RE: What do the ontologists want

From: "Charles F. Munat" <chas@munat.com>
Subject: RE: What do the ontologists want
Date: Wed, 16 May 2001 08:00:51 -0700

> Peter F. Patel-Schneider wrote:
> 
> The basic complaint of the first group is that people in the second group
> are going beyond what RDF is capable of.  People in the second group use
> the reification syntax, but have some extra meaning for it that is not
> shared by all interested in RDF.  It is the contention of the first group
> that the use of these extra meanings make RDF no longer be a true
> representation language, and thus ill suited for representing information
> in the WWW.
> 
> Reply:
> 
> Yes, that is pretty clear from the discussion. Also clear is that it isn't
> going anywhere anytime soon.
> 
> Is there a document somewhere that explains the limitations (clearly) of RDF
> according to the first group? In other words, could I start using RDF, limit
> myself to the sorts of uses that the first group approves of, and avoid
> controversy while the second group makes its case?
> 
> I'm just trying to pull something out of this that I can use right now while
> people who know a lot more than I do put together a "version 2" to solve
> these other problems.
> 
> Any help appreciated...
> 
> Sincerely,
> Charles F. Munat
> Seattle, Washington

I think that there have already been some discussions on this, but here are
a couple of points.

If you restrict yourself to basic triples, i.e., no reification, no
alternatives, probably no bags, in RDF, then there will be no problem.  If
you restrict yourself to basic classes, i.e., no new meta-classes, no new
classes of properties, no user-defined constraints, no multiple domains or
ranges, in RDFS, then there will be no problem.  Just about all of the
above limitations have undergone significant discussion.

It would also be best to limit your use of literals.

peter

Received on Wednesday, 16 May 2001 12:14:51 UTC