Re: Question on semantics of reified statements

Bob MacGregor wrote:

>
>
> After posting a example to the Jena e-mail list,
> I realized that there is a deeper issue, relating
> to the semantics of RDF reified statements, that
> I don't know the answer to.  Here is the example:
>
>
> "John states that (a ship) s1 visited Antwerp
> on March, 2003."
>
> "Sue states that s1 visited Antwerp
> on April, 2003."
>
> Below is a set of RDF statements intended to represent the above two
> sentences.  I'm trying to keep things simple -- I've left out the
> assertion that 's1' is a ship, and I'm using an n3-like notation that
> omits some namespaces:
>
> st1 rdf:type rdf:Statement.
> st1 rdf:subject s1.
> st1 rdf:predicate location.
> st1 rdf:object antwerp.
> st1 beginDate "March 2003".
> st1 endDate "March 2003".
> st1 author john.
>
> st2 rdf:type rdf:Statement.
> st2 rdf:subject s1.
> st2 rdf:predicate location.
> st2 rdf:object antwerp.
> st2 beginDate "April 2003".
> st2 endDate "April 2003".
> st2 author sue.
>
> My intent is that 'st1' and 'st2' are both reifications of the
> same statement   [s1 location antwerp].  My question is, are
> st1 and st2 the same resource, or are they two distinct
> resources? 


Whatever you want it to be - see the discussion of the subject of 
reification in http://www.w3.org/TR/rdf-primer/#reification

It sounds like you'd be better modelling in OWL, or something similar, 
what you are trying to model here would seem to fit more naturally 
there, and it's a simple process to generate the rdf:triples from OWL 
descriptions, in fact that is the only useful way to process OWL/XML 
(that I can see).


>
>
> Cheers, Bob
>
> P.S., I don't consider the above encoding to be a particularly good
> use of RDF reification, but that's not germane to my question
> in this posting.
>

Received on Tuesday, 9 September 2003 20:26:48 UTC