Re: [OEP] n-ary relations and topic maps

I think I agree with the point Lars is making, which is, I think, that the 
relationship between n-ary relations in topic maps and "workarounds" for 
RDF probably doesn't belong in the n-ary relations note. 

The general idea, that the limitation to binary properties in rdf requires 
a workaround that is not required in other languages and formats, does not 
need, IMHO, special treatment in a document that deals with RDF.  There 
are a lot of representation formats that naturally handle n-ary relations, 
probably the most important commercially is ER diagrams, probably the 
oldest is good old FOL.

So I don't think there is a need for an appendix in the n-ary relations 
note for topic maps,  because that opens the door to arbitrarily extending 
the note for every other format.  If there is a need to describe the 
relationship between the workaround in the n-ary note and topic maps, then 
I think it should be in a note about topic maps and there relationship to 
RDF.

-Chris

public-swbp-wg-request@w3.org wrote on 12/01/2004 03:35:07 AM:

> 
> 
> * Lars Marius Garshol
> |
> | This means that the RDFTM task force will need to consider:
> |
> |   a) how to convert binary and n-ary relations from RDF to topic
> |      maps,
> 
> * Natasha Noy
> | 
> | [...] I would like to point it out that the next version of the note
> | will contain an appendix written by Bernard Vatant on expressing
> | n-ary relations in Topic Maps:
> | http://lists.w3.org/Archives/Public/public-swbp-wg/2004Jul/0057.html
> |
> | This will probably address the first issue in your message.
> 
> I haven't seen what Bernard's done yet, so this isn't a reflection on
> his work, but I don't really see how an appendix on that subject can
> solve my issue a) above. The reason is that expressing n-ary relations
> in topic maps is not the problem (there are issues, of course, and I
> expect these are what Bernard wrote about). The real problem, I think,
> is recognizing when you see an n-ary relation in RDF, so that you know
> you should convert it into a single topic map association, as opposed
> to n topic map associations.
> 
> | Any feedback on that will of course be appreciated (I don't think
> | there was any feedback when Bernard posted the message).
> 
> I've had a brief look now, and will come back with more detailed
> feedback later. As far as I can tell, Bernard hasn't attempted to deal
> with the RDF half of the issue (and quite reasonably so, since he
> probably considered it out of scope).
> 
> -- 
> Lars Marius Garshol, Ontopian         <URL: http://www.ontopia.net >
> GSM: +47 98 21 55 50                  <URL: http://www.garshol.priv.no >
> 
>