RE: ISSUE: owl:Class name misleading; try owl:Set?

On Thu, 2003-01-02 at 13:34, Ziv Hellman wrote:
> My two sense:
> 
> 1. Classes in OWL _ARE NOT_ the same as sets,

Hmm... if I write YES, THEY ARE, does it convince you?

i.e. please explain why you think owl:Classes
are not the same as sets. (I know that rdfs:Classes
are different.)

Please provide an example of two owl classes that
have the same members but differ in some other
property; i.e. some ?C1, ?C2, ?P, and ?X
such that

	?C1 a owl:Class.
	?C2 a owl:Class.
	?C1 rdfs:subClassOf ?C2.
	?C2 rdfs:subClassOf ?C1.
	?C1 ?P ?X
but not
	?C2 ?P ?X.

>  as the word is generally defined and used in mathematics,
> so it would be disastrously wrong to change owl:class to owl:set
>  -- this distinction ought to be kept clear to the general public.
> 
> 2. The matter of how misleading sameClassAs can be as mentioned
>  below is indeed tricky, because it exactly highlights the
>  intensional/extensional distinction -- thinking of intension
>  and extension as two functions on the class of classes, we mean
>  by sameClassAs that the two items have different values from the
>  perspective of the intension function,

What's the intension function? I don't think this line of
argument helps me much. I'm trying to get a strictly
formal understanding.

>  but the same value from the
>  perspective of the extension function, so owl:sameExtensionAs is
>  more precise. But is it really requiring too much understanding
>  of logic to use that name when at the same time we expect users
>  to grok intensional and extensional distinctions if they are going
>  to use full OWL correctly?

sameMembersAs seems intuitive and correct.

-- 
Dan Connolly, W3C http://www.w3.org/People/Connolly/

Received on Thursday, 2 January 2003 16:54:45 UTC