Use areas (Re: State of the N-ary)

I think the Product Modelling XG is going to have many (not just  
units). I wanted to call out this post:
	http://lists.w3.org/Archives/Member/member-xg-w3pm/2008Jun/0016.html

For this quote:
	"""Inevitably mathematics is involved, and RDF/OWL does not do  
mathematics."""

I've heard this complaint several times over the years, e.g., by  
Ashok Malhotra (of Oracle; also on XML Schema group as an editor of  
some documents and active in web services, policy etc.). This makes  
it a *non*-starter in many areas. (C&P needs sophisticated predicates  
for policy analysis, though much of that would be "behind the  
scenes". Of course, C&P'd want to *also* expose that implementation  
to modelers. To put it another way, C&P is inclined to implement  
either way.)

Science, product management, engineering, configuration, medicine,  
bioinformatics...all cases where sophisticated mathematics (and  
string processing) are critical. Many of these use TBox reasoning  
(i.e., consistency and subsumption) as critical path in their  
applications (when they can use OWL at all).

This is all distinct from massaging data. That's needed too, but is  
often handled (and better) by custom programming. I don't think OWL  
is going to compete heavily there (though we should *work* with  
solutions in those problem spaces). Whereas, constraint reasoning +  
conceptual reasoning is an unique and useful package.

This is my high level motivation. Obviously, the devil is in the  
details. But this is why I think getting a robust level of n-ary  
support is a really wonderful thing. It opens lots of new areas.

Cheers,
Bijan.

Received on Wednesday, 2 July 2008 09:57:47 UTC