- From: Raphaël Troncy <raphael.troncy@eurecom.fr>
- Date: Tue, 07 Jan 2014 14:27:31 +0100
- To: Ghislain Atemezing <auguste.atemezing@eurecom.fr>
- CC: Kostis Kyzirakos <Kostis.Kyzirakos@cwi.nl>, Frans Knibbe | Geodan <frans.knibbe@geodan.nl>, LocAdd W3C CG Public Mailing list <public-locadd@w3.org>
> So we can similarly say for locn, as you suggested something like: > > locn:identifier a owl:DatatypeProperty, rdf:Property ; > rdfs:range [a rdfs:Datatype ; > owl:unionOf (xsd:URI rdfs:Literal) .] ; > Does it make sense? No. If you want that this URI to be interpreted as a URI and not as a literal, then the identifier property cannot be a owl:DatatypeProperty. It is an owl:ObjectProperty by definition. Since we want an unconstrained range, we are out of OWL anyway, so the property will just be a rdf:Property. > And add to the appropriate class in locn vocab an axiom like the one stated in org like: > locn:aClass owl:hasKey (locn:identifier) ; * What would be the owl class: locn:Location? locn:Geometry? Something else? * Why would you like to make this property a key? This prevents to have two geographic identifiers for the same object which sorts of ruins the interoperability effort we are trying to achieve with this property. I'm afraid I don't understand your proposal. Raphaël -- Raphaël Troncy EURECOM, Campus SophiaTech Multimedia Communications Department 450 route des Chappes, 06410 Biot, France. e-mail: raphael.troncy@eurecom.fr & raphael.troncy@gmail.com Tel: +33 (0)4 - 9300 8242 Fax: +33 (0)4 - 9000 8200 Web: http://www.eurecom.fr/~troncy/
Received on Tuesday, 7 January 2014 13:28:00 UTC