W3C home > Mailing lists > Public > www-rdf-interest@w3.org > November 2002

RE: RDF/XML without namespaces?

From: Danny Ayers <danny666@virgilio.it>
Date: Sat, 30 Nov 2002 14:46:52 +0100
To: "Ian Davis" <iand@internetalchemy.org>, <www-rdf-interest@w3.org>
Message-ID: <EBEPLGMHCDOJJJPCFHEFAELBIKAA.danny666@virgilio.it>

Interesting idea, but you will need one namespace to refer to your schema
(though if you (or your agent/app) are the only audience then you can of
course drop this).

There is also the issue of stating "interpret this as RDF" (using <rdf:RDF>
or the mime type). Though the XML syntax may be readable as valid RDF, there
is a bit of a grey area with the XML representation -> RDF model
transformation. In practice this shouldn't really be a problem, as it would
be fairly trivial to formally define a mapping between elements of a
particular XML format and their RDF counterparts. Going the other way,
RDF -> XML, again a formal mapping would be needed (probably, but not
necessarily the same as XML -> RDF). In practice you might well get the app
to output the preferred RDF syntax (verbose/abbrev) then run XSLT on it.

Ideally the mappings would occur between the RDF and XML models and the
domain model (in RSS, the humanly defined ideas of item, channel etc), so
they would each be a representation of the 'higher' abstraction. In other
words, keep the semantics well away from the syntax.

Your suggestion isn't far removed from what appears to be a fairly common
practice, using subclassing/subpropertying :

rss:title rdfs:subPropertyOf dc:title

One of the side effects is that it can produce a very plain XML-looking
serialization as you are suggesting.

There are a couple of advantages in this approach - the relation is defined
at a lower layer (RDFS rather than OWL) and the vocab writer is leaving
their options open for further tweaking/qualification of their term
(questionable in principle but invaluable in practice ;-)

It will be interesting to see how the development of vocabularies evolves
once everyone's got inference built into their RDF API to understand RDFS &
OWL relationships (particularly subclass/sameclass).


Danny Ayers

Semantic Web Log :

"The lyf so short, the craft so long to lerne." - Chaucer

>-----Original Message-----
>From: www-rdf-interest-request@w3.org
>[mailto:www-rdf-interest-request@w3.org]On Behalf Of Ian Davis
>Sent: 29 November 2002 19:32
>To: www-rdf-interest@w3.org
>Subject: RDF/XML without namespaces?
>This is just a musing.
>I wonder if it's possible to write RDF/XML without the use of
>namespaces if all the element names are aliased through an RDF schema
>to those in a different namespace using owl:sameClassAs. So long as
>basic striping rules are followed the resultant xml looks like, well,
>Can this thought be taken a step further? Is it possible to define xml
>attributes that are aliases for rdf:resource and rdf:about?
>For example, here's some RSS 1.0 (chosen only as an example of an RDF
>application that uses many different namespaces)
>  xmlns:rdf="http://www.w3.org/1999/02/22-rdf-syntax-ns#"
>  xmlns:dc="http://purl.org/dc/elements/1.1/"
>  xmlns:admin="http://webns.net/mvcb/"
>  xmlns="http://purl.org/rss/1.0/">
><channel rdf:about="http://internetalchemy.org/">
><title>Internet Alchemy</title>
><admin:generatorAgent rdf:resource="http://www.movabletype.org/?v=2.21" />
><rdf:li rdf:resource="http://internetalchemy.org/2002/11/weblogs.html" />
>With an appropriate RDF Schema maybe it's possible to write the
>following instead:
>  <Channel url="http://internetalchemy.org/">
>    <title>Internet Alchemy</title>
>    <link>http://internetalchemy.org</link>
>    <language>en</language>
>    <generator id="http://www.movabletype.org/?v=2.21" />
>    <items>
>      <List>
>        <item id="http://internetalchemy.org/2002/11/weblogs.html" />
>      </List>
>    </items>
>  </Channel>
>Syntactic sugar for the author only.
>-- Ian <iand@internetalchemy.org>
>"If we knew what it was we were doing, it would not be called
>research, would it?"
Received on Saturday, 30 November 2002 08:58:04 UTC

This archive was generated by hypermail 2.3.1 : Wednesday, 7 January 2015 15:07:43 UTC