perm filename QUAL.MSG[ADM,DBL] blob
sn#163797 filedate 1975-06-13 generic text, type T, neo UTF8
∂12-JUN-75 1640 THE,AJT
just to say that I liked the format of the AI qual a lot, and
think that it should be continued next year...
∂15-APR-75 1711 ESS,JMC
The syllabus is tolerable to me. I suggest you include my review of the
Lighthill report that appeared in the AI Journal, but this is not a strong
suggestion.
∂11-APR-75 1207 network site AI
Date: 11 APR 1975 1506-EDT
From: SUZIN at MIT-AI
To: dbl at SU-AI
hi. about your questions.....
1. yes, you have to pay the registration fee.
2. xgp'ed copy is ok as long as it is not covered with those little specks.
might be a good idea to hav a clean drum and/or webbing. just make
sure that it doesn't have that grayish background xgp sometimes has.
3. yes, its ok to paste up on the special sheets WITHIN THE COLUMNS.
4. about travel funds: there are none available from ijcai committee.
Prof. Winston does not have any suggestions for other sources. sorry.
cheers,
suzin l. jabari
∂02-APR-75 1057 LSP,AJT
I've put an AI QUAL questionaire in yr. box. I will be glad to discuss my
comments (which are necessarily cryptic) with you and/or Terry. What is your
role in this? arthur
∂21-MAR-75 1151 1,MLM
Am interested in AI qual; so what?
∂21-MAR-75 1151 1,MLM
Am interested in AI qual; so what?
∂15-MAR-75 1055 network site SRI
Date: 15 MAR 1975 1054-PDT
From: SHORTLIFFE at SRI-AI
Subject: AI QUAL - MYCIN REFERENCE
To: TW at SU-AI
cc: DBL at SU-AI
TERRY,
I was glancing through the material on the proposed content of the
AI qual in your file SYL.PUB[AI,TW] (saw Doug Lenat's note on the subject
when I signed on at AI recently) and noticed that you want to know the most
up-to-date short MYCIN reference. I think the best paper for your purposes
would be the one which will appear in "Computers And Biomedical Research",
Volume 8, in June of this year. A copy of the draft of this paper will
appear in Cordell's syllabus for CS206 this spring, so copies should be
readily available even though the paper will not yet have appeared in press.
It is authored by myself, Randy Davis, Stan Axline, Bruce Buchanan, Cordell,
and Stan Cohen. Title is "Computer-Based Consultations in Clinical Thera-
peutics: Explanation and Rule Acquisition Capabilities of the MYCIN
System".
The whole syllabus looks like a good summary of what is happening
in the field, so I'll look forward to listing the whole thing once SUMEX
gets on the net and I can FTP a copy down here. Seldom have a chance to
stop by the AI lab at present.
Regards,
Ted
-------
∂14-MAR-75 0132 SEC,TW
*** Doug, do you think we should try ordering copies from NTIS or
Xerox the ones we have, or what? Do whatever seems best and
if it involves a budget, go ahead and pay and I'll be sure
you get reimbursed. Thanks --t *****
∂13-MAR-75 0625 network site AI
Date: 13 MAR 1975 0925-EDT
From: SUZIN at MIT-AI
To: tw at SU-AI
hi there..got your messages.
bad news about the 1974 Progress Report. We are all out of it and it probably
won't be reprinted until May at which point we will have 1000 or so printed. It's
going to be revised and we can't afford to print up just a few copies. It is
available from National Technical Information Service, but they have not notified
me as to what the AD number is, though I'm sure you can still order it. Their
address is:
NTIS, Operations Division, Springfield, Virginia
Unfortunately I can't tell you how much they would charge. It cost us a fortune to
print so we really have to charge for it, especially for quantity.In the meantime,
I'll see if I can get an AD numer out of NTIS. It's a bitch to call them up
since one gets the royal gov't runaround!
love, suzin
∂12-MAR-75 1833 1,TW
no, don't merge in the whole list -- it's too long and scattered. I was
just suggesting it as a source of already machine-read-in detailed references
for those things we decided we wanted.
∂12-MAR-75 1827 1,TW
I just got an extesive reference list from Danny Bobrow. It is in
REFS.PUB[AI,TW] and should give the details for at least some
of the references we want. Just for consistency, why don't
we use the same format, and then the two lists can be
merged and cleaned up (the one he gave me is obviously a
merge, since there is lots of redundancy in it).
∂11-MAR-75 1923 1,TW
I did some more work on the syllabus. I made major changes to the
mechanics at the end, which I would like to discuss. I wanted to list
the references separately for two reasons. First, I think it is easier
to grasp a list if it appears on fewer pages. Including all the references
internally would expand it in terms of the amount you need to scan for
any one outline entry. Second, some items are referred to over and over,
and its annoying on a repeated reference to have to thumb back to some
other section to find the details. It's easier if they're all in a consistent
place. The Key words I have used are more or less mnemonic for me, but
I would be glad to change any individual ones. I do find that key word
mnemonics are better than plain reference numbers (worst) or dates (better but
still confusing). The reference list is then alphabetical by author and
key word and gives all the relevant details. I think the reference list
would be the right place to include annotations (like which chapters and
pages of a long document are most relevant and readable, or what earilier
work this references replaces). I believe the list is more or less complete,
and we need to think about some sort of priority marking since it is too
much to expect everyone to read. I would like to send it around to Cordell,
Dave Luckham, Tom Binford, etc. in the next couple of days to get their
comments on the sections they know best, and then leave you and David Levy
(if you have time) to clean it up, fill in details of publication and PUB
out a decent copy while I am gone (I am leaving Fri and won't be back until
April 2). I'll be around most of the day tomorrow (Wed) so lets try to
get together. --terry
∂05-MAR-75 1739 1,TW
SYL.PUB[AI,TW] is a start towards the qual syllabus. At the moment
I want to expand it to include everything we might possibly want,
including both topics and specific references. Later we can prune
or better, indicate priorities. Go ahead and dump suggestions and
comments right into the file. I would like to get a reasonable
version by the end of next week if at all possible.
--Terry
COMMENT ⊗ VALID 00003 PAGES
C REC PAGE DESCRIPTION
C00001 00001
C00002 00002 ∂25-FEB-75 1400 BPM,BPM
C00003 00003 ∂25-FEB-75 1307 BPM,BPM
C00006 ENDMK
C⊗;
∂25-FEB-75 1400 BPM,BPM
As of today there are nine people signed up for the AI Qual.
∂25-FEB-75 1307 BPM,BPM
It looks like Terry Winograd will be the AI Qual chairman. He is leaning toward
giving individual two-hour orals to everyone. Each oral would be given by
three people: one of the professors (Winograd, Green, Feigenbaum, and perhaps
Baskett), a research associate (e.g., Binford, Buchanan, Luckham, etc.), and
an outsider (e.g., Bobrow, Raphael, Nilsson, etc.). Winograd isn't too hot
on students being on orals committees. You may feel otherwise. In any case,
I told him that you were the student appointed to the AI Qual Committee, so you
should go talk to him about it. There is no guarantee the thing will be an
oral at this point; one of the things you could help on is check with the
Department office at Polya and find out who has signed up for the Qual and see
how they feel about a two-hour oral versus, say, an eight-hour written exam.
Even if Winograd decides on an oral (actually it should be the AI faculty
plus you that decides), he may want help organizing the reading list, oral
committees, exam times, etc. Good luck and let me know what's happening.