- From: Kingsley Idehen <kidehen@openlinksw.com>
- Date: Sat, 07 Jul 2007 23:57:10 -0400
- CC: Linking Open Data <linking-open-data@simile.mit.edu>, www-tag@w3.org, semantic-web@w3.org
Kingsley Idehen wrote: > Kingsley Idehen wrote: >>> >> Depends on the characteristics of the property "owl:sameAs" as used >> in Tim's "Universe of Discourse" (Data Space) and here is why: >> >> Yes, if in Tim's Data Space Definition he asserts that "owl:sameAs" >> is an "owl:SymmetricProperty" then his usage is unambiguous, as this >> should result in identical triples when you SPARQL against his Data >> Space using either URI (assuming we are not modeling the Web as a >> single Graph). Of course, this assumes the SPARQL processor has >> inference capabilities. Otherwise, "rdfs:seeAlso" would be the >> preferred property as this is unambiguous (i.e. the aforementioned >> URIs will produce varied Triples without introducing confusion). > In retrospect, the statement above is *incorrect* since "owl:sameAs" > means the URIs in the Domain and Range are to be treated as > "Identical". It's up to RDF Clients / Consumers (e.g. Browsers and > SPARQL Processors) to ensure that identical Triples are dereferenced > from said URIs. Thus, Tim's usage is unambiguous :-) > > > All, Missing from the above is a SPARQL example. Graph URI: http://www.w3.org/People/Berners-Lee/card#i Query1: select distinct ?s ?p ?o where {{<http://www.w3.org/People/Berners-Lee/card#i> ?p ?o } union {?s ?o <http://www.w3.org/People/Berners-Lee/card#i>} } Query2: select distinct ?s ?p ?o where {{<http://dbpedia.org/resource/Tim_Berners-Lee> ?p ?o } union {?s ?o <http://dbpedia.org/resource/Tim_Berners-Lee>} } If the SPARQL Processor supports inferencing then Queries 1 & 2 should produce identical results (Triples) for the Graph URI: http://www.w3.org/People/Berners-Lee/card#i -- Regards, Kingsley Idehen Weblog: http://www.openlinksw.com/blog/~kidehen President & CEO OpenLink Software Web: http://www.openlinksw.com
Received on Sunday, 8 July 2007 03:57:28 UTC