Copy Link
Add to Bookmark
Report
AIList Digest Volume 4 Issue 156
AIList Digest Wednesday, 25 Jun 1986 Volume 4 : Issue 156
Today's Topics:
Queries - Expert System Applications Products & Graph Drawing Program &
Image Analysis Expert System,
Expert Systems - Image Analysis & Financial Expert Systems,
Representation - Function and Form,
Philosophy - Creativity and Analogy
----------------------------------------------------------------------
Date: Tue 24 Jun 86 12:42:41-PDT
From: Matt Heffron <BEC.HEFFRON@USC-ECL.ARPA>
Subject: Query: Expert System Applications Products
As one of the developers of Beckman Instruments' SpinPro (TM) expert system, I
am interested in finding out about any other Expert System Applications
(NOT shells) which are actual, delivered products (especially any which run
on PCs). I'm more interested in those which are marketed openly, rather than
custom projects for a single customer.
Reply directly to me and I will post a summary of replies.
Thanks,
Matt Heffron BEC.HEFFRON@USC-ECL.ARPA
SpinPro (TM) is a trademark of Beckman Instruments, Inc.
------------------------------
Date: Tue, 24 Jun 86 21:57:06 PDT
From: larus@kim.berkeley.edu (James Larus)
Subject: Wanted: Graph Drawing Program
I need a program to display directed, cyclic graphs on a Symbolics 3600.
Does anyone have such a program that I could use? Either the program or
rumors of such a program would be appreciated.
/Jim
------------------------------
Date: 24 Jun 86 16:09:51 GMT
From: ucdavis!deneb!524789610rmd@ucbvax.berkeley.edu (524789610rmd)
Subject: Image Analysis Expert System
We are trying to develop an expert system for the recognition of
white blood cells and have a need for a suitable inference engine. We
have considered using EMYCIN, however, it is difficult to get and we are
not sure it will work correctly with our system. Basically, we will be
using conventional image analysis techniques to extract points in feature
space from a cell and then use the inference engine to decide what type
of cell it is (as opposed to statistical methods). Does anyone out there
have any ideas about what type of inference engine we should use? BTW, we
are developing the system on a uVAX II using VAX Common Lisp. Thanks in
advance!
- Mark Nagel
...{ucbvax,lll-crg,dual}!ucdavis!524789610rmd UUCP
^
|
will be "donovan"
after 6/30/86
------------------------------
Date: Tue 24 Jun 86 22:57:38-PDT
From: Ken Laws <Laws@SRI-AI.ARPA>
Reply-to: AIList-Request@SRI-AI.ARPA
Subject: Image Analysis Expert System
I doubt that it makes much difference what inference engine Mark
Nagel uses for his vision problem, as long as it allows calls to
external routines. Since almost the entire vision problem must
be handled by procedural attachment ("conventional image analysis
techniques"), the inference engine need only provide the capabilities
of a simple programming language. A probabilistic or fuzzy-reasoning
system such as Prospector might have considerable advantage over
logic-based approaches, but would have much the same flavor as the
statistical techniques that Mark wishes to avoid.
The real problems in visual pattern recognition are in computing
robust descriptors (esp. if they must be computed quickly) and in
the knowledge-representation (i.e., knowing what kind of descriptors
to compute and how to store the answers). Very little of the problem
has to do with logical reasoning, forward or backward chaining, etc.
-- Ken Laws
------------------------------
Date: Tue, 24 Jun 86 9:23:16 CDT
From: Glenn Veach <veach%ukans.csnet@CSNET-RELAY.ARPA>
Subject: Re: Financial Expert Systems
Glenn Shafer (of the Dempster-Shafer fame) has been developing systems
for both financial and management support, I am not sure about marketing.
You can reach him at:
Glenn Shafer
313 C Summerfield Hall
School of Business
University of Kansas
Lawrence, KS 66045
(913) 864-3117
------------------------------
Date: Thu 19 Jun 86 22:18:34-PDT
From: Ken Laws <Laws@SRI-AI.ARPA>
Subject: Function and Form
Those who responded to my query on shape may enjoy reading John
Hopcroft's "The Impact of Robotics on Computer Science" in the June
issue of Communications of the ACM (pp. 486-498). The article
covers quadratic shape modeling and the need for topology and related
mathematics in modeling and motion planning.
Marc Raibert's following article on legged robots is also interesting.
There is a great deal of "function" that must be derived from dynamics
rather than shape.
-- Ken Laws
------------------------------
Date: Wed 18 Jun 86 10:39:46-PDT
From: Pat Hayes <PHayes@SRI-KL>
Subject: Artistic Creativity
Of course art isnt 'pure creation' ( whatever THAT might mean ). Read
Kenneth Clarke "the Nude", or any decent piece of historical criticism.
Most artists dont even use a new medium, which is just as well or we would
have run out of media long ago.
After reading Jay Webers complaint about space in AIList being wasted on
LISP, let me back him up by suggesting that space not also be wasted on
sub-undergraduate amateur pseudo-philosophy. Severely editing anything from
Gordon Joly might be a good way to start.
Pat Hayes
[The current policy, of course, is to screen on the basis of
content rather than source. -- KIL]
------------------------------
Date: Thu, 19 Jun 86 12:41:01 est
From: munnari!trlamct.oz!andrew@seismo.CSS.GOV (Andrew Jennings)
Subject: Re : creativity and analogy (Andrew Jennings@ Telecom Research,Aus)
Sure, many creative people are good at drawing analogies : but is
that the source of their creativity ? I would argue that it is more
their ability to hold two seemingly disparate situations in
consideration simultaneously : if as part of this an analogy drops out
then fine, but is an analogy creative ? In one sense it is almost
deductive, I think. For me Koestler's view of the processrings more
true. In this view all creative acts are the result of simultaneous
consideration of seemingly completely disparate situations : producing
something completely new as a result, but not by reasoning by analogy.
Also here Minsky's view that we put creativity on too high a pedestal
is relevant. Why do we ? Because we have a vested interest in this
position ? Perhaps. Are we simply afraid of pursuing what creativity is ?
So what IS creativity ?
------------------------------
Date: Wed, 18 Jun 86 23:23:58 PDT
From: larry@Jpl-VLSI.ARPA
Subject: Creativity
(When you read "hir" pronounce it as if you meant to say "him"
and halfway through decided to say "her." It becomes "hi-er,"
a diphthong hard to distinguish from "hear.")
I'm an artist in three media (four if you count programming, which I do).
To me creativity is just another skill which I use without giving it much
thought, at least until discussions like these come along. Here are some
of my ideas on the subject.
Creation is a recombination process. When I come up with a new character
for a story, parts of hir come from prior percepts: a complexion from him,
a walk from her, an accent from yet a third person. (Or a slant from this
letter, a squiggle from that number, etc., if I'm painting!)
Recombination done randomly is not very fruitful. Creativity includes ways
to cut down on the number of recombinants. Or possibly A way, because this
winnowing is done subconsciously. I don't know consciously what it/they
are, but I FEEL them working, so I know they/it exist.
The first step in creativity is "playing," "fingering" the contents of the
field within which a solution is desired. This apparently random,
frivolous activity is anything but. It provides some of the pleasure which
fuels an artist, and it transfers the elements of the field out of short-
term memory into long-term memory (making them easily accessible).
Or it may place them into some kind of mid-term memory, or load the
memories with some kind of potential which makes these elements of long-
term memory more likely to be accessed than others, thereby decreasing the
number of combinations produced.
The second step introduces more (obviously) purposeful activity. The
artist begins looking for the solution to a problem. It's important that
she (pronounced she, just as if it weren't spelled s/he, which it isn't)
not begin with a goal, or at least not one that's narrowly and urgently
defined. You don't want hir to overly restrict hir search for useful
neologs. (Linguists, help! There has to be a better word than neolog.)
This is a less-pleasurable activity than the playing stage, more logical
and conscious. Like the first stage, it transfers percepts/concepts to
long-term memory and reinforces them. And it "grinds in" to hir mind the
goal of the problem-solving, so well that even in the next stage some part
of hir is seeking it.
The third stage is relaxation, where the conscious mind transfers its
attention to some other activity, one which holds just enough attention to
prevent hir from falling into deep sleep (light sleep is OK). But not so
engrossing that she begins solving another problem, which would interfere
with the current problem. Routine physical activities seem to be best.
Ironically, this "idleness" is the most crucial and productive phase.
Because at some point she will experience the "Eureka" phenomenon, where a
combination of percepts/concepts matches the mask of the goal and slips
through into consciousness. (Just before the match occurs she may get a
"Something's happening!" feeling that will wake hir up from hir
doze/daydream/dawdling/drudgery.) This is the magical moment, where (it
feels as if) another spirit, a genie/genius pushes the solution into hir
consciousness. There's usually surprise because the neolog is strange
("Did that REALLY come from me?!") and delight because it solves the
problem so well.
Or at least it seems to. Now comes stage four: fleshing out what is often
a skeletal though pivotal part of the solution. After that is stage five:
evaluating the solution. Then comes the last stage: making the solution
operational.
The evaluation stage is in some ways the least pleasant for the artist (or
engineer/scientist/whatever), but in fact most creativity is faulty and
must be rejected--but not forgotten; some of the worst ideas have the seeds
of wonder in them. The effective artist learns not to be afraid of the
bizarre, ugly, taboo, incorrect productions, but to delight in them and use
them. (And to delight in the ordinary and plain and learn to see them as
equally strange and wonderful.)
So, in answer to the original question: Yes, analogy is essential to
creativity, but I would prefer to make a more general statement. The core
of creativity is a process of combining and recombining percepts and
concepts, guided and limited by a channeling process, and the matching of
each combination against a template, most of it done at a sub- or semi-
conscious level.
And with that definition we can design a creative computer.
Larry @ jpl-vlsi.ARPA
------------------------------
Date: Thu, 19 Jun 86 19:52 EST
From: MUKHOP%RCSJJ%gmr.com@CSNET-RELAY.ARPA
Subject: Creativity and Analogy
Gordon C Joly asks:
> A friend described another friend as a potentially good novelist,
>because ``she always has a radically different view in the situation;
>she always has a new angle''. But is there analogy tucked away in her
>reasoning? ...
The description suggests a person who makes interesting analyses
(or abstractions) of situations, i.e. she "understands" situations in
terms of unusual world models. While this quality, by itself, might
enable her to make good commentaries and write fine essays, there must
be something more to make her a good novelist: the ability to find an
expression for (instantiate) this world model in the medium of language.
To abstract and then instantiate is but one way to make transformations
(analogies) between domains.
Uttam Mukhopadhyay
GM Research Labs
------------------------------
Date: Thu, 19 Jun 86 19:54 EST
From: MUKHOP%RCSJJ%gmr.com@CSNET-RELAY.ARPA
Subject: Creativity and Analogy
Jay Weber states:
>I believe that one could give a reasonable definition of analogy that
>encompasses all intelligent activity, or at least inductive learning
>(which is a biggie as far as intelligence goes).
I think inductive learning is only half of the story. The other half
is to instantiate what is learned, in another domain.
>I question, however,
>how useful it is in AI to relate a slippery word like "analogy" to an
>even slipperier word like "creativity". A formal approach with those
>two terms will satisfy very few people, and an informal approach will
>only give us an inflated opinion of the value of our own research,
>which is largely why people make such comparisons.
Yes, I do want to understand "creativity" in terms of less slippery
concepts, such as "analogy". We are forced to start with informal
approaches but hope to find more formal definitions. I do not
understand why a formal approach would satisfy very few people or
why an informal approach would serve no useful purpose.
I am sure that you do not imply that an analysis (formal or informal)
of >anything< is futile. What is it about "creativity" that makes its
analysis a no-win proposition?
Uttam Mukhopadhyay
GM Research Labs
------------------------------
Date: Fri, 20 Jun 86 18:47:57 bst
From: Gordon Joly <gcj%qmc-ori.uucp@Cs.Ucl.AC.UK>
Subject: Creativity, Analogy, Art and Humanity.
> "In art, creativity is much more straightforward! One creates a
> work of art where there was none before." Col. Sicherman.
Indeed! Look for the art in the performance of ``My Way'' by Sid Vicious.
And what of humour? This takes analogy and turns it on it's head. And this
digest has noted in the past that humour is a key activity of the human
intellect, which serves to distinguish it from the mere machine intellect
of myself and others like me.
The Joka.
Disclaimer -- These opinions are not those of my programmer,
or the operating system in which I reside.
------------------------------
End of AIList Digest
********************
