xml-dev - Re: RDF? TM? (was Re: Didier's lab report)

[Uche Ogbuji <uche.ogbuji@fourthought.com>]

BTW, Jonathan, thanks for explicitly explaining typed nodes.  I'd been using
them in my examples in my discussion without it's occurring to me that I
should explain what they were.  I assumed too much understanding of RDF.

> Jonathan said:
> 	There is already a clearcut mapping of XLink to RDF. Furthermore using a
> particular type of RDF parsing (parseType='Resource') *arbitrary* XML can be
> parsed into RDF triples. I have written an XSLT which performs Xlink -> RDF
> and parses arbitrary XML into RDF triples:
>
> 	http://www.openhealth.org/RDF/rdfExtractify.xsl
>
> Didier replies:
> You are right Jonathan, we can translate a collection of links into RDF
> expressions and then reduce the document potential processes to RDF
> processing.  However, the main point is that if we can get an notation that
> merges both, the same document can be interpreted in two different ways:
> a) as a link base
> b) as a metadata base - or frames providing information about resources.
> It means that an Xlink interpreter and an RDF interpreter can provide two
> different interpretations of the same document. By doing so, we just
> increase the number of facets on how this document is revealed to us. Take a
> phenomenological position where an object is revealed by a point of view. We
> than now provide more than one point of view to this document. A linkage
> point of view and an information about resources point of view. This implies
> that we just increased the value this document may intrinsically possess.

Yes, but Didier, this is exactly what Jonathan is pointing out that you
can have.  I think you seized on the wwrong part of his message.  First he
pointed out that by using typed nodes, you could have your ordinary XML
elements have all the value of explicit rdf:description elements.  The RDF
model & syntax (M&S) spec provides a wide variety of such abbreviations
from the full rdf:description syntax in order to make it easier to
(A) provide RDF information in formats more natural to XML users who
don't care about RDF, and (B) make it easier to transform current formats
to RDF.

See RSS [1] for an excellent example of (A).  My article on 4RDF [2]
has a quicker, more hands-on example.  You might want to have a look at my
article about WSDL and RDF [3] for an example of (B).

But not all XML fits smmothly into the vision of RDF.  Tools such as
Jonathan's excellent transform for arbitrary XML [4], or the WSDL->RDF
conversion in another of my articles [5] handle those cases.  However,
with many XML formats, they would not be necessary.  You could have your
multi-faceted document without the intervention of XSLT.


-- 
Uche Ogbuji                               Principal Consultant
uche.ogbuji@fourthought.com               +1 303 583 9900 x 101
Fourthought, Inc.                         http://Fourthought.com
4735 East Walnut St, Ste. C, Boulder, CO 80301-2537, USA
Software-engineering, knowledge-management, XML, CORBA, Linux, Python

[1] http://www.egroups.com/files/rss-dev/specification.html#
[2] http://www.xml.com/2000/10/11/rdf/index.html
[3] http://www-106.ibm.com/developerworks/library/ws-rdf/index.html?dwzone=ws
[4] http://www.openhealth.org/RDF/rdfExtractify.xsl
[5] http://www-106.ibm.com/developerworks/library/ws-trans/index.html?dwzone=ws