Re: In RDF what is the best practice to represent data provenance (source)?

On 21/01/07, Richard Cyganiak <> wrote:

> So, I agree with ChrisR: If you feel the need to make statements
> about relationships, then maybe the modelling is not adequate to your
> use case, and the relationship ought to be turned into a resource of
> its own. Some related advice is found in [1].

A caveat on that -  it's pretty easy to mess up the modelling with n-ary
relations such as when talking about physical quantities, as highlighted by
timbl in [2].

Seems to me there are a couple of closely related but different notions
getting slightly blurred in this thread:

saying things about the single statement { subject, predicate, object }
saying things about the predicate in { subject, predicate, object }

Whatever's said about the predicate will generally apply wherever the
predicate is used (as a coarse analogy, they behave like static variables in
OO languages), but that may not be a problem. I believe I'm ok in having:

    a pet:Cat ;
    foaf:name "Neo" ;
    x:anotherNameForNeo "Bouncer" .

- where that predicate is specialised to the extent that it will only be
used with a specific subject individual (I promise). In itself it's a bit
obscure to be much use, but if:

x:anotherNameForNeo rdfs:subPropertyOf foaf:name .

then it may still be useful information in the greater scheme of things.





Received on Sunday, 21 January 2007 12:54:52 UTC