A roundup of recent developments in the RDF world, including XSLT transformations for RDF generation, a schema-based approach for triple-generation, and simplified syntax discussions.
New initiatives are flourishing along the paths recently
drawn by Leigh Dodds
in his XML-Deviant column for finding ways to extract semantic information from XML sources.
XSLT: Running further on the XSLT highway, Dan Connolly
announced
an XSLT transformation
"to scrape RSS/RDF" out of the W3C home page,
allowing the publishing of a RDF/RSS 1.0 channel containing the W3C headlines.
Syntax: Following the syntax track and taking over on the long running "strawman" thread,
Jonathan Borden announced
a new simplified syntax--inspired by a previous proposal
by Sergey Melnik--and also an XSLT transformation converting this syntax directly
into RDF triples.
XLink: On the XLink road, Ron Daniel has asked
for a last call of comments about his XLink to RDF mapping,
and Jonathan Borden mentioned a
first partial XSLT implementation.
Schemas: While these approaches are either intrusive (by imposing some constraints on the XML
vocabularies as is the case with a simplified or a XLink syntax) or based on
the development of specific transformations for each application, James Tauber
has proposed
a schema-based approach where the mapping between XML and RDF
would be defined in separate schema files:
I believe that all descriptions of serialization of RDF should be separated out of the RDF Syntax specification and could be described merely in terms of the mapping language. I would imagine that powerful mapping language could be achieved simply by mapping XPaths to subject, predicate and object (and perhaps a handful of properties such as type and label.
Tauber is calling for comments and volunteers to work on his proposal, along with his Redfoot RDF server project:
My second question is: would anyone be interested in working on this (along with an implementation) as part of the Redfoot project?
Has anyone asked Information Architects to get involved with this project? They have SmartCode as they call it which promises some rather definite advantages in communicating bio-directionally using RDF and Metadata.
www.IA.com