RE: TexasThings and owl:equivalentThing

Mike, Ian

Thanks for picking up the issue. Better late that never :))

BTW my suggestion would be to give up the Texas Thingies altogether, and
maybe take the same examples than in the Reference document (Mozart's
operas). The point for it is a very simple and generic pedagogic principle:
any introductory example should offer as small as possible space for
interpretation and discussion. Which means that the reader should instantly
be provided with a clear notion of what is an instance and what is not. In
that regard, Texas Thing, even with a someValuesFrom restriction, is still
confusing. Take the following examples.
There is no indication of cardinality for "locatedIn". Is a company with
several offices, some located in Texas and some not, a Texas Thing?
There is now indication of time extension. The Texan boots I bought 23
years ago in San Angelo (authentic) have been located in France ever since.
They've been Texas Things at a time at least, but are they still now?
And so on. Trying to figure out the extension of the class, the reader is
led astray to such irrelevant paths by too open definition, and misses the
point of "equivalentClass". See my "BlueThing" example in rdf-logic ...
OTOH, providing a non-ambiguous definition of Texas Thing would be very
arbitrary and controversial, and would need more restrictions, which would
make the example confusing by its complexity.

It figures. Finding good examples is difficult but important. 99% of users
will build their understanding from examples, not from abstract syntax.

Bernard

Bernard Vatant
Senior Consultant
Knowledge Engineering
Mondeca - www.mondeca.com
bernard.vatant@mondeca.com


> -----Message d'origine-----
> De : public-webont-comments-request@w3.org
> [mailto:public-webont-comments-request@w3.org]De la part de Smith,
> Michael K
> Envoyé : jeudi 18 décembre 2003 18:34
> À : Ian Horrocks; Bernard Vatant
> Cc : public-webont-comments@w3.org
> Objet : RE: TexasThings and owl:equivalentThing
>
>
>
> Bernard,
>
> I meant to respond last week but was swamped.  Of course you are correct.
> This will be corrected in next version.
>
> - Mike
>
> Michael K. Smith (+1-512-404-6683)
>
>
>
> -----Original Message-----
> From: Ian Horrocks [mailto:horrocks@cs.man.ac.uk]
> Sent: Thursday, December 18, 2003 4:19 AM
> To: Bernard Vatant
> Cc: public-webont-comments@w3.org
> Subject: Re: TexasThings and owl:equivalentThing
>
>
>
> Bernard,
>
> Good points. Hopefully this can be fixed by the Guide editors.
>
> Ian
>
> On December 15, Bernard Vatant writes:
> >
> >
> > Seems to me that there is something wrong, or at least
> misleading with the
> > example of "TexasThings"
> > http://www.w3.org/TR/owl-guide/#equivalentClass1
> >
> > <owl:Class rdf:ID="TexasThings">
> >   <owl:equivalentClass>
> >     <owl:Restriction>
> >       <owl:onProperty rdf:resource="#locatedIn" />
> >       <owl:allValuesFrom rdf:resource="#TexasRegion" />
> >     </owl:Restriction>
> >   </owl:equivalentClass>
> > </owl:Class>
> >
> > First it would certainly be better to have the singular "TexasThing"
> rather
> > than plural "TexasThings" :)
> >
> > "TexasThings are exactly those things located in the Texas region ..."
> >
> > This is obviously wrong. There is an allValuesFrom, but not a
> > someValuesFrom here. As defined, it means : If TexasThings are located
> > somewhere, they are located in Texas region.
> >
> > An further on
> >
> > " ... The difference between using owl:equivalentClass here and using
> > rdfs:subClassOf is the difference between a necessary condition and a
> > necessary and sufficient condition. With subClassOf, things that are
> > located in Texas are not necessarily TexasThings. But, using
> > owl:equivalentClass, if something is located in Texas, then it
> must be in
> > the class of TexasThings."
> >
> > ... but not the other way round, unfortunately.
> >
> > In fact under this definition any thing located nowhere is a
> Texas Thing.
> > Thinking about it, maybe it makes sense after all. Nowhere is indeed in
> > Texas, and especially its middle ... :))
> >
> > Bernard Vatant
> > Senior Consultant
> > Knowledge Engineering
> > Mondeca - www.mondeca.com
> > bernard.vatant@mondeca.com
> >
>

Received on Friday, 19 December 2003 05:39:49 UTC