- From: Nick Matsakis <matsakis@mit.edu>
- Date: Thu, 30 Oct 2003 12:44:30 -0500 (EST)
- To: "Bass, Mick" <mick.bass@hp.com>
- Cc: www-rdf-dspace@w3.org
On Wed, 29 Oct 2003, Bass, Mick wrote: > Andy Seaborne has mentioned the importance of modelling the concepts > that the elements represent. When modelling, I believe it is crucial to be able to fall back and discuss the underlying concepts, but I don't believe that they should be a part of the model itself. If we have a good model, then we should rarely have to break the abstraction of that model. But when we do we need a language for it; In this demonstrator it is important to be able to say, "These concepts are not expressible in VRA core" or "These are not expressible in IMS". I don't feel that concepts need to be included into diagram of the mapping between VRA, Dublin, and IMS; that diagram is lucid as it stands (to me, at least). I was only making a (minor) objection to the previous Venn diagram, which seemed to be saying that there was overlap in VRA categories and IMS elements, but was really saying that there is overlap in the concepts expressible by each of those. Maybe I should say more generally that Venn diagrams are good for expressing containment and exclusion relationships between subsets of a larger set, but not so good for expressing projection and mapping relations between subsets. Nick
Received on Thursday, 30 October 2003 12:44:37 UTC