issue at https://github.com/w3c/dxwg/issues/747#issuecomment-463433522 using your GitHub account
Received on Thursday, 14 February 2019 00:29:51 UTC