W3C home > Mailing lists > Public > www-rdf-interest@w3.org > June 2003

RE: [Q] how to reference a rdf schema in an rdf instance file ?

From: David Menendez <zednenem@psualum.com>
Date: Thu, 5 Jun 2003 15:13:43 -0400
Message-Id: <a05210601bb05437d7db7@[]>
To: <danny666@virgilio.it>, <hgadm@cswebmail.com>, <www-rdf-interest@w3.org>

At 1:44 PM +0200 2003-06-03, Danny Ayers wrote:
>  > how do I reference a rdf schema that defines a certain
>  > rdf instance file within that instance file ?

You might do something like:

<rdf:Description about="">
     <dcmitype:Dataset about="http://example.com/schema-location">

(given the usual declarations for rdf:, rdfs:, and dc: and 
<http://purl.org/dc/dcmitype/> for dcmitype:)

This has the general meaning of "for more information about this 
resource, see this particular RDF/XML file".

This could be more succinct with some new vocabulary, for example:

<rdf:Description about="">
   <eg:seeSchema resource="http://example.com/schema-location"/>

eg:seeSchema may or may not be equivalent to owl:import. I haven't 
read the OWL documentation recently, but I think the domain of 
owl:import is owl:Ontology.

In any case, importing or referencing a schema document is likely to 
require fewer triples than referencing defining documents for each 
property and class. For vocabularies which are commonly used 
together, we could even have individual schemas that collect other 

On the other hand, in the long run it's probably better to have 
services like URIQA that you can ask "what's the deal with this 
property?" without depending on the author to include declarations.

>  > BTW, does rdf provide for the inclusion of XML schema
>>  files
>  > or do I need DTDs ?
>Inclusion in what sense? Remember that RDF isn't XML, just RDF/XML is just
>one serialization of the model, so the constructs of XML schema aren't
>necessarily appropriate. RDF can use values given as XML schema datatypes
>[2], but XML schema validation on RDF documents doesn't really work because
>the RDF semantics work on a different level. DTDs are a non-starter because
>of the alternate representations (full, abbreviated, a mix) possible for the
>XML representation.

This is something of a tangent, but I have been doing some 
experiments with defining subsets of RDF/XML which can be validated 
by an XML schema (although not using XML Schema; I find Relax NG much 
easier to use).

For example, I've been playing around with a format for describing 
the various syndicated feeds associated with a site. It's defined as 
XML to make life easier for non-RDF folks, but the definition is such 
that any valid document is also valid RDF/XML.


>[2] http://www.w3.org/TR/rdf-concepts/#section-use-xsd

Dave Menendez - zednenem@psualum.com - http://www.eyrie.org/~zednenem/
Received on Thursday, 5 June 2003 15:12:46 UTC

This archive was generated by hypermail 2.4.0 : Friday, 17 January 2020 22:44:42 UTC