what are the best primitives for RDF/OWL?

I'd like to modify the focus of "deprecated URIs" a little bit,
and summarize some of the discussion on this list in 2003 time frame.

First topic: SubClassOf vs. properSubClassOf.
By definition,  X  SubClassOf  Y has two possible meanings.
1)  X  sameAs Y
2)  X  properSubClassOf  Y
Although meaning (1) may seem to have nice mathematical properties,
I think it's harmful to always have to "dispose" of that possibility when
reasoning about classes.

Second topic: Entity vs. Class vs. ClassSet
1. Entities are the primary "things" that exist, and Entities have 
Properties.
2. Some Entities are related to other Entities.
3. There are also secondary "things" that exist -- Properties have 
Properties,
and some Properties are related to other Properties.
4. Since RDF and OWL do not have Entities, there has been a tendency
to use "Class" as a substitute for Entity.
5. Next we mix in a little of 1st topic above, and begin to say things like
Class  SubClassOf  Class, and  Thing  SubClassOf  Class, which really
"don't make much sense", but are mathematically correct because of the
sameAs possibility.
6. I have argued in the past that RDF/OWL would be significantly improved 
by:
using "ClassSet" in place of the current idea of "Class";
including "Entity" as one of the primitives of the RDF/OWL languages;
using "properSubClassOf" as a primitive instead of "SubClassOf".

In other words, you take "sameAs" and "properSubClassOf" as primitives,
representing two fundamentally different and disjoint possibilities.

Dick McCullough
http://mKRmKE.org/
Ayn Rand do speak od mKR done;
knowledge := man do identify od existent done;
knowledge haspart proposition list;
mKE do enhance od "Real Intelligence" done;

Received on Thursday, 10 July 2008 01:48:45 UTC