Re: PROV-ISSUE-144 (Tlebo): how is "reserved attribute 'type'" related to rdf:type? [Data Model]

My 2c:

I understand the data model to be independent of any particular serialization or 
representation model, including RDF.  As such, it would be inappropriate for DM 
to specify that type == rdf:type.

The mapping from DM to RDF, as specified by the ontology, would be the 
appropriate place for such a correspondence to be asserted.

#g
--

On 06/11/2011 01:32, Provenance Working Group Issue Tracker wrote:
>
> PROV-ISSUE-144 (Tlebo): how is "reserved attribute 'type'" related to rdf:type? [Data Model]
>
> http://www.w3.org/2011/prov/track/issues/144
>
> Raised by: Timothy Lebo
> On product: Data Model
>
> The reserved attribute "type" is used a few places, but with no mention to its relation to rdf:type. This becomes particularly concerning when "subtyping" is mentioned. Should we avoid reinventing the wheel? Can we say somewhere that "type" SHOULD BE considered as rdf:type, and RDFS should be used to handle subtyping semantics and inference?
>
>
>
> 6. PROV-DM Extensibility Points
>
> "Subtyping is allowed by means of the reserved attribute type."
>
>
>
>

Received on Thursday, 10 November 2011 16:16:40 UTC