Copy Link
Add to Bookmark
Report
AIList Digest Volume 2 Issue 042
AIList Digest Thursday, 5 Apr 1984 Volume 2 : Issue 42
Today's Topics:
AI Tools - Lisp Eigenanalysis Package Request,
Automata - PURR-PUSS References & Cellular Automata Request,
AI Publications - SIGBIO Newsletter,
Expert Systems - Nutrition System Request & Recipe Planner &
Legal Reasoning Systems,
AI Programming - Discussion
----------------------------------------------------------------------
Date: 30 Mar 84 11:07:45-PST (Fri)
From: hplabs!hao!seismo!rlgvax!cvl!umcp-cs!eneevax!phaedrus @ Ucb-Vax
Subject: Lisp equivalent of Linpack info wanted
Article-I.D.: eneevax.103
I was wondering if anybody knows of any packages in lisp that does the same
thing that linpack does (ie. finding eigenvalues, eigenvectors etc.).
But it must do it fast.
My problem is that I need to do some linear algebra stuff but I need to
be able to load it into vaxima (MACSYMA on a VAX running 4.1BSD. If you
have any suggestions I would be very grateful.
Thanks
Pravin Kumar
>From the contorted brain, and the rotted body of THE SOPHIST
ARPA: phaedrus%eneevax%umcp-cs@CSNet-Relay
UUCP: {seismo,allegra,brl-bmd}!umcp-cs!eneevax!phaedrus
------------------------------
Date: 4 Apr 84 21:37:21-EST (Wed)
From: ihnp4!ihuxv!portegys @ Ucb-Vax
Subject: Request for PURR-PUSS reference
Recently someone mentioned a system called PURR-PUSS (I think it was Ken
Laws), in connection with determining the configuration of a finite
state machine based on observation of input-output relationships. I'm
doing some work related to that, and would appreciate references to
PURR-PUSS.
Tom Portegys, Bell Labs Naperville, Ill., ihuxv!portegys
[I ran across PURR-PUSS in J.H. Andreae's "PURR-PUSS: Purposeful
Unprimed Rewardable Robot", Electrical Engineering Report No. 24,
Sept. 1974, Man-Machine Studies, Progress Report UC-DSE/4(1974) to
the Defence Scientific Establishment, Editor J.H. Andreae, Dept.
of EE, Univ. of Canterbury Christchurch, New Zealand, pp. 100-150.
This article describes several applications of the PUSS (Predictor
Using Slide and Strings) learning program, including the identification
of a repetition pattern in a seemingly random H/T sequence. (The
pattern was two random choices followed by a repeat of the second choice.)
References are given to earlier reports in this series.
I also have copies of reports 26, 27, and 28 (Sep. 1975); each has
at least one article on the use of PUSS learning/predicting modules
for the reasoning component in some application. -- KIL]
------------------------------
Date: 28 Mar 84 12:30:21-PST (Wed)
From: decvax!dartvax!lorien @ Ucb-Vax
Subject: Cellular Automata -- Request for pointers
Article-I.D.: dartvax.1024
I was fascinated by the description of "cellular automata" in last
month's Computer Recreations section of Scientific American. The mass
of interacting parallel processes that are described there seem
singularly appropriate for the simulation of phenomona of interest to
AI workers. With complex enough rules of interaction between elements
it seems one could simulate neurons in the brain or the evolutionary
process. I'm aware, from a course I took long ago in Cognitive
Psychology, that psychologists use dynamically interacting models of
this sort.
This note is to request pointers to any research that's currently
being done in this area, specifically as it relates to AI.
Thanks in advance,
--Lorien Y. Pratt
Dartmouth College Library
Hanover, NH 03755
decvax!dartvax!lorien
------------------------------
Date: Wed, 4 Apr 84 00:40:26 EST
From: "Roy Rada"@UMich-MTS.Mailnet
Subject: SIGBIO Newsletter
[...]
As Editor of the ACM SIGBIO Newsletter, I would
like to publish material on AI in Medicine. My
own research focuses on continuity in knowledge
acquisition for medical expert systems. Please
send me whatever you feel might be relevant.
--Roy Rada
------------------------------
Date: 3 Apr 1984 1630-PST
From: Nitsan Har-gil <har-gil@USC-ISIF>
Subject: Expert System for Nutrition
Does anyone know of expert systems dealing with Nutrition (Food, etc.).
Something that you can give a typical daily menu and will respond with
nutritional deficiencies, etc. Thanks in advance, Nitsan.
------------------------------
Date: Wed, 4 Apr 84 15:26:23 EST
From: Kris Hammond <Hammond@YALE.ARPA>
Subject: Re: Recipe Planner
[This is a response to a personal query about Kris' work in recipe planning;
he agreed to let me share it with the list. I would be interested in hearing
about other recipe-based systems, including those in the chemistry domain.
-- KIL]
I have a paper in AAAI83, Planning and Goal Interaction: The use of
past solutions in present situations. My work centers around the notion
of organizing planning knowledge around the interactions between
features rather than around individual features themselves. In the
cooking domain this means the planner has to anticipate the interactions
between different tastes and textures, and search for past plans that
have already dealt with this interaction in the past. The end result
is a system that looks at an input situation, (a request for a dish
that includes many items and tastes) and tries to find a recipe for
an analogous past situation.
The paper is the analysis of an example which uses knowledge of feature
interaction to 1) analyze the original input, 2) index into a useful
plan, 3) suggest the type of modifications that have to be made on
that plan, 4) search for problems in the resulting plan and 5) propose
general solutions to the problems encountered.
I [am now working on] a more general application of the idea of organizing
planning information in terms of goal and plan interaction. [...]
The cooking paper in on YALE-RES <A.HAMMOND.WORK>WOK.MSS.
Thanks for the interest.
Kris Hammond
------------------------------
Date: 1 Apr 84 10:28:53 CST (Sun)
From: ihnp4!utcsrgv!dave@Berkeley
Subject: expert systems and legal reasoning
A recent request asked for information about expert systems and legal
reasoning. I suggest anyone interested in that field get onto Law:Forum,
a discussion group running under CONFER on an MTS system at Wayne
State University in Michigan. Access is free, with computer charges
and Telenet charges being picked up by the Markle Foundation grant
which is funding the project. Most of the major people involved
in developing legal reasoning systems (Thorne McCarty, Layman Allen,
Jim Sprowl, several others) are involved in Law:Forum and participate
regularly.
If you want to get onto Law:Forum, and can be reached electronically,
send me your electronic address:
ihnp4!utcsrgv!dave@BERKELEY (ARPA)
dave.Toronto (CSnet)
ihnp4!utcsrgv!dave (UUCP)
If you have no electronic address, I can't ship out the access information
to you, so send a letter directly to the conference organizer:
Prof. Jennifer Bankier
Dalhousie Law School
Dalhousie University
Halifax, Nova Scotia
Canada (sorry, don't have the postal code handy)
Dave Sherman
The Law Society of Upper Canada
Toronto
(416) 947-3466
------------------------------
Date: Sun, 1 Apr 84 21:50:25 cst
From: Georgp R. Cross <cross%lsu.csnet@csnet-relay.arpa>
Subject: Legal AI Systems
We are developing a model of the Louisiana Civil Code
The representation language is called ANF (atomically
Normalized Form) and is being used to develop the
conceptual retrieval and reasoning system CCLIPS (Civil
Code Legal Information Processing System). Some references
are:
deBessonet, C.G., Hintze, S.J., and Waller, W., "Automated
Retrieval of Information: Toward the Development of a Formal
Language for Expressing Statutes, Southern University Law Review,
6(1), 1-14, 1979.
deBessonet, C.G., "A Proposal for Developing the Structural
Science of Codification," Rutgers Journal of Computers,
Technology and the Law, 1(8), 47-63, 1980.
deBessonet, C.G., "An Automated Approach to Scientific
Codification," Rutgers Computer and Technology Law Journal,
9(1), 27-75, 1982.
deBessonet, C.G.
"An Automated Intelligent System Based on a Model of a Legal
System," Rutgers Journal of Computers, Technology, and the
Law, 10, to appear, 1983.
Technical Reports:
83-011 Formalization of Legal Information
83-023 Natural Language Generation for a Legal Reasoning
System
83-002 Processing and Representing Statutory Formalisms
84-006 Representation of Some Aspects of Legal Causality
83-005 Representation of Legal Knowledge
Copies of the above Technical Reports may be requested from
<techrep%lsu@csnet-relay> or from
Technical Reports Secretary
Department of Computer Science
Louisiana State University
Baton Rouge, LA 70803-4020
George R. Cross Cary G. deBessonet
<cross%lsu@csnet-relay> <debesson%lsu@csnet-relay>
------------------------------
Date: Tue, 3 Apr 84 13:35 PST
From: DSchmitz.es@Xerox.ARPA
Subject: Legal AI research
For all who requested, I am maintaining a copy of all the responses to
my request for information about ongoing AI research in legal-related
fields in the following file: [Oly]<DSchmitz>LegalAI.mail
There are about 15 responses in there now (including those who asked to
be copied on the responses) and I will be adding any new ones I receive
as they arrive.
David
------------------------------
Date: 1 Apr 84 22:35:06 EST
From: Louis Steinberg <STEINBERG@RUTGERS.ARPA>
Subject: Re: Stolfo's call for discussion
One way AI programming is different from much of the programming in other
fields is that for AI it is often impossible to produce a complete set of
specifications before beginning to code.
The accepted wisdom of software engineering is that one should have a
complete, final set of specifications for a program before writing a
single line of code. It is recognized that this is an ideal, not
typical reality, since often it is only during coding that one finds
the last bugs in the specs. However, it is held up as a goal to
be approached as closely as possible.
In AI programming, on the other hand, it is often the case that an
initial draft of the code is an essential tool in the process of
developing the final specs. This is certainly the case with the
current "expert system" style of programing, where one gets an expert
in some field to state an initial set of rules, implements them, and
then uses the performance of this implementation to help the expert
refine and extend rules. I would argue it is also the case in fields
like Natural Language and other areas of AI, to the extent that we
approach these problems by writing simple programs, seeing how they
fail, and then elaborating them.
A classic example of this seems to be the R1 system, which DEC uses to
configure orders for VAXen. An attempt was made to write this program
using a standard programming approach, but it failed. An attempt was
then made using an expert system approach, which succeeded. Once the
program was in existence, written in a production system language, it
was successfully recoded into a more standard programming language.
Can anyone out there in net-land confirm that it was problems with
specification which killed the initial attempt, and that the final
attempt succeeded because the production system version acted as the
specs?
------------------------------
End of AIList Digest
********************
