[
Lists Home |
Date Index |
Thread Index
]
Bullard, Claude L (Len) wrote:
>Thanks John! So essentially it comes down to
>striping the XML with the RDF namespace such
>that the RDF processor can... then do exactly what?
>
Loose it ;-)
See below.
>I guess I am asking about the processes an
>RDF system would use to read and interpret the
>XML. I can guess but a guess isn't good enough.
>I am considering the semiote (a sign processor)
>and thinking that the markup for that could
>be striped with the RDF and make the data
>available to an RDF engine without using RDF
>per se. That would be cool.
>
>Manos: thoughts on that?
>
RDF has a simple but strict model, meaning the way it's XML is
de-serialized into triples to form the RDF graph. That's what makes RDF
processors able of dealing with any RDF. Vanilla XML on the other hand
is unpredictable in structure (as well as the actual meaning of that
structure) and levels of depth - from the RDF point of view, XML is
ugly, low level and meaningless. The operation of viewing any non-RDF
markup using RDF rules to interpret it will result in loss of
information and missinterpretation.
Some people may want to look at [1] for an explanation of the
differences between RDF and XML.
>To me, this idea that one can make their XML
>RDF-friendly is pretty powerful as a means
>of bridging from the common web to the semantic
>web.
>
If you put it this way, things are different. Producing RDF-friendly XML
is indeed easier than most people think, John Cowan already covered most
syntax rules.
Also, converting XML to RDF automatically is simple, the problem is
that you will just come up with an RDF serialization of the XML Infoset ;-)
[1] http://www.w3.org/DesignIssues/RDF-XML.html
Kindest regards,
Manos
|
