RE: Close to final draft of "classes as values" note

Natasha

Some comments about the final draft. There are still a few things I'm uneasy with, from a
practitioner viewpoint.

First I would like to see more precise definition of the use cases in each approach : what
is given (the legacy), what are the objectives of the use case, and what is to be built to
achieve them.
Basically we have two KOS : on one side an ontology, on the other side a
classification-indexing system (a library system, to make it short).

It's unclear in each approach if the use case is :

#1 : Both ontology and library system are considered as given in the legacy, have been
built and managed independently, and the problem is to map them, from one side or another.
The use case should make it clear which is the master and which is the slave.
#2 : The two systems are developed together in an integrated environment. This looks like
a "closed" use case not really in the scope of the Semantic Web deployment. Maybe we
should skip that one.
#3 : One of the system is built to be best interoperable with the other, which is given
(can be both ways).

Reviewing the approaches, seems to me that the document takes it implicitly for granted
that the ontology is the legacy, but it's not obvious. OTOH in the current state of the
real world, there is more legacy library systems than legacy ontologies. I would like for
example to figure how I can leverage the following Web legacy of indexing categories:

http://isbn.nu/sisbn/lions/
http://isbn.nu/sisbn/mammals/
http://isbn.nu/sisbn/animals/

http://dmoz.org/Kids_and_Teens/School_Time/Science/Living_Things/Animals/Mammals/Lions/
http://dmoz.org/Science/Biology/Flora_and_Fauna/Animalia/Chordata/Mammalia/Carnivora/Felid
ae/
http://dir.yahoo.com/Science/Biology/Zoology/Animals__Insects__and_Pets/Mammals/Cats/Wild_
Cats/Lions/

I'm not aware of any real-world OWL ontology where the class "Lion" is actually defined,
but it would be good to have one to set a use case of type #1.
Use case #3 would actually build an OWL ontology (e.g. derived from some authoritative
zoological taxonomy) that would help to leverage the above legacy, e.g. support a
federated request over isbn.nu, dmoz.org and dir.yahoo.com indexes, with capacity of
inference.

Another point : I'm quite unesay with Approach 2, where the class "Lion" is defined
completely differently and in a very odd way. How many instances do you have for this
class? One by distinct subclass? Strange and potentially confusing. Would not it be better
to have a consistent and as "natural" as possible definition of the reference ontology
classes (Animal > Lion > African Lion) throughout the approaches, of which the individual
"Lenny the lion next door" is an instance? Of course, that means either striking Approach
2, or have a completely different use case for this approach.

Minor point. Seems there is a problem with N3 examples. My browser don't like them :(

Cheers

Bernard


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


> -----Message d'origine-----
> De : public-swbp-wg-request@w3.org
> [mailto:public-swbp-wg-request@w3.org]De la part de Natasha Noy
> Envoye : mercredi 12 mai 2004 22:24
> A : swbp
> Objet : Close to final draft of "classes as values" note
>
>
>
>
> You can find the next (and, hopefully, close to last iteration) of the
> "Classes as values" note at
>
> http://lists.w3.org/Archives/Public/www-archive/2004May/att-0019/
> ClassesAsValues-v3.html
>
> At this point, as we've agreed in the previous telecon, I would like to
> propose this document as an official note.
>
> Thanks to everyone for their comments and suggestions.
>
> Major changes:
>
> 1. I've added one more approach (approach 4) that approximates the
> solution by identifying the book as a  book about *some* lions. It is
> indeed an approximation, but the question has come up often enough that
> I felt it was worth at least mentioning and explaining what's different
> about such an approach. Please take a look at the considerations there,
> since this stuff is all new.
>
> 2. I've changed to real book examples, as Bernard suggested. I used
> rdfs:seeAlso to refer to the isbn site (cool resource indeed)
>
> 3. I have not used SKOS properties in approach 3 directly (avoiding the
> addition of another complexity level), but I have mentioned it (and
> also used seeAlso to point to the corresponding SKOS property). I hope
> it's a reasonable compromise.
>
> Natasha
>

Received on Thursday, 13 May 2004 05:42:31 UTC