LAWS@SRI-AI.ARPA (06/05/85)
From: AIList Moderator Kenneth Laws <AIList-REQUEST@SRI-AI>
AIList Digest Wednesday, 5 Jun 1985 Volume 3 : Issue 73
Today's Topics:
Queries - AI and Proctology & Range Data Images for Robotics &
Explorer/Symbolics Compatibility & LISP on SUN Workstations &
Lisp User Survey & AI and the CAIS & Man-Machine Interaction,
Natural Language - McDonnell Douglas Announcement,
Expert Systems - Expert Systems vs "Conventional" Programming,
Conference - IJCAI-85 UCLA Campus Housing Available
----------------------------------------------------------------------
Date: Sat, 1 Jun 85 22:34:36 PDT
From: Michael Pazzani <pazzani@UCLA-LOCUS.ARPA>
Subject: AI and proctology
I am about to start my Phd research on an expert system for proctology.
I am interested in finding out what others have done in this area.
Also I'd like to have some opinions on the best language for implementing
expert systems: PROLOG EURISKO or OPS5?
Thanks,
Mike Pazzani
------------------------------
Date: Mon 3 Jun 85 17:02:30-MDT
From: Tom Henderson <Henderson@UTAH-20.ARPA>
Subject: Range Data Images for Robotics
The IEEE Computer Society Technical Committee on Robotics has asked me to look
into the creation of a database of range data images. We would like to get an
idea of both the availability of range imagery and the demand for such imagery.
If you have range data images (perhaps with registered intensity images) or you
would like to have access to such images, please answer the following questions
and send to henderson@utah-20.
Name:
Address (Net and US Post):
Range Data Available (give format):
Would like access to range data (give any special requirements):
------------------------------
Date: Tue, 4 Jun 85 13:18:21 edt
From: Mark Weiser <mark@tove>
Subject: Explorer/Symbolics compatibility
We have a Symbolics 3600, and are looking at Explorers from TI now.
What is the software compatibility between these two? Is it the very
same zetalisp, windowing, etc.? Or is the only software similarity
that of ancient common ancestry? Thanks for any help or advice.
-mark
------------------------------
Date: Thu, 30 May 85 14:41:20 BST
From: Ewgorc@ucl-cs.arpa
Subject: LISP on SUN Workstations
Can anyone gine me information about LISP availability on the SUN
workstation range?
I am interested in what variants are available, COMMON Lisp, Inter Lisp
or whatever.
Comments from actual users on how they have found the implementation would be
very useful.
------------------------------
Date: Mon, 03 Jun 85 20:32:36 CDT
From: marick@GSWD-VMS
Subject: Lisp user survey
I'm making a survey of what users need from Lisp systems. If you use Lisp
for work other than developing Lisp itself and you would mail me answers to
the following questions, I'd much appreciate it. I will of course summarize
the results to the net.
1. (I hesitate to start this battle.) What do you consider the most
cost-effective system on which to develop programs? Special-purpose
Lisp machines? General purpose personal workstations with Lisp
(e.g., Suns, Apollos)? Timeshared systems? What about the most
cost-effective system on which to deliver programs?
2. How important is the ability to call functions written in another
language? If it's important, will those functions be specially written
for the application or will they tend to be parts of libraries like
the IMSL math library or a plotting package? If they're specially
written, are they making up for some deficiency in your Lisp (e.g., no
or weak support for handling strings)?
3. Should the editor be a part of the Lisp environment or should it be
separate? What are the important differences between an embedded
Emacs-like editor and an external Emacs-like editor that allows you
to run Lisp in a window?
4. How often could your programs run "standalone", without the support of
the full Lisp environment? Would it be useful to have some way to dump
a set of Lisp functions as an executable file, so that you could, for
example, write a reasonably-sized UNIX filter in Lisp?
5. How important is some sort of object-oriented programming package,
like Flavors or Loops? Do you have any preferences?
6. How important is the speed of your compiled code?
Obviously, everyone wants a Cray on their desk, but how many of your
programs are truly limited by machine speed or compiler quality and
would be significantly more useful if they ran faster? What operations
are most common in your tight inner loops?
7. How adequate is the standard set of debugging tools (trace,
single-stepper, debugger that lets you poke around in the stack)?
8. What other things make a Lisp system better?
If there's a question I haven't asked, please tell me.
Brian Marick
Gould Computer Systems -- Urbana
USENET: ...ihnp4!uiucdcs!ccvaxa!marick
ARPA: Marick@GSWD-VMS
------------------------------
Date: 3 Jun 1985 08:10-PDT
From: JWOLFE@USC-ECLB
Subject: AI and the CAIS
[Also forwarded by Larry@JPL-VLSI.]
Dear Colleagues: The Institute for Defense Analysis (IDA) is
investigating the Common APSE Interface Set (CAIS) for Ada
Programming Support Environments (APSE). Although it is a
proposed DoD standard, it is unlike other standards in that it
will evolve as interface needs are identified. My particular
interest is in what needs are generated by the introduction of AI
languages and techniques (e.g. LISP, PROLOG, Knowledge Bases,
Inferencing, etc.) to software engineering environments. This
issue may effect the CAIS at two levels: First, tools may be
written using AI languages and techniques; second, applications
may use AI languages or techniques which may require unique
tools. I am coming to the Ada and AI communities to solicit your
help and advise. I would appreciate hearing from anyone who has
experience in:
* Using AI languages and techniques to develop software
tools.
* Developing tools for the life cycle maintenance of
AI software.
* Anyone else who can contribute.
Since the results of this study will influence the evolution
of the CAIS standard, it is important to have the participation
of government, industry, and acedemia. Please note that since
IDA is a FCRC, issues of proprietary information and
non-disclosure CAN be resolved. Your participation is needed and
appreciated.
James Wolfe
Institute for Defense Analysis
1801 N. Beauregard Street
Alexandria, VA 22311
ARPA: JWOLFE@USC-ECLB.ARPA
Voice: (703) 845-2109
------------------------------
Date: Tue, 4 Jun 85 17:31:28 edt
From: Rob Jacob <jacob@nrl-css>
Subject: Query - AI and man-machine interaction
I'm trying to put together an entry on "man-machine interaction" for an
encyclopedia of artificial intelligence. My initial reaction is that the
main areas are natural language, speech, and maybe intelligent CAI-type
dialogues. Another area might be user interface techniques first pioneered
by AI people (as a byproduct of their work), like the Interlisp-D user
interface.
Have I missed your favorite topic? I'd appreciate any comments, pointers,
suggestions, vicious attacks, etc. on this topic. Please reply to me
directly, because I'm often too far behind in reading the lists.
Thanks in advance,
Rob Jacob
ARPA: jacob@nrl-css
UUCP: ...!decvax!nrl-css!jacob
SNAIL: Code 7590, Naval Research Lab, Washington, D.C. 20375
------------------------------
Date: 4-Jun-85 12:26 PDT
From: Kirk Kelley <KIRK.TYM@OFFICE-2>
Subject: McDonnell Douglas announcement
McDonnell Douglas has announced a "breakthough" in Natural Language processing
on the back cover of this week's Business Week. Of course there are no
specific references. Does anyone know what they are talking about or if there
is any substance to the hype:
... practical thinkers at our McDonnell Douglas Computer Systems Company
have created the first computer that accepts you as you are -- human.
They emulated the two halves of the human brain with two-level software:
One level with a dictionary of facts and a second level to interpret them.
The resulting Natural Language processor understands everyday conversational
English. So it knows what you mean, no matter how you express yourself. It
also learns your idiosyncracies, forgives your errors, and tells you how to
find what you're looking for.
Now, virtually anyone who can read and write can use a computer.
...
That is essentially all the "information" given. If you know what this refers
to, or what the nature of the "breakthrough" is please let me know.
-- kirk
------------------------------
Date: Tue 4 Jun 85 13:16:32-PDT
From: Michael Walker <WALKER@SUMEX-AIM.ARPA>
Subject: expert systems vs "conventional" programming
Curt,
You sent a message to the ailist some time ago on expert
systems vs. "conventional" programming. I drafted a reply, but have
been slow in sending it to you. Here it is, perhaps better late than never.
There is a great deal of value in asking questions about what
expert systems are good for. It is important for an emerging science
to seek clear statements of its capabilities and boundaries. I think
your observations on the value of ES are accurate. There is a third
advantage which I feel is moderately well substantiated: that
non-programmers can build certain types of programs using an expert
system shell that they could not otherwise build. A specific example
is the Blue-Box pharmacology therapy advisor built by two
non-programmer medical students using Emycin. A reference is:
Benoit Mulsant and David Servan-Schreiber (1984) "Knowledge
Engineering: A Daily Activity on a Hospital Ward" Computers and
Biomedical Research 17, pp 71-91.
I recommended that they use Emycin when they visited during
their medical clercship at Stanford. Neither of them had the knowledge
to build the system in an ordinary programming language, but were
successful in building their advisor in a remarkably short time using
the expert system shell.
Your question addressed the issue of what expert systems were
good for. We should probably also ask what expert systems research, as
opposed to just expert systems, is good for. I think that expert
systems research is valuable, in part, because it is concerned with
seeking ways to represent and use knowledge that cannot currently be
readily represented or used. In this sense, we can view knowledge
representation research as being concerned with developing new data
structures and algorithms, appropriate for representing and
manipulating symbolic information. There are many other arguments that
could be made about the value of expert systems research, but I think
I'll leave that for further discussion.
Cheers,
Mike Walker
------------------------------
Date: Mon, 3 Jun 85 15:39:31 PDT
From: Phyllis O'Neil <oneil@AEROSPACE.ARPA>
Subject: IJCAI-85: UCLA campus housing available
Economical UCLA dormitory rooms and suites are still available for
IJCAI-85 this August. You can stay in dorm rooms for the 6-night
period beginning Saturday, August 17, or for the 4-night period
beginning Monday, August 19. Dorm rooms sleep one or two people.
Residential suites are available starting Monday, August 19. Suites
sleep two to four people; each suite has two bedrooms, bath, and living
room.
Breakfast and dinner are included when you stay in the dorms or suites.
Additionally, you can use UCLA's swimming pools, raquetball courts, and
other atheletic facilities.
Campus housing forms are on pages 27 and 29 of the IJCAI-85 conference
brochure. For copies of the brochure, contact:
AAAI
445 Burgess Dr.
Menlo Park, CA 94025
(213) 328-3123
------------------------------
End of AIList Digest
********************