Re: RDF named graph use case and requirement

On 22/09/2011 15:40, Satya Sahoo wrote:
> I am a bit confused here - originally Graham used Pra and Prb as "an RDF
> graph", so they can be treated as resources and we can make statements about
> it - creator and date. Also, the "event" will be the equivalent to a PE
> "statementMaking" and the output will be Pra and Prb.
>
> I agree with Sandro that Graham's original use of Pra and Prb maps them to
> g-snaps from RDF WG terminology. Looking at the definitions of g-box and
> g-snap, they almost seems to be a class-instance correspondence between
> them?

As far as I'm aware, there is no formal semantics for the notions of "g-box" and 
"g-snap", so I can't tell if that is true or not.

Strictly, in my original message, Pra and Prb are "RDF graphs" used as RDF graph 
*nodes* in the containing graph.  The denotations of those nodes then follows 
from the RDF semantics.  (http://www.w3.org/TR/2002/WD-rdf-mt-20020429/#urisandlit)

But _if_ Pra and Pra are graph *literal nodes* in the RDF, then by the RDF 
semantics they denote the corresponding graph literals and nothing else (*). 
Thus, if they are the same graph literal, they denote the same thing.

[(*) Or, strictly, if Datatyped literals are considered, a fixed mapping from 
the literal string to some value, but the conclusion still fllows.]

(See also my response to Luc sent to the prov WG list)

#g
--

Received on Friday, 23 September 2011 07:13:44 UTC