Re: TEST: Re: notes for 6/6 until 1:10 (oneOf/sameClassAs)

From: "Lynn Andrea Stein" <lynn.stein@olin.edu>
Subject: Re: TEST: Re: notes for 6/6 until 1:10 (oneOf/sameClassAs)
Date: Wed, 26 Jun 2002 13:42:46 -0400

> 
> Stepping off a [BIG ;_)] cliff here....


> Text quoted in the following is excerpted from the full message at
> http://lists.w3.org/Archives/Public/www-webont-wg/2002Jun/0183.html  Context
> is given only where it is necessary to understand what I'm saying, not why or
> who said what's quoted.
> 
> There were several examples in the cited email (and its predecessors) that I'd
> like to propose restating with explicit assumptions about the implicit
> existentials.  I want to know whether the restatements are unacceptable.  If
> they are, then we may really have to deal with issues like
> 5.10-DAML+OIL-semantics-is-too-weak, but if the restatements are acceptable
> maybe they provide a way out.
> 
> Example 1:
> > >i.e. from nothing, conclude:
> > >
> > >[ owl:oneOf ( :a :a :b ) ] owl:sameClassAs [ owl:oneOf ( :b :a :a ) ] .
> 
> Restatement:
> 
> from
> [owl:oneOf (:a :a :b)]           the class owl:oneOf (:a :a :b) exists
> and
> [owl:oneOf (:b :a :a)]           the class owl:oneOf (:b :a :a) exists
> conclude
> [ owl:oneOf ( :a :a :b ) ] owl:sameClassAs [ owl:oneOf ( :b :a :a ) ] .

My view is that this begs the question entirely.  I would think that
everyone in the WG would agree with the above restatement.  However, I view
this restatement as a completely different example.

> or even
> 
> from
> classExists :a
> and
> classEexists :b

As already stated, :a and :b do not have to be (and were not thought of as)
classes.  

> conclude
> [ owl:oneOf ( :a :a :b ) ] owl:sameClassAs [ owl:oneOf ( :b :a :a ) ] .
> 
> 
> (where classExists is just a shorthand for a statement that the class exists,
> either using existing owl vocabulary or through some new mechanism)
> 
> After all, if the classes :a and :b don't exist, what would the consequent
> statement mean, anyway?

Well, if :a and :b are in the vocabulary, then the consequent says that the
two oneofs are the same.  (Yes, the requirement that :a and :b be in the
vocabulary may be missing.)

> Example 2:
> > ...the inference for (P and Q) from (Q and P)... in class
> > language requires...:
> > the existence of the class (P intersect Q) entails the existence
> > of the class (Q intersect P).
> 
> 
> Restatement:
> 
> from
> exists P, exists Q, 

Again, the requirement that P and Q are in the vocabulary might be missing.

> (P and Q)

Again, I view this condition as completely changing the example.
 
> conclude
> (Q and P)
> 
> or
> 
> from
> classExists (P intersect Q)
> conclude
> classExists P
> and
> classExists Q
> and
> classExists (Q intersect P)

Ditto.

> Example 3:
> 
> >   (forall (?x ?L)
> >     (=> (ow:item ?L ?x)
> >         (exists (?ex1) (and (rdf:type ?x ?ex1) (owl:oneOf ?ex1 ?L))) ) )
> 
> Restatement:
> pull the existential (as a special predicate) into the antecedent of the
> implication, i.e.
> 
>  (forall (?x ?L ?ex1)
>     (=> (and (classExists (?ex1))
>                 (ow:item ?L ?x))
>          (and (rdf:type ?x ?ex1)
>                (owl:oneOf ?ex1 ?L))) ) )
> 
> Again, if ?ex1 doesn't exist, then what on earth could the original statement
> mean?  (And the existence as a class of ?ex1 could be derived automatically
> from the existence as a class of each of the elements in ?L....)

The *whole* idea is that ?ex1 is forced to exist.  Your change above either
does nothing to the example or modifies it in the same unacceptable-to-me
manner that you modified the previous examples.

peter

Received on Saturday, 29 June 2002 00:06:35 UTC