The RDF view, which has been used as the basis for
discussing DC datamodels, is essentially OO anyway.
Also, I and others have mused about DC in UML occasionally.
Your view of objects and interfaces has probably been implicit in
some of our thinking, but it might be interesting to do this more
explicitly, for sure. "DC as an interface" is quite similar to
"DC as one view of the description" (which will probably actually
be stored in some local schema in most cases).
Anthony Finkelstein wrote:
>
> Has anybody considered an alternative view of DC - that is based on
> object-orientation? The information resource is an object and DC is
> an interface to that object. We replace the existing DC with an
> interface description and use an IDL in place of RDF. Thus metadata
> is not an attribute of passive data but a capability of an active
> resource.
--
Best Simon
|