[Fwd: Re: SUO: Category Theory]
FYI.
-------- Original Message --------
Subject: Re: SUO: Category Theory
Date: Fri, 24 Jan 2003 00:25:01 -0500
From: Leo Obrst <lobrst@mitre.org>
Organization: The MITRE Corporation
To: "John F. Sowa" <sowa@bestweb.net>
CC: "Robert E. Kent" <rekent@ontologos.org>,Jim Farrugia
<jim@spatial.maine.edu>
References: <3E22CEE5.7CCA5F88@oakland.edu> <3E236B88.20200@bestweb.net>
<002501c2bc26$b4b2efc0$7b76e4ce@Dell> <3E254ED7.2000508@bestweb.net>
I understand your concern, John. You are not alone. One issue here is
that
Bob Kent is really performing RESEARCH that will potentially affect us
all.
Yes, it is in the guise of a candidate for an emerging standard, and
hence
should not ideally be research, but really that is also the case with
the
other candidates SUMO and OpenCyc: they were or have been research.
They
are not really results that attempt to "codify and systematize the best
engineering practices" -- because there are not yet any engineering
practices.
All of the candidates attempt to leap full blown from the head of the
Zeus
of research into "best engineering practices" because we all know we
need
these best practices now, before we know exactly what they are.
What Bob is doing is a huge intellectual effort, which in other possible
worlds, would be fully funded and supported (to me, he must be
independently wealthy to pursue such an effort). However, there are no
government agencies -- comparable to our SUO membership -- who can
understand his effort enough to appreciate it. But Bob is working like
older maniac scholars and attempting to push the boundaries. Is it good
and
right? I don't know. I do know that in principle it is a reasonable
approach, and addresses tough mapping issues we in general dither about.
It
attempts to systematize the kinds of mappings we know we need for the
eventual "theory of lattices": mappings between theories, between
interpretations, between theories and interpretations (and sub-portions
of
these ala microtheories in Cyc). But it is a lot of formal apparatus to
swallow. The only comparable research I know about (besides Barwise &
Seligman's Information Flow Theory) is the Local Model Semantics (an
attempt to formalize context) of:
Giunchiglia, Fausto; Ghidini, Chiara. 1998. Local Models Semantics, or
Contextual Reasoning = Locality + Compatibility. Principles of Knowledge
Representation and Reasoning (KR'98), Proceedings of the Sixth
International Conference, Trento, Italy, June 2-5, 1998, Anthony Cohn,
Lenhart Schubert, Stuart Shapiro, eds., pp. 282-289.
And related citations. We wrote two papers ourselves focusing on
context/semantic interoperability in 1999, that we tangentially related
to
the above and the IFT):
Obrst, L., G. Whittaker, A. Meng. Semantic Context for Object Exchange,
AAAI Workshop on Context in AAI Applications, Orlando, FL, July 19,
1999.
Obrst, L., G. Whittaker, A. Meng. Semantic Interoperability via Context
Interpretation, submitted to Context-99, Trento, Italy, April, 1999.
If I had much more time, I would be a stronger advocate for IFF, and
trying
to "popularize" it, since I roughly understand category theory and
roughly
understand B&S's IFT, though not to Bob's detail. I am a poor advocate:
it
would take much more time and energy to be a good advocate. Again,
should
you swallow the whale? No. Does it make for a good standard as far as we
understand standards: no. Is it worthwhile? I would say, yes.
Leo
ps. I will post this also to the SUO list, if Bob desires: Bob, just let
me
know.
"John F. Sowa" wrote:
> Robert,
>
> I haven't read all of your papers, but I have read or skimmed enough
> of them to realize that you are doing some very important work,
> which I strongly urge you to continue.
>
> > In the meantime I have started an FAQ located at
> > http://suo.ieee.org/IFF/FAQ.html and linked off the SUO
> > IFF main page http://suo.ieee.org/IFF/
>
> This is very good, and it should be very much appreciated by
> everybody who wants to know what is happening with the project.
>
> What you are doing is entirely appropriate for cutting-edge research,
> and I believe that you are on the right track. However, a good
> standard is never based on cutting-edge research. The purpose of
> a standard is to codify and systematize the best engineering practices.
>
> The most important task right now is to begin developing some
> engineering experience in applying the theory to real-world problems.
> That work would require an implementation that is sufficient to
> support the engineering efforts.
>
> Right now, the most important task is to begin the development
> of an implementation. That first implementation will almost certainly
> be a prototype that is intended to be thrown away or totally rewritten
> before the standard is finished. But without it, it is impossible
> to finish the standard.
>
> John Sowa
--
_____________________________________________
Dr. Leo Obrst The MITRE Corporation
mailto:lobrst@mitre.org Intelligent Information Management/Exploitation
Voice: 703-883-6770 7515 Colshire Drive, M/S H305
Fax: 703-883-1379 McLean, VA 22102-7508, USA