RE: abstract model

Brian,

While I find your approach to reification very attractive, I think I am 
forced to conclude that it represents an *extension* to the RDF model as 
defined.  Specifically, the "mapping called Reification" is a new concept 
that is not present in the original, and the only way to find that mapping 
using the original model is to find the set of 4 statements that comprise 
the reification.

I was looking at your JENA interfaces earlier today.  I think it's quite 
legitimate to use the statement resource as a 'handle' to the reification, 
and that creates a legitimate RDF model.  But if the reification is 
incomplete then theres no (safe) way to recognize it in a "raw" set of triples.

#g
--

At 01:10 PM 9/12/00 +0100, McBride, Brian wrote:
> > > 6.  There is a mapping called Reification which a maps
> > >     each member s of Statements onto a unique member r
> > >     of Resources.  r is known as the reficiation of s.
> > >     Here unique means given s1 and s2 members of
> > >     Statements, Reification(s1) = Reification(s2) iff
> > >     s1 = s2.
> >
> > I don't read the RDF spec to imply this. Instead something like:
> >
> >       there is a relationship Reifies over {(r,s)} where
> >       r Reifies s (wrt a model m)
> >               iff
> >       m contains the statements
> >               r -[rdf:subject]-> (subject(s))
> >               r -[rdf:predicate]-> ...etc
> >
> > in other words, "a reification" instead of "the reification" is the
> > right way to look at this.
>
>You can certainly look at it as you suggest.  Lets say I do that
>and I have a model which contains only the reification of some
>statement s by your definition.  I now delete - lets say the
>statement with rdf:subject property.  It is arguable that the
>resource which is the subject of the other statements is still
>the same resource and still denotes the same thing it always did.
>
>The reason I chose to represent it the way I did, was to
>move the definition of what properties a reified statement must
>have from the abstract model to the definition of the transform
>between a serialization and an abstract model.  That means that
>a parser can essentially choose what statements to add when it
>reifies, and in practise I think that is a good thing.  Many
>times one just not want all those extra subject, property
>statements included.
>
>Brian

------------
Graham Klyne
(GK@ACM.ORG)

Received on Tuesday, 12 September 2000 11:56:47 UTC