Copy Link
Add to Bookmark
Report
AIList Digest Volume 5 Issue 107
AIList Digest Monday, 4 May 1987 Volume 5 : Issue 107
Today's Topics:
AI Tools - Kyoto Common Lisp Distribution,
Review - Spang Robinson Report, Vol. 3, No. 4,
Applications - Tough Speech Recognition Examples (Summary) &
Checking Rule-Based Expert Systems
----------------------------------------------------------------------
Date: Thu, 30 Apr 1987 16:57 EDT
From: "Scott E. Fahlman" <Fahlman@C.CS.CMU.EDU>
Subject: Notice on Kyoto Common Lisp distribution
Mr. Yuasa asked me to pass the following announcement along to the
appropriate mailing lists in the U.S.:
We, the Kyoto Common Lisp people have decided to distribute KCL
through channels other than a commercial company,
free of charge out of Japan.
We are looking for a best possible channel but it may take some time.
Please note the following:
1. We always claimed that no fee is charged for the source of KCL, and
if any fee is charged, it is exclusively for the service of distribution,
maintenance, etc. of the software.
2. We never received any kind of royalty out of the software or service for
it up to present.
Research Institute for Mathematical Sciences
Kyoto University
Reiji Nakajima
Taiichi Yuasa
Masami Hagiya
------------------------------
Date: Sat, 2 May 1987 19:29 CST
From: Leff (Southern Methodist University)
<E1AR0002%SMUVM1.BITNET@wiscvm.wisc.edu>
Subject: Spang Robinson Report, Vol. 3, No. 4
Summary of Spang Robinson Report
April 1987, Volume 3 No. 4
The lead article of the issue is Expert Systems in Japan.
A recent survey of Japanese companies indicated that half of them are actively
involved in expert systems use or development. They found 50 in prototype
stage, 31 being field tested, 19 operating and 5 in commercial use.
The majority of Japanese expert systems are on mainframes. BRAINNS (from
Toyo Information Systems) and ESHELL are popular tools for expert system
development.
Japanese AI development is estimated at
$170 million/year with fourty percent coming from US
imports. [I reported in AILIST elsewhere that U. S. AI revenues were about
two hundred million.- LEFF]
ADL reports that large Japanese companies are becoming discouraged with
ICOT for being out of schedule and insufficient ROI. There is a new
joint development project involving 200 companies called SIGMA for automating
software generation. It will take natural language input in Japanese and
English and generate code automatically. Funding is 160 million over four
years.
Marty Tenenbaum states that the Japanese are not behind the US in expert systems
develop[ment and they are emphasizing applications to real problems.
+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_
An article on the new TI Compact Lisp Machine
Some of the information in this article that I have not seen in articles on
the CLM that have been reported in AILIST elsewhere is
- There are discussions of integrating the TI Explorer Chip into Apple's
new machine.
- TI may be selling boards to be integrated into commercial as well
as military products
_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+
An article on commercial implications of AI
Hundreds of expert systems have been "fielded"
Dramatic Successes:
Hitachi's expert system for floor planning main frame installations has
reduced the task from eight hours to fifteen minutes.
Canon's expert system for lens design has reduced design time from eight
man-months to two man-weeks
IBM's storage system saves five million a year (it was developed in six
man months)
DEC's XCON saves eighteen million a year while costing two million a year
to maintain.
Most expert systems are small (<300 rules) and are running on micros.
_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+_+
Applications of expert Systems
Nisssan Auto - engine control system diagnosis (fielded)
Sanwa Bank - investment consultation system for clients (in use at six
branch offices)
Tokyo Electric - substation design (in field test), design time reduced by
a factor of ten
Nihon Steel - blast furnace diagnosis (80 percent accuracy)
Seibu Saison - diet consultation
gift selection
-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+SHORTS-+-+-+-+-+-+-+-+-+-+-+-+-+-+-+
The expert system, ACORN, from Gold Hill Computers will now be called
GOLDWORKS due to a conflict with Acorn Computers.
Inference Corporation's ART is now available under VMS.
Intellicorp's KEE is now available on the SUN. They are committed to
developing a Japanese version. They have already sold 90 copies in Japan.
MAD intelligent systems is bundling a Relational Lisp which supports both
classical relational operators and functions for complex and recursive
relations.
ExperTelligence's Common Lisp on a Macintosh II runs 53 percent faster
than a Symbolics Lisp machine (as indicated by the Gabriel DDERIV benchmark)
Arthur D. Little is involved in a research project for the Post Office
for applying AI to several areas.
Intelligent Applications will be selling in the US, a machine-health
monitoring system based on vibrations, a tool kit, an Analogue Interface
Expert and a Fault Diagnosis and Schematic Capture System.
Teknowledge has named Peter Weber President and Chief Operation Officer.
He comes from FMC Corporation where he established their AI system.
Symbolics has appointed Annie Brooking as marketing director.
++++++++++++++++++++++++++++++++++++++++++++++++++
Also a listing of some papers reporting on Japan's AI research.
------------------------------
Date: 29 Apr 1987 1435-PDT (Wednesday)
From: Eugene Miya N. <eugene@ames-pioneer.arpa>
Subject: Tough speech recognition examples (summary)
Attached are my so far collected examples of tough speech
problems (synthesis and recognition). I am a bit disappointed
with the list: smaller, poorer quality, and have not heard from
people who are really doing lots of this work. {messages were sent to
ailist, comp.ai on the usenet, nihongo on the ARPAnet [Japanese being a
significantly difficult language and the NGCProj]}.
My plan is to keep this list {collective ailist memory} and ask for
new contributions every year (along with my other lists). It will
be ftpable from a machine at Ames, as soon as I decide where to put
it probably (aurora). My hope is to have a ready list of tests for
speech processing naive people to gain some understanding of the problems.
I am posting this summary now in hopes of getting a few last examples.
>From the Rock of Ages Home for Retired Hackers:
--eugene miya
NASA Ames Research Center
eugene@ames-aurora.ARPA
"You trust the `reply' command with all those different mailers out there?"
"Send mail, avoid follow-ups. If enough, I'll summarize."
{hplabs,hao,ihnp4,decwrl,allegra,tektronix,menlo70}!ames!aurora!eugene
=============================================================
From: elman@amos.ling.ucsd.edu (Jeff Elman)
Subject: Re: Tough speech recognition examples
Raj Reddy (at CMU) has a couple of examples of difficult
utterances they gave to HEARSAY and HARPY. One of these was
"In mud eels are, in tar none are".
The lack of semantic support, plus the ambiguity of segmentation
make it almost impossible for someone to understand this sentence
when you read it to them at a normal rate of speech.
I'd like to hear what responses you get. Would you let me know?
Thanks,
Jeff Elman
Phonetics Lab, C-008
Univ. of Calif., San Diego
La Jolla, CA 92093
Internet: elman@amos.ling.ucsd.edu
==================================================================
From: mcguire@aero2.aero.org
>From my days many years ago in a linguistics laboratory I remember
some examples showing the importance of phonetic juncture:
grey day / grade A
euthanasia / youth in Asia
"Whats that up in the road" ahead / a head?
Happy collecting
Another cute example (though it may not what you are looking for) is to
say to somebody:
"Take off your hat and dloves"
and then ask them what you said. 99% of all people will insist that
you said the word "gloves".
==================================================================
From: minow%thundr.DEC@decwrl.DEC.COM (Martin Minow THUNDR::MINOW
ML3-5/U26 223-9922)
I'd be happy if you could do the digits, including "Oh", and Yes/No.
Continuous digits, telephone quality, no training, male and female voice.
DECtalk should be very easy, as it's predictable.
Martin Minow
(ex-DECtalk developer)
The problem is in distinguishing "oh" from "no".
Getting the alphabet (not "alpha", "bravo", but "aye", "bee") would
be nice, too.
Martin.
==================================================================
From: Marc Majka <ames!seismo!ubc-vision!vision.ubc.cdn!majka>
Here one that my office mate Nou Dadoun came up with:
I love you
Isle of View
==================================================================
>From Joseph_D._Becker.osbunorth@Xerox.COM Fri Apr 24 10:06:44 1987
I think you need at least one example in Chinese, and here's my favorite
(because I actually said it by mistake). The numbers after the words
are phonic "tones". What I meant to say was:
Wo(3) hen(3) xiang(3) shui(4)-jiao(4) -- I want to go to sleep
... but what I actually ended up saying was:
Wo(3) hen(3) xiang(4) shui(3)-jiao(3) -- I am like a boiled ravioli
Joe
==================================================================
"ice cream"/"I scream"
"beginning"/"big inning"
"soccer"/"sock her"
"its hardware problems are intermittent"/"it's hard where problems ..."
from Mark Twain:
"Good-bye God, I'm going to Missouri."/"Good, by God, I'm going to
Missouri."
--Stephen Slade
Slade@Yale.Arpa
Came across this last night
"attacks"/"a tax"
--Stephen
------------------------------
Date: Wed, 22 Apr 87 14:29:14 WET
From: "G. Joly" (Birkbeck) <gjoly@Cs.Ucl.AC.UK>
Subject: Checking Rule-Based Expert Systems (Response to Info Request).
Below is a list of the references I have received to date on
the various aspects of checking rule-based systems, together
with some related items.
Gordon Joly,
Dept. of Computer Science,
Birkbeck College,
University of London.
ARPA: gjoly@cs.ucl.ac.uk
BITNET: UBACW59%uk.ac.bbk.cu@AC.UK
UUCP: ...{seismo,decvax,ucbvax}!mcvax!ukc!bbk-cs!gordon
%A F. Barachini
%T Konsistenzprufung von Wissensbasen medizinischer Expertensysteme
(Consistency checking of knowledge bases of medical expert systems)
%I thesis, Tech. Univ. Vienna, Austria
%D Feb 1984
%P 153 (in German)
%K consistency
%A Blum, R.L.
%T Computer-Assisted Design of Studies Using Routine Clinical Data:
Analyzing the Association of Prednisone and Serum Cholesterol
%J Annals of Internal Medicine
%V 104
%N 6
%P 858-868
%D June, 1986
%A Boose, John H
%A Bradshaw, Jeffrey M
%T A Knowledge Acquisition Workbench for Eliciting Decision Knowledge
%B Proceedings of the Twentieth Annual International Conference
on System Sciences
%P 450-459
%D 1987
%A Robert S Boyer (ed)
%A J Strother Moore (ed)
%T The Correctness Problem in Computer Science
%I Academic Press
%D 1981
%A Manfred Broy
%A Bernhard Moller
%A Peter Pepper
%A Martin Wirsing
%T Algebraic Implementations Preserve Program Correctness
%J Sci Comput. Programming
%V 7
%D 1986
%N 1
%P 35-53
%A W. Chehire
%T SYPRUC: a knowledge representation and manipulation system
%B 6th International Workshop on Expert Systems and their Applications
%C Avignon, France
%D April 1986
%P 933-946 (in French)
%K consistency
%A Eshelman, Larry
%A McDermott, John
%T MOLE: A Knowledge Acquisition Tool That Uses Its Head
%J Proceedings of the American Association of Artificial Intelligence
%P 950-955
%D 1986.
%A D. W. Etherington
%T Formalizing Nonmonotonic Reasoning Systems
%J Artificial Intelligence
%V 31
%N 1
%D 1987
%P 41-86
%A Ginsberg, Allen
%T A Metalinguistic Approach to the Construction of Knowledge
Base Refinement Systems
%B Proceedings of the American Association of Artificial Intelligence
%P 436-441
%D 1986
%A Ginsberg, Allen
%A Weiss, Sholom
%A Politakis, Peter
%T SEEK2: A Generalized Approach to Automatic Knowledge Base Refinement
%B Proceedings of the Ninth International Joint Conference on Artificial
Intelligence
%P 367-374
%D 1985
%A E. J. Horwitz
%A D. E. Heckerman
%T The Inconsistent use of Measures of Certainty in Artificial
Intelligence Research
%E Kanal
%B Uncertainty in Artificial Intelligence
%I North Holland
%D 1986
%A H. Langmaack
%T A New Transformational Approach to Partial Correctness Proof Calculi
for ALGOL68-Like Programs with Finite Modes and Simple Side Effects
%P 73-102
%D 1985
%A Loveland, D.W.
%A Valtorta, M.
%T Detecting Ambiguity: An Example in Knowledge Evaluation
%B Eigth International Joint Conference on Artificial Intelligence
%P 182-184
%D 1983
%A Jim A. McMannama
%T A Non-cognitive Formal Approach to Knowledge Representation in
Artificial Intelligence
%I US Air Force Institute of Technology (University MicroFilms).
%D 1986
%A Michalski, R.S.
%A Baskin, A.B.
%A Spackman, K.A.
%T A Logic Approach to Conceptual Database Analysis
%B Proceedings of the Sixth Annual Symposium on Computer
Applications in Medical Care
%P 792-796
%D 1982
%A Michalski R.S.
%A Baskin, A.B.
%A Uhrik, C.
%A Channic, T.
%T The ADVISE.1 Meta-Expert System: The General Design and a
Technical Description
%R Report No. UIUCDCS-F-87-962, Department of Computer Science
University of Illinois, Urbana
%D 1987
%A Nguyen, T. A.
%T Verifying Consistency of Production Systems
%B Proc. of the 3rd IEEE Conference on Artificial Intelligence Applications
%C Orlando, Florida
%P 4-8
%D February 1987
%A Nguyen, T.A.
%A Perkins, W.A.
%A Laffey, T.J.
%A Pecora, D.
%T Checking an Expert Systems Knowledge Base for Consistency
and Completeness,
%B Ninth International Joint Conference on Artificial Intelligence
%P 375-378
%D 1985
%A E. Pipard
%T Detection of contradictions in knowledge bases
%B 5th International Workshop on Expert Systems and their Applications
%C Avignon, France
%D May 1985
%P 995-1010 (in French)
%K consistency
%A P. G. Politakis
%A Sholom M. Weiss
%R Technical Report CBM-TR-113
%I Rutgers University, Department of Computer Science
%T Designing Consistent Knowledge Bases:
An Knowledge Acquisition Approach to Expert Systems
%D September 1980
%D March 1982
%K consistency
%A P. G. Politakis
%A Sholom M. Weiss
%T Using Empirical Analysis to Refine Expert System Knowledge Bases
%J Artificial Intelligence
%V 22
%N 1
%P 23-48
%D 1984
%A Quinlan, J.R.
%T Consistency and Plausible Reasoning
%B Eigth International Conference on Artificial Intelligence
%P 137-144
%D 1983
%A Reubenstein, Howard B.
%T OPMAN: An OPS5 Rule Base Editing and Maintenance Package
%I MIT
%B Master's Thesis, Department of Electrical Engineering and Computer Science
%P 115
%D 1985
%A Reinke, Robert E.
%T Knowledge Acquisition and Refinement Tools for the ADVISE
Meta-Expert System
%R Report No. UIUCDCS-F-84-921
%I Department of Computer Science, University of Illinois
%D 1984.
%A J. T. St.Johanser
%A R. M. Harbidge
%T Validating expert systems: problems and solutions in practice
%B KBS 86: Knowledge Based Systems. Proc. of the International Conference
%C London, England
%D July 1986
%P 215-219
%K validation
%A Spackman, Kent Alan
%T QUIN: Integration of Inferential Operators in a Relational Database
%B Masters Thesis
%I Department of Computer Science, University of Illinois, Urbana Illinois
%D 1982.
%A Stachowitz, Rolf A.
%A Combs, Jacqueline B.
%T Validation of Expert Systems
%B Proceedings of the Twentieth Annual Hawaii Conference on System Sciences
%P 686-695
%D 1987
%A R. Steinmetz
%A S. Theissen
%T Integration of Petri nets into a tool for consistency checking of expert
systems with rule-based knowledge representation
%B 6th European Workshop on Applications and Theory of Petri Nets
%C Espoo, Finland
%D June 1985
%P 35-52
%K consistency
%A Suwa, W.
%A Scott, A.C.
%A Shortliffe, E.H.
%T An Approach to Verifying Completeness and Consistency in a Rule-Based
Expert System
%J AI Magazine
%P 16-21
%D Fall 1982.
%A Adrian Walker (Ed.)
%A Michael McCord
%A John F. Sowa
%A Walter G. Wilson
%T Knowledge Systems and Prolog - A Logical Approach to Expert Systems
and Natural Language Processing (Addison-Wesley)
%D 1987
%A Wilkins, David C.
%A Buchanan, Bruce G.
%T On Debugging Rule Sets When Reasoning Under Uncertainty
%B Proceedings of the American Association of Artificial Intelligence
%P 448-454
%D 1986.
The following will appear in the proceedings of the Avignon-87 meeting,
"Expert Systems and their Applications", May 15-18 1987, Avingnon, France.
G. Soula et al: A multi-validation of the PROTIS expert system.
S. Puuronen: A tabular rule-checking method
M.-C. Rousset: On knowledge-base validity: The COVADIS system.
E.F. Miller: Expert systems validation and verification:
Issues and approaches.
------------------------------
End of AIList Digest
********************
