Re: Knowledge Representation Problem
dwig@markv.com
To: cg@cs.umn.edu
Cc: cg@cs.umn.edu, interlingua@isi.edu, srkb@cs.umbc.edu, sowa@ranch.poly.edu
In-reply-to: <9409280959.AA11003@ranch.poly.edu> (sowa@ranch.poly.edu)
Subject: Re: Knowledge Representation Problem
Date: Wed, 28 Sep 94 11:12:25 PDT
From: dwig@markv.com
Message-id: <9409281112.aa19380@hermix.markv.com>
Source-Info: From (or Sender) name not authenticated.
Sender: owner-srkb@cs.umbc.edu
Precedence: bulk
John Sowa writes:
> This example is typical of the kind of information that must be
> represented for business problems. Some of the participants will
> also try to put together an engineering/design problem. There is
> no shortage of candidate problems, but it is difficult to find
> something that can be described in a page or two and still be
> representative of the kinds of problems that arise in real life.
It's definitely a good example. In reading it, what occurred to me to be a
logical next step would be to provide examples of how the requirements might
change, and see how and whether the KR systems can accommodate the changes.
Even nicer would be the ability to track the changes, allowing one to review
the course of evolution, and possibly roll back changes that didn't work
out.
Don Dwiggins "If you can't control where you're going,
Mark V Systems, Inc. at least keep track of where you've been."
dwig@markv.com -- Hermann Puterschein