- From: Dan Brickley <danbri@w3.org>
- Date: Wed, 6 Nov 2002 14:13:09 -0500 (EST)
- To: Pedro Assis in Oporto <passis@dee.isep.ipp.pt>
- cc: <www-rdf-interest@w3.org>
On Wed, 6 Nov 2002, Pedro Assis in Oporto wrote: > [freed from spam filter -rrs] > > Date: Wed, 6 Nov 2002 04:42:09 -0500 (EST) > Message-ID: <Pine.LNX.4.44.0211060912530.13945-100000@douro.dee.isep.ipp.pt> > From: Pedro Assis in Oporto <passis@dee.isep.ipp.pt> > To: www-rdf-interest@w3.org > > Hi, > > I've read several W3C and DAML docs about RDF, RDFS and DAML (DAML+OIL), > and their role in the Semantic Web project, but I'm still puzzel by the > answers given to the basic question: "why use RDF-based technology rather > than only XML-based?" If one use an OO modeling (expressive issue is > address here) to describe the knowledge related to the target domain, > translating the resulting schemata to XML, shouldn't the XML output > address the same functionality as a RDF-based solution? Is it a mater of > logic/inference capability? If you took this approach, the XML you generated *could* losslessly be converted back into a rich body of information. Unfortunately, nobody but you and your immediate collaborators would be able to do this without guesswork or special-purpose code, since the mapping that translates between expressive OO model and the XML serialization would be a 'private' (custom, proprietary, special-purpose etc) one. RDF's XML syntax provides one strategy for mapping between expresive OO-style models and an XML notation. Until we have richer XML schema annotation languages, RDF/XML may be our best bet for ensuring that rich information models can be turned into widely-consumable XML and back again easily. Instead of each application inventing its own conventions for translating into markup, we each agree to use RDF/XML as a 'good enough' middle ground. That's the theory, anyway... Dan -- mailto:danbri@w3.org http://www.w3.org/People/DanBri/
Received on Wednesday, 6 November 2002 14:13:47 UTC