RE: SUO: Maintenance - related issues
Dear All,
. I think John has aired some important points below,
which hedge around my stance that the benefits of the current thrusts will
be minor compared to that achievable if the SUO structure focuses on the
formats of an unambiguous adaptation of natural language.
. The SUO is currently instead heading down the road of
becoming a highly symbolic ontology that will be very foreign and unreadable
to the average lay person. I think this line of development arises naturally
out of eliciting support from admittedly keen logicians, who understandably
want to stick to their familiar "symbo-logical" structures, because that is
where they feel at home. However, I feel this means it will remain buried in
the secret files of industry designers and software developers, and probably
not even of managers.
. On the other hand, the general public probably doesn't yet
even realise the potentials of such leveraging of what can loosely be called
"human - machine interaction", let alone the potential benefits of
standardisation in this area. So, perhaps I am championing a path ahead of
its time.
. As someone who first developed knowledge-based decision
making software a quarter of a century ago, I am somewhat dismayed by the
fact that neither of the above reasons constitutes a technological barrier
to achieving the more beneficial end at this time. Perhaps, too, I am
proposing something beyond the purview of IEEE, etc., though I don't really
believe this should be so.
. The main benefit I see from a "natural language based SUO"
would be a greater clarity of logical implications of situations and devices
(both physical and conceptual) to the average person across societies and
populations around the globe. I would include in conceptual devices those
that are educational, legal and sociological, so I suggest the benefits
would extend right across society.
. A small example of the benefits I could see from a "natural
language based SUO" could be a release from the frustrations of the current
unintuitive and inconsistent behaviour of Windows, which many of us have to
tolerate every day. This would arise because such an ontology would make it
much more evident to far more people that the current structure of that
application is far from tidy. The market rejection and complaints would be
more intense than the frequent cursing of it many currently do, if under our
breaths.
. What will be the consequence? Basically I feel the current
work will form a relatively clunky SUO that will be later supplanted by a
more natural one that ordinary people will relate to far better. What I
wonder is how soon afterwards such a more natural SUO will arise, along with
the benefits it will bring.
. Anyway, enough of this hobby horse of mine. I'd better get
back to other things.
Cheers Graham Horn
Australian Institute of Health and Welfare
================================================
Phone: 02.6244.1094
Fax: 02.6244.1199
Email: Graham.Horn@aihw.gov.au <mailto:graham.horn@aihw.gov.au>
-----Original Message-----
From: John F. Sowa [mailto:sowa@bestweb.net]
Sent: Tuesday, October 17, 2000 9:53 AM
To: Josiah Lee Auspitz; Frank Farance; standard-upper-ontology@ieee.org
Subject: Re: SUO: Maintenance
Lee, Frank, et al.,
There are many complex issues involved in the recent correspondence, but I
would like to emphasize the following paragraph of Lee's recent note:
>3) seek alliances and other avenues for development in the area, and
self-situate the SUO effort in a niche relative to other efforts. This I
take to be John Sowa's point not only in his plea for greater cooperation
with international efforts (below) but also in his previous requests that we
not cut off the old lines of development in NCITS/ANSI/ISO and the
university/foundation world-- lines which succeeded in placing in the public
domain, and on the SUO archive, previously proprietary materials from IBM,
Cyc, Mikrocosmos, the Japanese EDS, and others. I have sensed an
undercurrent in the email traffic between Frank Farance and Bob Spillers
that there has been an effort make the SUO in IEEE the unique focus for
American sponsored ontologies standards work by damping down the
>previous lines of development not only with respect to ontologies but also
>in related areas. To the extent the squealching of other avenues is
conscious and successful, as I gather it already has been to some degree,
the SUO effort will diminish rather than extend American-sponsored efforts
>in this genuinely international and public arena.
The ontology project is potentially immense. Cyc has been working on it for
16 years. The Japanese EDS project spent many billions of yen. It is not
something that can be "fast-tracked", but it is something that involves many
pieces, some of which are closer to completion than others.
There are also a large number of very smart people who have been working on
important pieces of the puzzle, but who do not subscribe to the current SUO
mailing list. It is important to get them involved as critics, supporters,
contributors, and reviewers. But we have already reached a point where our
mailboxes are overflowing with just the current level of participation.
I don't know what the best organization should be, but I do know that the
ultimate goals will take a lot of money, people, time, and work. And I
believe that we should keep our options and contacts open.
John Sowa
-----Original Message-----
From: Horn, Graham
Sent: Wednesday, October 11, 2000 1:27 PM
To: 'sowa@bestweb.net'; Thompson, John A;
standard-upper-ontology@ieee.org
Subject: RE: SUO: KIF & Naming Problems
Johns et al,
. I can't help but be struck by the fact that we can
communicate succinctly and unambiguously in the English language if we take
sufficient care to. Having said that, I have found some of the SUO dialogue
lacking some of that very care, leading to occasional requirements for
clarifications, etc. I feel this is often because one leaves much of what
one wants to communicate implied, and the recipient doesn't always catch the
implication.
. I support John's proposition that: "The only satisfactory
solution is to adopt a standard naming convention for SUO.", as the only
really practical option.
. Here is a proposition for people to validate or otherwise,
to provide something very close to what we want.
1st selecting the main direct (as opposed to metaphorical) meaning of
each word we need from the OED, and requiring all other meanings to be put
in similar ways to the way those meanings are explained on that august
(better replace this word by "revered") reference;
2nd adopting inflections to distinguish noun, verb, adjective, adverb
(and if necessary any other parts of speech), to accommodate words often
used in more than one such role with the same spelling;
3rd tightening up those definitions that are imprecise, and here again
requiring the thereby excluded interpretations to be denoted by elaboration;
4th in certain cases, adopting specific technical meanings for
particular words, though in most cases of common words it would be
preferable to adopt distinctive spellings (much as advertisers and trade
mark registrants often do in order to avoid loss of such marks through
common usage - I'm sure most of know about the famous "Biro" case); and
5th formalising, and possibly restricting, allowable grammatical
structures, most of which would be pre-existing.
. Of course this would require all metaphors to be expressed
as, at least, similes, and more often directly described. However, I suspect
a lot of misunderstandings arise from differing foci on the application of
metaphors, so I feel this is likely to a requirement for ensuring precision
anyway.
. In addition to machine interpretability and the other claims
for the SUO, a significant additional benefit of such an approach would
greatly facilitate human interpretability: It could even facilitate
automated human speech and document interpretation and comprehension.
. Of course, those of us who are of the baby boom generation
will recall that such an approach is reminiscent of the language and grammar
we were taught at school, in that very structured and rule-bound era. I
certainly recall, long ago, parsing sentences and adding in the implied
words in brackets, so the sentence had the required grammatical structure,
as a means of determining whether the sentence was grammatically correct or
otherwise.
. This in no way detracts from the requirement for sorting out
the requirements for allowable expression logics that many participants are
working on.
. I would be very interested in any reason why such an
approach would not work, and why it may not considerably simplify our task.
Graham Horn
Australian Institute of Health and Welfare
================================================
Phone: 02.6244.1094
Fax: 02.6244.1199
Email: Graham.Horn@aihw.gov.au <mailto:graham.horn@aihw.gov.au>