GUIDE review

I did a review of the Guide draft when coming back from the ftf, but 
never got around sending it I checked my comments against the last 
version, and it seems almost all of them are gone, including the errors 
in wine.owl (I ran it through oru parser).
I have a few comments left, but there may be some toher changes required 
dependning on the outcome of the discussions about the points Jeremy 
raised.

1. "WineDescriptor" is defined as the superclass of classes such as 
"WineColor". These subclasses are defined as enumerations of their 
values (e.g red, white and rose for WineColor). This means that the 
WineDescriptor classs has as its instances the union of all these 
values. This looks a bit strange. Actually, it looks like a C-as-I 
problemL WineColor is really an instance of WineDescriptor. Probably too 
much work to change at this stage, byt a remark would be in order (also 
stating the C-as-I version would have been OK in Full).

2. Some properties have owl:Thing as their domain, which is by 
definition true.

3. InverseFunctionalProperty: the Guide says in OWL Full datatype 
properties can be of this type. If I understand the semantics correctly, 
the real story is a bit dufferent. In OWL Full all properties are object 
properties (owl:ObjectPorperty sameClassAs rdf:property), because data 
values belong to the domain of individuals. This makes datatype 
properties are in gacy a subclass of object properties, and thus there 
is no problem applying inversefunctional to such a property. This is at 
least how I explain it currently in the Reference. If my view is wrong, 
I will have to chnage that.

That's all, excellent work.

Guus




http://www.swi.psy.uva.nl/projects/SWI-Prolog/packages/sgml/online.html
http://www.swi.psy.uva.nl/projects/SWI-Prolog/packages/sgml/online.html

-- 
A. Th. Schreiber, SWI, University of Amsterdam,
http://www.swi.psy.uva.nl/usr/Schreiber/home.html

Received on Wednesday, 22 January 2003 09:38:55 UTC