 |
 |
On Tue, 9 Sep 2003, Mikael Nilsson wrote: > On Tue, 2003-09-09 at 15:35, Sigfrid Lundberg wrote: > > > > What is the difference between a an abstract representation of an > > aggregation of properties used to form a "record" and XML or RDF > > schemas? > > The difference is that RDF schema cannot (by design) handle collections > of properties as a "record"; there is simply no such concept in RDF > schema. It only talks about the semantics of individual classes and > properties, and their relations, not about aggregating them. > > OWL has minimum cardinality contraints for properties, but this still > does not mean that every property needs to reside in the same RDF > document :-) That is the strength and the weakness of RDF 8-) ... > > Yes, and there are several RDF-based (experimental) languages for > specifying such things. When I think of it, isn't there ISO standard about data dictionaries which is meant for this kind of descriptions. If I remember that correct we even had some of our documents structured that way for some time, until we choose a way which was, sort of, more compatible with RDF/XML way of thinking. Wasn't Renato fond of ISO data dictionaries, and so was IMS at some point. This aggregation concept is also closely related to "application profiles"... If we go on with this discussion we might even reinvent the DCMI Registry. Sigge