Re: possible semantic bugs concerning domain and range

>From: pat hayes <phayes@ai.uwf.edu>
>Subject: Re: possible semantic bugs concerning domain and range
>Date: Tue, 24 Sep 2002 18:09:38 -0500
>
>[...]
>
>  >
>>  Heres a way to phrase the difference: call a class a protorange iff
>>  it contains all the values of a property. For you, protoranges are
>>  ranges. For me, only some of the protoranges need be ranges. Clearly,
>>  iff semantics is appropriate for protoranges; but many applications
>>  of the notion of range require us to be able to identify particular
>>  protoranges as the ones to which other information is associated,
>
>I would like to know about these applications.  In particular, I would like
>to know about these applications that actually work correctly in the
>presence of super-properties with different ranges.

Well, I have to confess to not being able to cite particular ones 
from here in my ivory tower, but I have been given that strong 
impression by others in the WG. And I can certainly imagine some: for 
example, suppose that one had a class of classes corresponding to the 
categories that are known to have associated implemented systems 
(say, membership in which can be checked by a particular kind of 
efficient application); and one wanted to say that a property range 
was in that class.

YOu are right to observe that this kind of reasoning requires some 
care with subPropertyOf. HOwever, I have noticed that almost ALL uses 
of subPropertyOf require care, and indeed that many users never use 
it for just this reason.

>
>>  and
>>  if we make the identification then this ability to distinguish
>>  particular protoranges is lost (or requires extra machinery). Whereas
>>  it seems to me that applying the iff semantics provides no useful
>>  extra entailments. It allows one to conclude that many more classes
>>  are ranges, of course, but all this does is make manifest that the
>>  notion of 'range' has been (from my point of view) fatally weakened.
>
>Please present some indication of how the iff definition of range fatally
>weakens the notion of `range'.

Well, perhaps fatally is rather too strong. Given your semantics, I 
could hack round it by introducing a class of classes called 'real 
ranges' or some such, and then what I call a range is a pfps-range 
that is also a real range. So I guess its more of an aesthetic 
difference than a fatal one. But since the difference is only 
aesthetic, I see no good reason to change what we already have.

>
>>  These can all be expressed using my notion of range and
>>  rdfs:subClassOf or rdf:type. The important inferences about ranges -
>>  notably, the kind that arise from an association of a datatype with a
>>  range - apply in both semantics, but require more care to state in
>>  yours. 
>
>Do these inferences actually work?  I thought that RDF Core had decided
>that they didn't work in the presence of super-properties.

No, the problem was the possibility of an XML datatype value space 
being included in another when the datatype mappings were 
incompatible, and we basically decided to punt on that one. I don't 
think superproperties pose any problem, but maybe I missed something: 
what do you see as the problem there?

Pat
-- 
---------------------------------------------------------------------
IHMC					(850)434 8903   home
40 South Alcaniz St.			(850)202 4416   office
Pensacola,  FL 32501			(850)202 4440   fax
phayes@ai.uwf.edu 
http://www.coginst.uwf.edu/~phayes

Received on Wednesday, 25 September 2002 13:13:31 UTC