LAWS@SRI-AI.ARPA (04/17/85)
From: AIList Moderator Kenneth Laws <AIList-REQUEST@SRI-AI>
AIList Digest Wednesday, 17 Apr 1985 Volume 3 : Issue 47
Today's Topics:
Bindings - Walter Reitman,
Request - Advanced AI Programming Topics,
AI Tools - RuleMaster & Lisp for IBM PC & GCLISP Review,
Expert Systems - Articles & Comparative Report,
Publications - Flores' & Lisp Conferences Proceedings,
Conference - IJCAI-85 Registration & Housing,
Seminars - A Relational Database with Logic Programming (SRI) &
Phrase Structure and Parsing (BBN) &
Expert Systems at Lockheed (SU) &
The Japanese Lisp Machine (SU)
----------------------------------------------------------------------
Date: Wed, 3 Apr 85 15:43:54 EST
From: David_West%UMich-MTS.Mailnet@MIT-MULTICS.ARPA
Subject: Bindings - Walter Reitman
Walter Reitman, whose address was given as here in AIlist v3 #42,
in fact moved east a few years ago. We are currently forwarding
his mail to 25 Oak St., Rumson NJ <no zip available>
I am not sure whether he continued to work on Go after leaving here.
David West (David_West%umich-mts.mailnet@mit-multics.arpa)
MHRI, University of Michigan, Ann Arbor.
------------------------------
Date: 16 Apr 85 16:31:38 EST (Tue)
From: Carlo J. Rodriguez <rodrigue@mitre.ARPA>
Subject: Advanced AI Programming Topics
Hello! I was wondering if I could stir up some interest on the part of
some of the more experienced programmers and designers that follow the
AIList to share some of their knowledge of programming techniques and
structures, data structure implementations and favorites tricks, shortcuts
and other similar tidbits with those among us less experienced in this
magic. Topics such as information hiding, graph structure applications,
parsing techniques and other similar methods typically employed in AI
programming are the types of things I'd like to see discussed. It would
be nice to see some real hard-core, nitty-gritty technical information
exchange.
Let us all know what you think about this topic.
--- Carlo J. Rodriguez
aka: carlor@mitre
------------------------------
Date: 10 Apr 85 15:37:28 EST
From: Mary.Lou.Maher@CMU-RI-CIVE
Subject: RuleMaster
I recently attended a training session for RuleMaster at Radian
Corporation. RuleMaster is an expert system development tool that allows
the programmer to put the knowledge in the form of examples and RuleMaster
will induce the rules. RuleMaster essentially accepts rules in the form
of decision tables and induces an efficient IF THEN representation.
The decision tables form a hierarchy of modules and allow variable
declaration and access to external programs or databases. RuleMaster
was developed with consultation problems in mind and may possibly be
used for robotic control. I found a few limitations that would make it
difficult to use RuleMaster for design problems (one of which is the
requirement that you have to declare all your variables before execution).
------------------------------
Date: Tue, 16 Apr 85 10:22:42 BST
From: Fitch@ucl-cs.arpa
Subject: Lisp for IBM PC
The number is growing, but I reckon the best is UOLISP from Jed Marti
(Marti@Rand-unix on arpa net). It has a compiler, screen editor and other
goodies. I have seen it run a large part of the REDUCE algebra system, which
for a PC is good going.
Alternatives are Golden Common Lisp and MuLisp (Stoutemyer in Hawaii)
------------------------------
Date: Tue 16 Apr 85 12:53:10-PST
From: Ken Laws <Laws@SRI-AI.ARPA>
Subject: GCLISP & Expert Systems Articles
For information on GCLISP, see Jonathon Amsterdam's review in the May
issue of Popular Computing. The same issue also features overview
articles about AI and expert systems, and a review of four AI books.
(Mike Nicita and Ron Petrusha liked Expert Systems by Paul Harmon
and David King, didn't like The AI Business by Winston and Prendergast,
and had mixed comments about Into the Heart of the Mind by Frank Rose
and The Cognitive Computer by Schank and Childers.)
For a business view of expert systems, see Exploring Expert Systems,
by Elisabeth Horwitt, in Business Computer Systems, March 1985, pp. 48-57.
The article concludes with a list of vendors of expert-system hardware,
software, and services.
-- Ken Laws
------------------------------
Date: Fri 12 Apr 85 14:55:27-PST
From: Mark Richer <RICHER@SUMEX-AIM.ARPA>
Subject: report on ES tools
I have written a report (for a class) on five software tools for building
knowledge-based systems: ART, DUCK, KEE, S1, and SRL. The report is based
on written information I have collected and demos I have seen of KEE and S1.
(Unfortunately no hands-on experience). Nevertheless, several people have
found the paper imformative so I am offering it to others. Send me your
US mail address if you are interested. I will wait a little while and if
there aren't too many requests I'll mail you a copy. Otherwise, I'll make
arrangements to post somewhere on the net where you can ftp it. (Actually
it was written with Macintosh word, and another alternative is to send me
a floppy with a self-addressed mailer).
mark
------------------------------
Date: Tue, 16 Apr 85 09:30:57 PST
From: Charlie Crummer <crummer@AEROSPACE.ARPA>
Subject: Flores Publications (AIList Digest V3 #46)
Reply to Paul Cashman: More by Flores that you might not have seen:
"Management and Communication in the Office of the Future" (dissertation)
'Understanding Computers and Cognition' with Winograd
You can get these through Hermenet Inc. in the SF Bay area (415) 474-3400.
--Charlie
------------------------------
Date: 11 Apr 85 1225 PST
From: Yoni Malachi <YM@SU-AI.ARPA>
Subject: Lisp conferences proceedings
[from SIGACT News]
ACM SIGPLAN has republished the conference proceeding of
previous Lisp conferences.
Price
Order No. Members Others
1980 Lisp Conf. 552800 $15 $21
1982 Lisp Conf. 552820 $18 $26
1984 Lisp Conf. 552840 $20 $27
Ordering address (prepaid)
ACM Order Dept.
P.O.Box 64145
Baltimore, MD 21264
------------------------------
Date: 09 Apr 85 09:38:31 PST (Tue)
From: Phil Klahr <klahr@rand-unix>
Subject: IJCAI-85 Registration -- Please post
The International Joint Conference on Artificial Intelligence will be
meeting in Los Angeles (at UCLA) August 18-23, 1985. Conference
brochures (including registration information) have already been mailed
out. If you have not received one, or would like extras, contact
IJCAI-85
c/o AAAI
445 Burgess Drive
Menlo Park, CA 94025
415-328-3123 or 415-321-1118
Registration will be limited to 5,000 people. Based on early projections,
up to 7,000 people may wish to attend, so early registration is highly
encouraged (if not necessary).
As a bonus, early registrants will receive a substantial reduction in
registration costs. Through June 28, registration fees are $175 ($80 for
students); for registrations received after June 28 but prior to July 26,
fees will be $225 ($100 for students); and for on-site registration (if
available), fees will be $275 ($125 for students). Substantial reductions
for early tutorial registrations are also in effect.
Further information on the technical conference, the tutorials, the
exhibition, and housing can be found in the conference brochure.
------------------------------
Date: Thu, 11 Apr 85 14:01:06 PST
From: Phyllis O'Neil <oneil@AEROSPACE.ARPA>
Subject: IJCAI-85 housing
The UCLA Guest House is booked up for IJCAI-85.
Many residential suites and residence hall rooms are available, as well
as hotel rooms. Mail your IJCAI Housing Bureau Reservation Form with
the deposit as soon as you can, to be assured a room for IJCAI-85.
------------------------------
Date: 11 Apr 1985 1524-PST
From: GOGUEN@SRI-CSL.ARPA
Subject: Seminar - A Relational Database with Logic Programming (SRI)
[Forwarded from the SRI bboard by Laws@SRI-AI.]
CSL SEMINAR SERIES: 17 April 1985 at 2pm in EL381
EXTENDING A RELATIONAL DATABASE WITH LOGIC PROGRAMMING FACILITIES
H. van Emde Boas - Lubsen, IBM INSDC, Uithoorn Netherlands
P. van Emde Boas, IBM San Jose, on leave from IIW/FVI, Univ.
Amsterdam
C.F.J. Doedens, Univ. of Amsterdam, while at IBM-INSDC Uithoorn
Netherlands
The observation that there exists a strong connection
between the Relational Database Model on the one side, and a
convincing interpretation of Logic Programs on the other
hand has been made frequently. The real issue is how to
exploit this connection in order to let Logic Programming
share the fruits of years of research and development in
Database technology.
Various groups have worked on the problem of interfacing
Logic Programming and Relational Databases by giving the
logic programming system access to an existing Database
system, or by incorporating Database facilities into the
Logic Programming System. Our approach is radically
different: we incorporate Logic Programming Facilities into
an existing Database system by extending the functionality
of the latter.
In the presentation we illustrate this strategy by defining
a simple subset of PROLOG. Our prototype system translates
programs from this subset into high level code for the
Database System Business System 12, a commercially
available relational Database system for time sharing use
developed at IBM-INSDC, Uithoorn, the Netherlands. Facts
are translated into rows which are inserted in Database
tables. Clauses are compiled into View definitions and goals
are represented by Database queries. The existing BS12-API
language provides a powerful view mechanism, allowing us to
design this translation. The query generated by compiling
the goal statement is processed by the BS12 system, and all
answers generated are returned to the Prolog system without
further need for interfacing.
The subset of PROLOG presently covered by our system
excludes several important facilities. They will be
discussed together with the required enhancements both to
our interface and the existing BS12 system needed for
incorporating these features. We also compare our work with
projects performed by others.
Disclaimer
This presentation contains reference to, or information
about, possible future enhancements to IBM products
(Business System 12). Such reference or information must
not be construed to mean that IBM intends to implement these
enhancements. The contents are entirely the responsibility
of the authors and reflect their personal opinion.
------------------------------
Date: 11 Apr 1985 12:17-EST
From: AHAAS at BBNG.ARPA
Subject: Seminar - Phrase Structure and Parsing (BBN)
[Forwarded from the MIT bboard by SASW@MIT-MC.]
BBN's next AI seminar is at 10:30 on Thursday, April 18, in the 3rd
floor large conference room at 10 Moulton Street. Craig Thiersch, of
the University of Connecticutt and the University of Koeln, will speak
on "Scrambling, the "VP node" and the nature of projections". His
abstract:
This talk consists of two parts: the first deals with several
linguistic problems, illustrated by a specific example from the
structure of German; the second discusses the effect which the
resolution of these problems has on (1) the nature of phrase-
structure, and (2) applications to parsing.
In the past, parsers have quite generally assumed that one needed
separate "packets" for different constituent types, NP, VP, S, etc.
which had rather different structures from one another, in spite of
attempts to reduce them to a common structure.
If my arguments are correct, there is only one constituent type,
XP, and its structure is uniform; the "satellites" of the Head are
licensed by it and given interpretations which vary little from
constituent to constituent. This isolates and minimizes the
difference between constituents, and locates the few differences where
they belong, namely in the lexicon.
The parser itself, on the other hand, is the same for all
constituents, and "expects" the same structure for all. This means
that given a particular place in the parse tree, the parser has a more
complete set of expectations. This presumably would be particularly
useful in processing fragmentary or garbled utterances, as in speech
rather than text processing.
------------------------------
Date: Fri 12 Apr 85 09:00:51-PST
From: Elliott Levinthal <LEVINTHAL@SU-SIERRA.ARPA>
Subject: Seminar - Expert Systems at Lockheed (SU)
[Forwarded from the SRI bboard by Laws@SRI-AI.]
The next seminar will take place Wednesday, April l7, in Terman 2l7
from 2:l5 - 3:30.
Speaker: J.R. Zumsteg
Advanced Software Laboratory
Topic: "An Overview of Engineering Expert Systems Research
at Lockheed Palo Alto Research Laboratory
Expert system technology has progressed to the point where it can
be used to assist engineers and designers during the design and
analysis of aerospace structures. Because structural design and
analysis relies heavily on computer-based modelling, an engineering
expert system must be able to utilize existing application programs.
The software interface required to provide this capability must set
up the input for the program, run the program, and extract the
needed results from the output of the program, in much the same
manner as a human expert would use the progeram. In addition,
the knowledge base of the expert system must include facts and
rules to decide when and how to use an application program during
the course of a design or analysis.
Two engineering expert systems have been under development at the
Lockheed Palo Alto Research Laboratory for over a year. The
Composites Design Assistant (CDA) uses a PROLOG-based expert
system framework, which has been interfaced to the RIM relational
database manager and the ADVLAM laminate analysis code, to assist
an engineer in the design of sandwich panels. The Buckling Expert
uses the Lockheed Expert System (LES), an expert system framework,
written in analysis code and the PANDA shell design code during
the analysis of a cylindrical shell. The design and operation
of these engineering expert systems, along with the software and
logic issues encountered during their development, will be
described. An example of each system in use will also be presented.
------------------------------
Date: Mon 15 Apr 85 16:03:46-PST
From: Carol Wright <WRIGHT@SUMEX-AIM.ARPA>
Subject: Seminar - The Japanese Lisp Machine (SU)
[Forwarded from the Stanford SIGLUNCH distribution by Laws@SRI-AI.]
DATE: Friday, April 19, 1985
LOCATION: Chemistry Gazebo, between Physical & Organic Chemistry
TIME: 12:05
SPEAKER: Shigeki Goto
Musashino Electrical Communication Laboratory,
NTT, Japan
Visitor SU Computer Science Department
TITLE: Japanese Lisp Machine ELIS and it's Language TAO
NTT is a Japanese telephone company. Electrical Communication
Laboratories are often called "Japanese Bell Labs". At the Feburary
15 Siglunch, Professor Feigenbaum spoke about NTT's new Lisp machine
being ten times faster than Symbolics 3600. This talk is a follow
up on his talk, and will cover the following points:
1) A brief overview of NTT, and Research activities at Electrical
Communication Laboratories.
2) Lisp Machine ELIS: It is safe to say that ELIS is at least six times
faster than Symbolics 3600. Because the interpreter is implemented
fully by microcode, interpreted code runs faster than most dedicated
machines.
3) A dialect of Lisp: TAO
Although TAO looks like Zetalisp, it takes in features of Prolog and
Smalltalk. Users can program by selecting and mixing the programming
paradigms.
------------------------------
End of AIList Digest
********************