W3C home > Mailing lists > Public > public-owl-dev@w3.org > July to September 2007

Re: declaredAs

From: Dan Brickley <danbri@danbri.org>
Date: Sat, 11 Aug 2007 10:52:58 +0100
Message-ID: <46BD86FA.6030607@danbri.org>
To: Pat Hayes <phayes@ihmc.us>
CC: John McClure <jmcclure@hypergrove.com>, Bijan Parsia <bparsia@cs.man.ac.uk>, Owl Dev <public-owl-dev@w3.org>

Pat Hayes wrote:
> 
>>
>>> But rdf:ID doesn't get you *any sort of
>>> declaration*. It's *just a funny way of making an assertion*.
>>
>> Nope, I don't agree -- I don't see this in the specs anywhere.
> 
> Well, the RDF specs say explicitly that RDF consists entirely of 
> assertions, they give a normative semantics which defines the 
> assertions, and they do not mention declarations anywhere. Seems pretty 
> damn clear to me.

Yup, if you're looking for "declarations", you might try to see it 
social wrappings around pure formal RDF. Like if I use RDFS/OWL to 
describe some classes and properties, then write that out as an RDF/XML 
UTF-8 document and publish it, then mail semantic-web@w3.org and say ... 
'hey everyone, use my new ontology'. Perhaps after all that we could say 
the classes and properties have been declared? RDF no more has 
declarations than it has lies; but people can lie and declare using RDF.

<rdf:Property rdf:ID="permits"/> in an RDF/XML document 
http://example.com/somerdf ...might be read as somehow ensuring that the 
URI for the term "permits" that it expands to is new, previously unused, 
etc. And I guess this is why it gets seen as a declaration, since it 
gives (illusory) impression we can be sure that whoever wrote this is 
the first to do so. That assumptio is a mistake, since the interaction 
with xml:base allows pre-existing properties and classes to be further 
described via rdf:ID.

cheers,

Dan

ps. testcase:

<rdf:RDF xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
   xml:base="http://creativecommons.org/ns#"
   xmlns:dc="http://purl.org/dc/elements/1.1/"
 >
<rdf:Property rdf:ID="permits"
dc:description="assertions about cc:permits go here, but if anyone 
'declared' it, it wasn't me."/>
</rdf:RDF>

gives 2 triples:

http://creativecommons.org/ns#permits 
http://www.w3.org/1999/02/22-rdf-syntax-ns#type 
http://www.w3.org/1999/02/22-rdf-syntax-ns#Property .

http://creativecommons.org/ns#permits 
http://purl.org/dc/elements/1.1/description 	"assertions about 
cc:permits go here, but if anyone 'declared' it, it wasn't me." .
Received on Saturday, 11 August 2007 09:53:17 GMT

This archive was generated by hypermail 2.3.1 : Wednesday, 27 March 2013 09:32:54 GMT