Re: [OEP] minutes of 5/26 telecon

There is a new Editor's draft available that takes care of the things  
we agreed at the telecon. More comments in-line.

> We discussed the "unintended models" point as well.  It turns out  
> that the
> comment about RDF treating two triples with the same S,P,O as "the  
> same"
> is not accurate.  As a result, the point is more general than just  
> n-ary
> relations, and is also more complicated than the bullet describes. We
> resolved to remove this bullet and move that point to the  
> "pitfalls" note,
> with perhaps a forward reference to it.

For the moment I just removed it since I couldn't come up with a good  
way to put the forward reference (without getting into the details)  
or a place for it. If someone wants to add it and sees a way to do  
it, please go ahead. My feeling at the moment is that the issue is  
indeed not specific to n-ary relations, but rather shows up pretty  
much everywhere and hence not worth a special forward reference. It  
certainly should be part of the Pitfalls document.

> Finally, we discussed the proposed standard vocabulary for reified
> relationships.  Natasha suggested that specific vocabulary for  
> mapping OWL
> to other languages does not belong in this note, in particular the
> "argNum" property in the proposed vocabulary is for mapping n-ary
> relations to languages that use argument position to encode the  
> role. The
> "use cases" for this standard vocabulary were 1) tools that treat  
> reified
> n-ary relationships in some special way and thus need to know which  
> ones
> they are, and 2) translating OWL & RDF to other languages that support
> n-ary relations in the syntax.

I felt that while the part of the vocabulary describing  (1) belonged  
in this document, the part for (2) did not. Several reasons: First,  
the mapping part allows you to number the roles in your n-ary  
relation to specify which positions they should go into when you are  
mapping to another ontology in another language. These numbers change  
if you are mapping to another ontology in that language. All of this  
requires a fair amount of explanation (we don't ever mention that  
order of roles is relevant anywhere, and, except for mapping it is  
not). I am not convinced that this mapping discussion belongs in this  
note.

Second, as was pointed out at the telecon, the vocabulary that allows  
you to number the positions of different roles may not be sufficient  
or appropriate for mapping to all possible languages (in general,  
saying that this vocabulary is all that you will ever need to map n- 
ary relations from OWL ontologies anywhere else seems to be a bit  
ambitious and hard to prove). Evan and I brought up the example of  
Association classes in UML. To map to an association class, you  
basically need to identify the two roles among all your roles that  
would go as the two ends of the relation, and all the rest will be  
encapsulated in this additional class. This would require a different  
kind of vocabulary.

I would suggest that we include in the note the vocabulary for  
telling the world you are an n-ary relations and have a separate note  
(edited by Chris) that talks about mapping vocabulary for this and  
other cases.

However, as I pointed out at the telecon, I was not  the only editor  
of the note and I would like to let Alan weigh in as well (he  was  
not at the telecon).

Natasha

Received on Thursday, 26 May 2005 22:43:53 UTC