Cart
[Home] [By Thread] [By Date] [Recent Entries]
Miles Sabin <msabin@c...> writes:
> If all you're interested in is representing a formal graph
> structure, then fine. But surely to be interesting we have to
> be able to _interpret_ this stuff (I'm intending the contrast
> between 'formal' and 'interpreted' here in the same sense it's
> used in semantic theories).
The idea of using RDF (or any extensible, distributed
object-representation format) is that you can work with partial
knowledge. RDF allows that in two ways:
1. If you do not recognize a class, it is still possible to recognize
some of its properties and act accordingly, i.e.
unrecognized class megg:Foo has a recognized property megg:rating
2. If you don't recognize a class or property, you may recognize one
of its superclasses or superproperties, and be able to act
accordingly, i.e.
unrecognized class megg:Bungalow is a kind of
recognized class megg:House
unrecognized property megg:length is a kind of
recognized property megg:linear-measurement
Traditional OO implementations allow only #2; I actually question how
usable #2 will be for distributed XML objects, since it depends very
heavily on schemas, but #1 is clearly useful.
All the best,
David
--
David Megginson david@m...
http://www.megginson.com/
***************************************************************************
This is xml-dev, the mailing list for XML developers.
To unsubscribe, mailto:majordomo@x...&BODY=unsubscribe%20xml-dev
List archives are available at http://xml.org/archives/xml-dev/threads.html
***************************************************************************
|
