Date: Mon 9 May 1988 01:31-PDT From: AIList Moderator Kenneth Laws Reply-To: AIList@KL.SRI.COM Us-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList V6 #100 - New Moderator, Queries, AI News, AI-ED SIG To: AIList@KL.SRI.COM Status: RO AIList Digest Monday, 9 May 1988 Volume 6 : Issue 100 Today's Topics: Administrivia - Matriculation, Queries - Looking back on "The Fifth Generation" & Neural Network Capabilities & Proof Checker & AI and Cytology & Explanation Generation & Knowledge Engineering Courses & Reasoning by Analogy, Review - Spang Robinson V4 N4, Education - New AI-ED SIG ---------------------------------------------------------------------- Date: Mon, 9 May 88 01:30:20 PDT From: Ken Laws Reply-to: AIList-Request@SRI.COM Subject: Matriculation This issue, the 100th of this year, is the last AIList Digest I will edit. After almost exactly five years, I am abandoning my career in volunteer publishing and giving the chair to Nick Papadakis. I'm sure all of you are as glad as I that Nick stepped forward, whether or not it was of his own free will. Future submissions should be mailed to AIList@AI.AI.MIT.EDU, with administrative mail to AIList-Request@AI.AI.MIT.EDU. (I will set up aliases so that mail to the old addresses will be rerouted to MIT.) It may take a while to develop new mailing procedures, so be patient. If you want to help Nick, add meaningful Subject fields to your messages and be sure to notify him if you are going to drop off the net. I will be leaving for NSF in just over a month, arriving about June 30. Sorry I haven't time to visit you all on the drive over. Lily and I have to get to Washington with our three kids -- Brandon, 8; Kelsey, 6; and Devon, 1 -- to look for a house. (If you know of a bargain 4-bedroom rental in Bethesda, Falls Church, or any really great school district there, we'd like to hear about it.) I look forward to meeting many of you during my stay at NSF. Part of my job, as Program Manager for Robotics and Machine Intelligence, will be to send out proposals for peer review. I have started to build a database of addresses and professional interests of potential reviewers (and people able to suggest good reviewers). If you wish to volunteer, just send me a note with keywords for your areas of interest. (Otherwise I'll take names from conference announcements and journal articles.) Good luck to all of you, and especially to Nick! -- Ken ------------------------------ Date: Thu, 5 May 88 08:11:16 PDT From: John B. Nagle Subject: Looking back on "The Fifth Generation" It has been five years since Feigenbaum and McCorduck's "The Fifth Generation", and over five years since the five-year fifth generation program began in Japan. It is thus time for a serious retrospective. Is someone writing one? John Nagle ------------------------------ Date: Thu, 5 May 88 12:11:31 edt From: drb@cscadm.ncsu.edu (Dennis R. Bahler) Subject: Neural Network Capabilities There is a new edition of the Minsky/Papert Perceptron book out which contains new material, including I am told a theorem/proof on the (in)adequacies of the newer neural net models. Unfortunately, the book is back-ordered and I can't get my hands on a copy locally. Would someone (Prof. Minsky himself, perhaps?) care to offer a precis of the new chapter while I wait for my copy to arrive? Dennis Bahler Dept. of Computer Science INTERNET - drb@cscadm.ncsu.edu North Carolina State University CSNET - drb%cscadm.ncsu.edu@relay.cs.net Raleigh, NC 27695-8206 UUCP - ...!decvax!mcnc!ncsu!cscadm!drb ------------------------------ Date: 5 May 88 22:39:58 GMT From: abbott@aerospace.aero.org (Russell J. Abbott) Reply-to: abbott@aero.uucp (Russell J. Abbott) Subject: Proof Checker Does anyone have or know of a public domain (or cheap) proof checker that can be used by undergraduates to write and check simple proofs. I' teaching an automata theory and formal languages course and the students are having a hard time formalizing their thinking. It would be nice if they could practice with an automated proof checker. A simple example problem is: prove that all strings in the set denoted by the regular expression (01 + 10)* have the same number of 0's as 1. The proof is straightforward by induction on the length of the string. The proof checker should have built into it knowledge of set notation, i.e., {X | p(X)} and of inductive proofs. It should also have a basic knowledge of simple arithmetic. Of course it also needs to be able to use results that are proved earlier or given to it as axioms. Thanks, -- Russ Abbott ------------------------------ Date: 4 May 88 17:56:46 GMT From: paul.rutgers.edu!eagles.rutgers.edu!lubinsky@rutgers.edu (David Lubinsky) Subject: Request for refernces on AI and Cytology I would very much appreciate any references on the automatic analysis of cytological samples, especially cervical smears. I would also like to know about current work in robot microscopes. If anyone else is interested, send me your address and I will mail a summary. Thanks David ------------------------------ Date: 7 May 88 00:18:46 GMT From: munnari!moncsbruce.oz.au!pps@uunet.UU.NET (Peter Sember) Subject: explanation generation bib I would greatly appreciate any information on the availability of a bibliography on Explanation Generation Systems. Thanks in Advance. Peter Sember, | ARPA: munnari!moncsbruce.oz!pps@uunet.uu.net Computer Science Dept., | CSNET and ACSNET: pps@moncsbruce.oz Monash University, | Victoria 3168, | Australia | ------------------------------ Date: 6 May 88 15:32:38 GMT From: mcvax!philmds!philtis!grant@uunet.uu.net (Joe Grant) Subject: Knowledge Engineering Courses I am posting this in the hope that people out there can help me. I have just begun working in a Knowledge Based Systems group, and am hoping to train as a knowledge engineer. My background is as a software engineer, and the only exposure that I have had to knowledge engineering to date is a brief introduction to the are of expert systems in college, plus this and the other Comp.ai.* newsgroups which I have been reading for the last 6 weeks. What I would like to do is take some time to find out what courses are available, and plan out a training schedule that will at least give me a good grounding in the current knowledge engineering techniques. These may take the form of College, Company, or correspondence training, which ever suits the situation best, though the College option may very well not prove viable -- I can't see my employers letting me wonder off to anything but a 9--5 course, and all the college courses I took were far from 9--5 on one topic. I have heard that there are a number of in-house courses run by the likes of DEC, to train their own prospective knowledge engineers, any information on such courses would be greatly appreciated. I don't know how well such firms take to applications form non-employees, but if I can get the information at least I can test the water. As I am pretty green in this area I feel that the best way to approach this is to look at all the options. So if you know of any training course or aid, I would be indebted to you if you could let me know about it, and if you have any information as to its usefulness that too would be greatly appreciated. I am not so familiar with the network, so I don't know what sort of addressing format will be required to reach me. However I do know that the route from here to mcvax -- the European gateway I believe -- is via a system called philmds, if I were to send to there I'd enter philmds!mcvax!userid Hopefully that will be enough information for you to reach me. I hope to hear from some of you out there. Many thanks in advance. Joe Grant. ------------------------------ Date: 6 May 88 16:54:53 GMT From: mcvax!ukc!mupsy!liv-cs!stian@uunet.uu.net Subject: Reasoning by Analogy Does anyone know of any work done on reasoning by analogy. Any references received gratefully. Thanks in advance, Ian Finch --------- Postal: Dept. of Computer Science, Chadwick Tower, University of Liverpool, P.O. Box 147, Liverpool L69 3BX Janet: stian@uk.ac.liv.csvax uucp: ...!mcvax!ukc!mupsy!liv-cs!stian arpa: stian%csvax.liv.ac.uk@nss.cs.ucl.ac.uk ------------------------------ Date: Sat, 7 May 88 20:41:21 CDT From: smu!leff@uunet.UU.NET (Laurence Leff) Subject: Review - Spang Robinson V4 N4 Spang Robinson Report on Artificial Intelligence Volume 4, No. 4, April 1988 Lead article is on Expert system applications and MIS. There is a discussion on how MIS people can avoid the high costs of "knowledge engineers," CASE tools, the use of low end systems and availability of tools for mainframes. * Expert Systems never eliminated employees * Cultural issues * Problems in using COBOL to do the job Blue Cross, claim analysis system reduce the time to evaluate a claim from two weeks to fifteen-30 minutes. 145 out of 155 claims were handled by a medical review system. Also provided is a table of some applications with information such as whether the systems paid for themselves, barriers overcame and software: Blue Cross - claims analysis and medical review Data General - MIS department DEC - order handling McDermott Corporation - used Bachman Re-engineering tool set Northern Telecom - Engineering Change Manager Provident Life - Credit Union loan analysis Shearson Lehman - Bachman Re-engineering tools Oil Company - Two Engineering Application ________________________________________ IEEE Conference on Artificial Intelligence Applications Review. There were 620 attendees consisting of people developing commerical AI applications. Complaint was lack of technical content. Session on expert system project failures showed "cultural planning" and "focus on real business planning rather than on technical experimentation" necessary to avoid failure ________________________________________ Carnegie Group of Pittsburgh and Texas Instruments have announced an expert system shell for troubleshooting com plex machine failures and process planners. The development enginee runs on Explorer with capability to translate applications to C or IBM PC/AT. ________________________________________ Shorts: Integrated Analytics sells Market Mind which alerts traders to significant trading patterns. It is written in C. AI Corp has announced "KBMS" available by June 1. It supports IBM mainframes including MVS, VM, CICS, TSO, IMS, CMS, DB/2 and SQL/DS. Cooperative Technology is a new start up that offers management consulting to custom software development. It is headed by William Turpin who developed the Personal Consultant Series from TI. DEC announced that in its network and communications products division, AI integration has generated a 300 percent increase in revenues, a 20 percent rise in gross profit margins and a 285 percent increase in product reliability. Lucid will be bundled with KEE on 80386 and Wisdom Systems' Concept Modeller. Nihon will be giving Symbolics a million dollars in funding for the development of a system based on the Ivory chip. Anza is now selling a diskette-based bibliographic reference list for neural networks. John McDermott has left Carnegie Mellon University to go to DEC. ------------------------------ Date: 29 Apr 88 2144-PDT From: Moderator Steve Barnhouse... Subject: New AI-ED SIG [Forwarded from the IRList Digest.] AI-ED Digest Saturday, 30 Apr 1988 Volume 3 : Issue 16 Date: 15 Apr 88 13:05:00 EST From: "ARTIC::PSOTKA" Subject: New AERA AI & ED SIG ARTIFICIAL INTELLIGENCE & EDUCATION A NEW Special Interest Group was formed at the annual meeting of AERA in New Orleans in April. I am pleased to announce that Wallace Feurzeig of BBN Laboratories is the first Chair of the SIG on AI&ED. Cathie Norris of the University of North Texas is the Newsletter Editor, and Joe Psotka of the U.S. Army Research Institute is the Secretary/Treasurer. It was the consensus of the SIG organizational meeting that it is time to put AI to use in schools by getting a broader range of researchers to be aware of its potential. Artificial Intelligence is advancing rapidly on a broad front of research issues. Many of these issues are directly relevant to the interests and needs of teachers and researchers at all levels of instruction, and in many different settings. This SIG will provide an overview of the current issues that may have the strongest effect on education. Artificial Intelligence is increasingly becoming more applicable to practical use in education. In part, this is because the technology of AI, based as it is on specific algorithms and understanding derived from areas of computer science and cognitive science somewhat remote from the mainstream of educational research, is maturing steadily and becoming less arcane and more generally useful for instruction. The other main reason for this increasing practicality of AI technology is the continuing increase in the power of available personal computer technology at an affordable price for schools and workplaces to purchase. The outstanding example of this is the Hypercard environment on MACs, and the powerful Lisp environments on all new PCs. Both of these factors make it more important that educational researchers understand and become familiar with AI technology. This SIG will offer us an opportunity to introduce other educational researchers to these topics. The SIG on AI & ED will be mainly concerned with the use of AI and cognitive science technologies for education. Primary areas for reporting research in these technologies will be within Authoring systems for CBI and ICAI; intelligent microworlds; machine learning; complex environments for instruction; knowledge representation; qualitative modelling techniques; structures of declarative knowledge; computer thinking tools; rule systems for procedural knowledge; student modelling; student diagnosis; teacher amplifiers; hypertext systems; natural language processing; and other important outgrowths of AI that offer significant potential for improving education in the schools, workplace, and at home. For more information, inquiries, suggestions for symposia, and other offers of support, please contact : Wallace Feurzeig, BBN Laboratories 10 Moulton St Cambridge, MA 02238 at (617)873-3448 or Feurzeig@g.bbn.com.arpa Send newsletter contributions, announcements, and other information to: Cathie Norris University of North Texas P. O. Box 5155 Denton, TX 76203 at (817)565-4189 Joseph Psotka, Ph.D. Army Research Institute 5001 Eisenhower Avenue Alexandria, Virginia 22333-5600 OR CALL: (202)274-5540 or Psotka@ARI-HQ1.Arpa If you would like to join, please send this application. [Contact the message author. -- KIL] ------------------------------ End of AIList Digest ********************