Re: [dcat] rdf graphs and documents

On 10-04-22 19:19, Cory Casanave wrote:
> RDF graphs are "documents".
>   

A bit of nuance here. RDF (named) graphs *can be thought of as*
documents and *can be used as* documents but need not necessarily be.
The fourth element in a quad store can be used in whatever way you like
- there is a school of thought (that I don't subscribe to) where the
fourth element is used as a unique ID for the triple for avoiding
reification (which some consider problematic using full reification).

I would be tempted to posit that in RDF a document is the result of
dereferencing a URI. That said, in our implementation, the result of
deferencing a URI is in fact a graph contained in the quadstore, but
again, this needn't necessarily be the case (could be the result of a
DESCRIBE query on the default (all) graph).

Related to this, I've been doing som work with changesets recently for
the quadstore beneath semantic.ckan.net, and I've found that I often
need to have a changeset that updates multiple graphs. Now the talis
changeset schema is good as far as it goes, and specifies changes with
reified triples. We have rdf:subject, rdf:predicate, rdf:object but
there is no such thing as rdf:graph to mention the fourth element. I've
invented an equivalent, but does anyone know if there is such a
predicate defined anywhere? Is it worth attempting to suggest an update
to the core rdf vocabulary to have this added (also with a commensurate
rdf:Graph class)?

We are lacking in tools for talking about graphs in rdf itself it seems...

Cheers,
-w

-- 
William Waites           <william.waites@okfn.org>
Mob: +44 789 798 9965    Open Knowledge Foundation
Fax: +44 131 464 4948                Edinburgh, UK

Received on Thursday, 29 April 2010 17:11:54 UTC