W3C home > Mailing lists > Public > public-rdf-wg@w3.org > December 2011

Re: Graph-State Resources

From: Kingsley Idehen <kidehen@openlinksw.com>
Date: Fri, 16 Dec 2011 21:00:52 -0500
Message-ID: <4EEBF7D4.1090604@openlinksw.com>
To: public-rdf-wg@w3.org
On 12/16/11 3:56 PM, Andy Seaborne wrote:
> INSERT INTO <http://www.w3.org/2000/10/swap/log.rdf>
> {?s rdfs:isDefinedBy <http://www.w3.org/2000/10/swap/log#> .
> <http://www.w3.org/2000/10/swap/log#>
> <http://open.vocab.org/terms/defines> ?s .
> <http://www.w3.org/2000/10/swap/log#> a owl:Ontology .
> ?s <http://www.w3.org/2007/05/powder-s#describedby>
> <http://www.w3.org/2000/10/swap/log.rdf>}
> FROM <http://www.w3.org/2000/10/swap/log.rdf>
> WHERE {?s a ?o  } 

Andy,

I guess my "SPARQL embellishment" comment threw this contribution off 
topic a little. I pasted the statement above as part of the discussion 
about named graphs. The statement is actually SPASQL that works inside 
our SQL query processor as well as via our SPARQL endpoints. Anyway, 
here is the gist of the what its about.

RDF Resource URL/Address: http://www.w3.org/2000/10/swap/log.rdf
Ontology URI/Name: http://www.w3.org/2000/10/swap/log#
Key Predicates:
1. http://www.w3.org/2007/05/powder-s#describedby -- for a relation that 
associates a URI with the address of a resource that bears the 
description of its referent
2. rdfs:isDefinedBy  -- for associating a Class or Property URI with the 
URI of the Ontology that defines its referent.

When the statement above is executed, the following happens:

1. HTTP GET against origin resource URL: 
http://www.w3.org/2000/10/swap/log.rdf, subject to cache invalidation 
scheme and pre existence of a named graph with that IRI

2. Additional triples added to the named graph due to missing 
isDefinedBy and wdrs:describedby relations in the origin resource 
(g-box) at: http://www.w3.org/2000/10/swap/log.rdf .

Looking at: 
http://lod.openlinksw.com/describe/?url=http%3A%2F%2Fwww.w3.org%2F2000%2F10%2Fswap%2Flog%23semantics 
, shows the result of the aforementioned actions.

I was hoping this would make the matter of named graphs a little clearer 
by walking through what's been outlined above. Basically, there's data 
at a location, its retrieved, and post-processed en route to a final 
representation of the description of a number of data items / objects / 
entities. Each endowed with a URI that resolves to the description of 
the URI's referent.

Caveat:

Which is the correct relation, bearing in mind Name / Address 
disambiguation:

?s <http://www.w3.org/2007/05/powder-s#describedby> 
<http://www.w3.org/2000/10/swap/log.rdf>

or

?s <http://www.w3.org/2007/05/powder-s#describedby> 
<http://www.w3.org/2000/10/swap/log.rdf#this> .

This is the riddle, as I see it. Is there a special case were 
Name/Address ambiguity is accepted such that object identity and object 
values can be conflated, I doubt it, but it begs the question.

I think:
?s <http://www.w3.org/2007/05/powder-s#describedby> 
<http://www.w3.org/2000/10/swap/log.rdf#this> is the cleaner relation.

-- 

Regards,

Kingsley Idehen	
Founder&  CEO
OpenLink Software
Company Web: http://www.openlinksw.com
Personal Weblog: http://www.openlinksw.com/blog/~kidehen
Twitter/Identi.ca handle: @kidehen
Google+ Profile: https://plus.google.com/112399767740508618350/about
LinkedIn Profile: http://www.linkedin.com/in/kidehen








Received on Saturday, 17 December 2011 02:02:12 GMT

This archive was generated by hypermail 2.3.1 : Tuesday, 26 March 2013 16:25:46 GMT