From comsat@vpics1 Mon Oct 21 22:34:41 1985 Date: Mon, 21 Oct 85 22:34:37 edt From: comsat@vpics1.VPI To: fox@opus (MILLER,FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a019013; 21 Oct 85 0:35 EDT Date: Sun 20 Oct 1985 20:40-PDT Reply-to: AIList%sri-ai.arpa@CSNET-RELAY US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #150 To: AIList%sri-ai.arpa@CSNET-RELAY Received: from rand-relay by vpi; Mon, 21 Oct 85 22:23 EST AIList Digest Sunday, 20 Oct 1985 Volume 3 : Issue 150 Today's Topics: Seminars - Program Logics (UPenn) & Meaning, Information and Possibility (UCB) & Machine Learning and Knowledge Representation (NU) & Intelligent Mail Manipulation (MIT) & A Logic for Defeasible Rules (Buffalo) & Learning From Multiple Analogies (GTE) & Computational Discourse Analysis Using DEREDEC (MIT) & RESEARCHER and Patent Analogies (CMU) ---------------------------------------------------------------------- Date: Mon, 14 Oct 85 19:26 EDT From: Tim Finin Subject: Seminar - Program Logics (UPenn) REASONING ABOUT PROGRAMS: CONCEPTUAL AND METHOLOGICAL DISTINCTIONS DANIEL LEIVANT, COMPUTER SCIENCE, CMU 3:00 pm Tuesday, October 15, 1985 216 Moore, University of Pennsylvania Reasoning about programs can be done explicitly, in first-order or higher-order mathematical theories, or implicitly, in modal logics of programs (Hoare Logic, Dynamic Logic...). One wants the latter, but the former are better suited for metamathematical analysis (semantics, calibration of proof-theoretic strength). However, modal logics are interpretable in explicit theories, so we can eat the cake and have it. In particular, we can distinguish in modal logics of programs a purely logical component and an analytical component. For example, Hoare's Logic captures exactly logical reasoning about partial-correctness assertions over WHILE-programs. We argue that this type of completeness is more informative than relative completeness. ------------------------------ Date: Wed, 16 Oct 85 14:22:50 PDT From: admin@ucbcogsci.Berkeley.EDU (Cognitive Science Program) Subject: Seminar - Meaning, Information and Possibility (UCB) BERKELEY COGNITIVE SCIENCE PROGRAM Cognitive Science Seminar - IDS 237A Tuesday, October 22, 11:00 - 12:30 240 Bechtel Engineering Center Discussion: 12:30 - 1:30 in 200 Building T-4 ``Meaning, Information and Possibility'' L. A. Zadeh Computer Science Division, U.C. Berkeley Our approach to the connection between meaning and information is in the spirit of the Carnap--Bar-Hillel theory of state descriptions. However, our point of departure is the assump- tion that any proposition, p, may be expressed as a generalized assignment statement of the form X isr C, where X is a variable which is usually implicit in p, C is an elastic constraint on the values which X can take in a universe of discourse U, and the suffix r in the copula isr is a variable whose values define the role of C in relation to X. The principal roles are those in which r is d, in which case C is a disjunctive con- straint; and r is c, p and g, in which cases C is conjunctive, probabilistic and granular, respectively. In the case of a disjunctive constraint, isd is written for short as is, and C plays the role of a graded possibility distribution which asso- ciates with each point (or, equivalently, state-description) the degree to which it can be assigned as a value to X. This possibility distribution, then, is interpreted as the informa- tion conveyed by p. Based on this interpretation, we can con- struct a set of rules of inference which allow the possibility distribution of a conclusion to be deduced from the possibility distributions of the premises. In general, the process of inference reduces to the solution of a nonlinear program. The connection between the solution of a nonlinear program and the traditional methods of deduction in first-order logic are explained and illustrated by examples. ELSEWHERE ON CAMPUS William Clancy of Stanford University will speak on ``Heuristic Classification'' at the SESAME Colloquium on Monday, Oct. 21, 4:00pm, 2515 Tolman Hall. Ruth Maki of North Dakota State University will speak on ``Meta- comprehension: Knowing that you understand'' at the Cognitive Psychology Colloquium, Friday, October 25, 4:00pm, Beach Room, 3105 Tolman Hall. ------------------------------ Date: Thu, 17 Oct 85 15:02 EDT From: Carole D Hafner Subject: Seminar - Machine Learning and Knowledge Representation (NU) Northeastern University College of Computer Science Colloquium 4p.m. Wednesday, October 30 Brittleness, Tunnel Vision, Machine Learning and Knowledge Representation Prof. Steve Gallant Northeastern University A system is brittle if it fails when presented with slight deviations from expected input. This is a major problem with knowledge representation schemes and particularly with expert systems which use them. This talk defines the notion of Tunnel Vision and shows it to be a major cause of brittleness. As a consequence it will be claimed that commonly used schemes for machine learning and knowledge representation are pre- disposed toward brittle behavior. These include decision trees, frames, and disjunctive normal form expressions. Some systems which are free from tunnel vision will be described. Place: 405 Robinson Hall Northeastern University 360 Huntington Ave. Boston MA ------------------------------ Date: Sun, 13 Oct 1985 16:53 EDT From: Peter de Jong Reply-to: Cog-Sci-Request%MIT-OZ Subject: Seminar - Intelligent Mail Manipulation (MIT) [Forwarded from the MIT bboard by SASW@MIT-MC.] Thursday 17, October 4:00pm Room: NE43- 8th floor Playroom The Artificial Intelligence Lab Revolving Seminar Series "The Information Lens: An Intelligent System for Finding, Filtering, and Sorting Electronic Messages" Thomas W. Malone MIT Sloan School of Management This talk will describe an intelligent system to help people share, filter, and sort information communicated by computer-based messaging systems. The system exploits concepts from artificial intelligence such as frames, production rules, and inheritance networks, but it avoids the unsolved problems of natural language understanding by providing users with a rich set of semi-structured message templates. A consistent set of "direct manipulation" editors simplifies the use of the system by individuals, and an incremental enhancement path simplifies the adoption of the system by groups. The talk will also include an overview of the other projects and research goals in the Organizational Systems Laboratory at MIT. ------------------------------ Date: Fri, 18 Oct 85 08:30:03 EDT From: "William J. Rapaport" Subject: Seminar - A Logic for Defeasible Rules (Buffalo) UNIVERSITY AT BUFFALO STATE UNIVERSITY OF NEW YORK DEPARTMENT OF COMPUTER SCIENCE COLLOQUIUM DONALD NUTE Advanced Computational Methods Center and Department of Philosophy University of Georgia A LOGIC FOR DEFEASIBLE RULES Humans reason using defeasible and sometimes conflicting rules like `Matches burn when struck' and `Wet things don't burn'. A formal language for representing sentential versions of such rules is presented together with a derivability relation for this language. The resulting system, LDR, is non-monotonic. Inspired by work in conditional logic, the non-monotonic rules of LDR correspond to simple subjunctive and `might' conditionals. Chaining of these rules is restricted in LDR just as the transi- tivity of the conditional is restricted in conditional logics. Several notions of consistency and coherency are defined. LDR is of special importance for research in automated reasoning, since its language is PROLOG-like and its derivability relation can be implemented in PROLOG. Thursday, November 7, 1985 3:30 P.M. Bell 337, Amherst Campus Wine and cheese will be served at 4:30 P.M., 224 Bell Hall For further information, contact: William J. Rapaport Assistant Professor Dept. of Computer Science, SUNY Buffalo, Buffalo, NY 14260 (716) 636-3193, 3181 uucp: ...{allegra,decvax,watmath}!sunybcs!rapaport ...{cmc12,hao,harpo}!seismo!rochester!rocksvax!sunybcs!rapaport cs/arpanet: rapaport%buffalo@csnet-relay ------------------------------ Date: Fri, 18 Oct 85 23:44:39 EDT From: Bernard Silver Subject: Seminar - Learning From Multiple Analogies (GTE) GTE LABS INCORPORATED MACHINE LEARNING SEMINAR Title: Learning from Multiple Analogies Speaker: Mark H. Burstein BBN Labs. Date: Monday October 21, 10am Place: GTE Labs 40 Sylvan Rd, Waltham MA 02254 Students learning about an unfamiliar new subject under the guidance of a teacher or textbook, are often taught basic concepts by analogies to things that they are more familiar with. Although this seems to be a very powerful form of instruction, the process by which students make use of this kind of instruction has been little studied by AI learning theorists. A cognitive process model of how students make use of such analogies will be presented. The model was motivated by examples of the behavior of several students who were tutored on the programming language BASIC, and focusses in detail on the development of knowledge about the concept of a program variable, and its use in assignment statements. It suggests how several analogies can be used together to form new concepts where no one analogy would have been sufficient. Errors produced by one reasoning from one analogy can be corrected by another. As an illustration of the main principles of the model, a computer program, CARL, is presented that learns to use variables in BASIC assignment statements. While learning about variables, CARL generates many of the same erroneous hypotheses seen in the recorded protocols of students learning the same material given the same set of analogies. The learning process results in a single target model that retains some aspects of each of the analogies presented. For more information, contact Bernard Silver (617) 576-6212 ------------------------------ Date: 11am 10/22/85 From: Alker@mc Subject: Seminar - Computational Discourse Analysis Using DEREDEC (MIT) [Forwarded from the MIT bboard by SASW@MIT-MC.] Computational Discourse Analysis Using DEREDEC: An Analysis of Balzac's Sarrasine Jaqueline Leon and Jean-Marie Marandin Centre National de la Recherche Scientifique Paris, France We present research in computational discourse analysis and discuss an example for the case of Balzac's Sarrasine. We use P. Plante's DEREDEC programming system in this work because of its suitability for natural language processing. After a bottom-up syntactic parser for French grammar produces a syntactic derivation, we perform pattern matching on the output to achieve a linguistic and literary interpretation. We describe how we use these programs to capture two different aspects of a text: the thematic segmentation and density. Time: 11-12:30, Tuesday, October 22, 1985 Place: Millikan Room, E53-482 Host: Professor Hayward R. Alker, Jr., Department of Political Science, MIT ------------------------------ Date: 18 Oct 85 10:14:51 EDT From: Jeanne.Bennardo@CMU-RI-ISL1 Subject: Seminar - RESEARCHER and Patent Analogies (CMU) Topic: Presentation of RESEARCHER project. Speaker: John C. Akbari Place: DH3313 Date: Wednesday, Oct. 23 Time: 10:00am - 11:00am Speaker: John C. Akbari is a Masters student at Columbia University's Department of Computer Science. He is interested in joining the Intelligent Systems Laboratory's Phoenix project. Below is a description of his artificial intelligence research. Both projects described below investigate different aspects of RESEARCHER, a prototype intelligent information system being developed at Columbia University under the direction of Professor Michael Lebowitz. The domain of investigation is disc drive patents. The result of this research is being implemented in LISP as a component of RESEARCHER. MS Thesis Research involves generating "catalogue descriptions" of hierarchical objects, determining salience as a function of similarity between an instance of an object and the prototype of the object. This will be used in generating information to be passed to a case grammar generator to produce the actual text. We hope to develop a method of determining importance of static information (via "filtering through" the prototype) relative to context. We are studying the interaction of structural, attributive, and functional information on the quality of the description. Further work will investigate the need for different prototypes for different users as an aspect of user modelling, so that a patent lawyer would receive a different description from an engineer, given the same instance. Thesis advisor: Prof. Michael Lebowitz Natural language We are enhancing RESEARCHER's parser to utilize syntactic aspects of relations that cause focus of attention to shift within sentences. This involves modifying memory-based parsing to determine when syntax cues are sufficiently strong to over-ride the need to search memory. Supervisor: Prof. Michael Lebowitz ------------------------------ End of AIList Digest ******************** From comsat@vpics1 Mon Oct 21 05:02:55 1985 Date: Mon, 21 Oct 85 05:02:52 edt From: comsat@vpics1.VPI To: fox@opus (MILLER,FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a019489; 21 Oct 85 2:10 EDT Date: Sun 20 Oct 1985 20:46-PDT Reply-to: AIList%sri-ai.arpa@CSNET-RELAY US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #151 To: AIList%sri-ai.arpa@CSNET-RELAY Received: from rand-relay by vpi; Mon, 21 Oct 85 04:53 EST AIList Digest Sunday, 20 Oct 1985 Volume 3 : Issue 151 Today's Topics: Seminar Summary - Situation Theory and Situation Semantics, Conferences - Symposium in Logic on Computer Science & The Computerized Oxford English Dictionary ---------------------------------------------------------------------- Date: Wed 16 Oct 85 17:12:46-PDT From: Emma Pease Subject: Seminar Summary - Situation Theory and Situation Semantics [Excerpted from the CSLI Newsletter by Laws@SRI-AI.] CSLI SEMINAR SUMMARY Notes from the STASS Underground October 3, 1985 David Israel gave an overview of the motivation behind the formation of the Group on Situation Theory and Situation Semantics (STASS). The aim of the group is the development of Situation Theory as a framework within which to express, analyse, and compare treatments of a wide range of problems and phenomena. Among the ``applications areas'' are the semantics of natural languages, the semantics of programming and other computer languages, the nature of informational content, the nature of computational processes, problems in the theory of representation, problems about the nature of truth, etc. The method of development is essentially a close and continuous interaction between those working on Situation Theory itself and those looking to use the theory within their own areas of interest. This interaction is enhanced because everybody in the group is doing both things, often simultaneously---though not, of course, equally. In the respect of being a background theory within which to develop theories of more delimited domains, Situation Theory is analogous to Set Theory. Thus, for instance, Montague's treatment of phenomena in the semantics of natural language was carried out within set theory. So, too, was the treatment by Barwise and Perry in ``Situations and Attitudes.'' The crucial transition between the account in that book and the present approach is precisely the abandonment of the strategy (or was it anyway only a tactic?) of modelling all but a small number of basic kinds of things in set theory. Thus, for instance, in ``Situations and Attitudes'' there was no real attempt to explicate the nature of propositions---though much of the interest of the book was said to lie in its treatment of the propositional attitudes. The reason for this uncomfortable state of affairs was that there was no good way of modelling propositions set theoretically. The aim now is a direct, non-reductionist treatment of the various kinds of entities only modelled in the book---thus, of states of affairs and facts, conditions, situations, propositions, etc. This is thought to have a number of happy side effects. One is that it makes it much easier to expose the various modes of modelling to analysis---easier simply because one has not committed oneself to modelling as one's major theoretical technique. The second stems from the fact that Situation Theory is not only analogous to Set Theory in a certain respect; Situation Theory is intended both to encompass and to be modellable by Set Theory. Thus, the demand that Set Theory be capable of providing models of Situation Theory imposes constraints on our conception of sets. A crucial example of such a constraint is that there be non wellfounded sets. --David Israel ------------------------------ Date: Fri, 11 Oct 1985 20:18 EDT From: MEYER@MIT-XX.ARPA Subject: Symposium in Logic on Computer Science ANNOUNCEMENT AND CALL FOR PAPERS SYMPOSIUM ON LOGIC IN COMPUTER SCIENCE JUNE 16-18, 1986, CAMBRIDGE, MASS., USA The Conference will cover a wide range of theoretical and practical issues in Computer Science broadly relating to Logic, including algebraic and topological approaches. Many of these areas have been represented separately, but not in a general LICS conference. Some suggested, not exclusive, topics are: Abstract data types, computer theorem proving, concurrency, constructive proofs as programs, data base theory, foundations of logic programming, logic-based programming languages, logic in complexity theory, logics of programs, knowledge and belief, program verification, semantics of programs, software specifications, type theory. Organizing Committee J. Barwise E. Engeler A. Meyer W. Bledsoe J. Goguen R. Parikh A.Chandra,Chair D. Kozen G. Plotkin E. Dijkstra Z. Manna D. Scott Program Committee R. Boyer W. Damm S. German D. Gries M. Hennessy G. Huet D. Kozen A. Meyer,Chair J. Mitchell R. Parikh J. Reynolds J. Robinson D. Scott M. Vardi R. Waldinger Paper Submission: Authors should send 16 copies of a detailed abstract by Dec. 23, 1985 to the program committee chairman: Albert R. Meyer - LICS Program tele:(617)253 6024 MIT Lab. for Computer Science Arpanet: MEYER@XX 545 Technology Square, NE43-315 Cambridge, MA 02139 USA (If reproduction facilities are not available to the author, a single copy of the abstract will be accepted.) The abstract should be at most 4500 words, but should provide sufficient detail, including references and comparisons to related work, to allow the Program Committee to assess its technical merits. The time between abstract due-date and committee review is short, so late submissions run a high risk of elimination. Authors will be notified of acceptance by Jan. 24, 1986. Photo-ready copies of accepted papers typed on special forms are due March 31, 1986. General Chairman: A. K. Chandra, IBM Thomas J. Watson Research Center, P.O. Box 218, Yorktown Heights, NY 10598, tele: (914) 945-1752, CSNET: ASHOK.YKTVMV at IBM. Local Arrangements Chairman: A. J. Kfoury, Dept. of Computer Science, Boston University, Boston, MA 02215, tele: (617) 353-8911, CSNET: KFOURY at BOSTONU. Sponsorship: IEEE Computer Society, Technical Committee on Mathematical Foundations of Computing, in cooperation with ACM SIGACT and Association for Symbolic Logic (request pending). ------------------------------ Date: Tue, 15 Oct 85 17:27:37 edt From: lesk%petrus@mouton.ARPA (Michael E. Lesk) Subject: Conference on the computerized Oxford English Dictionary The University of Waterloo Centre for the New OED is starting research projects using the machine-readable form of the OED now being prepared. The plan is to have not just typesetting tapes, but an electronic database representing the history and use of the English language, as shown in the dictionary. A one-day meeting at Waterloo, from 7pm Thursday Nov. 7 through 4:30pm Friday Nov. 8, 1985, will examine research areas related to the OED and machine-readable dictionaries. The program is: Introduction John Simpson, Oxford University Press, "The New OED Project" John Stubbs, University of Waterloo, "The UW Centre for the New OED" Using On-Line Dictionaries (Michael Lesk, session chair) Henry Kucera, Brown University, "The Problem of Structural Ambiguity in the Lexicon" Donald Walker, Bell Communications Research, "Knowledge Resource Tools for Accessing Large Text Files" George Miller, Princeton University, "Wordnet: A Dictionary Browser" The Use and Misuse of Dictionaries (Neil Hultin, session chair) Gisele Losier, U. Waterloo, "Using the OED for the Study of Loan Words" Christopher Dean, U. Saskatchewan, "The OED: The Study of Local Regional Dialects and Historical Dialet Dictionaries" Knowledge Databases (Robin Cohen, session chair) Randy Goebel, U. Waterloo, "What is a Knowledge Representation System?" John Sowa, IBM, "Using Knowledge Representation to Capture the Semantic Information of a Lexicon" Summary (Frank Tompa, U. Waterloo, plus other session chairs) Those interested in attending should send $25 US or $35 Canadian, along with their name, address and phone numbers, to: Centre for the New OED Dana Porter Library, rm 105 University of Waterloo Waterloo, Ontario, Canada N2L 3G1 ------------------------------ End of AIList Digest ******************** From csvpi@vpics1 Mon Oct 21 22:45:16 1985 Date: Mon, 21 Oct 85 22:45:07 edt From: csvpi@vpics1.VPI To: fox@opus (MILLER,FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a019805; 21 Oct 85 3:29 EDT Date: Sun 20 Oct 1985 20:57-PDT Reply-to: AIList%sri-ai.arpa@CSNET-RELAY US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #152 To: AIList%sri-ai.arpa@CSNET-RELAY Received: from rand-relay by vpi; Mon, 21 Oct 85 22:26 EST AIList Digest Monday, 21 Oct 1985 Volume 3 : Issue 152 Today's Topics: Query - Classic AI Books, Administrivia - Move Discussion to Symbolics List, Games - Go, Humor - IM Hockey!, AI Tools - The future of POP-11 ---------------------------------------------------------------------- Date: Saturday, 19 October 1985 20:05:27 EDT From: Dan.Miller@a.sei.cmu.edu Subject: Classic AI Books This type of question was asked a while ago on the SOFT-ENG (software engineering) bboard, and now I'd like to ask the same of AI'ers: WHAT, IN YOUR OPINION, ARE THE "CLASSIC" AI BOOKS??? Please send your replies or pointers directly to dhm@sei.cmu.edu (old style: dhm@cmu-sei.arpa). I'll post the "netwide" consensus. --- Daniel "Dan" H. MIller Software Engineering Institute dhm@sei.cmu.edu (dhm@cmu-sei.arpa) Carnegie-Mellon University (412)578-7700 Pittsburgh, PA 15213 USA "The views expressed are my own, and not of my employer" ------------------------------ Date: 19 Oct 85 15:17 PDT From: Fischer.pa@Xerox.ARPA Subject: Move discussion to Symbolics list In your message to AIList with header: Date: Wed, 16 Oct 85 10:21 CDT From: Joseph_Tatem Subject: AI machines You prepare to discuss the redesign of the Symbolics window system. That should probably be done on the mailing list reserved for that specific machine. I doubt everyone on AIList has a Symbolics system (I don't), and those folks with Xerox Lisp machines are pretty happy with their existing window interface. (ron) Ron Fischer Xerox AI Systems Palo Alto, CA ------------------------------ Date: Sat, 19 Oct 85 13:38:01 EDT From: "Keith F. Lynch" Subject: the ancient oriental game of Go Reply to David Nicol : Robert Maas, REM%IMSSS@SU-SCORE.ARPA, has a Go playing program. ...Keith ------------------------------ Date: 16 Oct 1985 22:14 EDT (Wed) From: "David A. Brown" Subject: IM Hockey! [Forwarded from the MIT bboard by SASW@MIT-MC.] INTRAMURAL HOCKEY or A Project in Cooperative Real-Time Solution of 12-body Problems Consider a compact subset of the Euclidean plane, H; cover this compact set with substance W, cool that substance past its freezing point. Let P and Q be sets of independently controlled self-interested actors. Each actor is unaware of the specific knowledge of the other actors (imperfect information) and also unaware of the effects of its own actions (perfect incompetence). Posit a game plan "gp" in which an object, "puck", originally lying on the plane H, is attracted strangely to one or both of two attractors, g1 and g2. It is well known that this problem is decidable. We propose to implement a distributed system for solving this problem. A simple strategy for solving this is known (see Oilers, 1985), but difficult to implement in general (see Rangers, 1900-.....). A simulation of this 12-body problem will be performed under the auspices of MIT intramurals; bodies are needed. [...] ------------------------------ Date: 18 Oct 1985 22:17:47-BST From: Aaron Sloman Subject: The future of POP-11 Fri Oct 18 22:17:37 BST 1985 Ken, I've received a letter about POP-11 which may be of some interest to AIList readers. I am forwarding it for you to veto if appropriate. First some background. POP-11 is a much expanded version of POP-2, the AI language originally developed at Edinburgh University. POP-11 is being increasingly widely used in the UK, both in universities and in commercial organisations taking part in Britain's Government-funded Alvey Programme, which has a major AI component. It has many Lisp-like features, including garbage collection, built in list processing, incremental compilation, integrated screen editor etc. However, the syntax and the philosophy are very different from Lisp: e.g. procedures are just values of variables. People in the UK have been discussing the future of POP-11, some arguing that it should be abandoned in favour of Common Lisp, others arguing that it has too many advantages and should not be thrown out just because it is less widely used. POP-11 is the core language of POPLOG, developed at Sussex University and marketed by Systems Designers. I recently had a letter on POP-11 from Steve Knight who currently works for GEC, one of Britain's largest high tech firms, and is about to move to Hewlett Packard AI Labs in Bristol. He doesn't at present have access to Arpanet, so agreed that I could forward an extract from his letter. As we at Sussex University have an interest in this matter, being the designers of POPLOG, I thought I should leave it to you to decide whether the contents of this message are worth broadcasting. His current address (mail will be forwarded after he moves): Steve Knight, 35 Baker St, Chelmsford, Essex, CM2 OSA, England. Steve writes: As you know, my main interests in computing are to do with designing high quality software in as little time as possible. What do I mean by high quality? Specifically, having high reliability, good space/time performance, being written in a natural and elegant way, and most important of all being what the client really wanted. By as little time as possible, I mean timescales that are orders of magnitude shorter than the ones we accept now. Not unnaturally these are regarded as contradictory aims. The very short timescales are frequently the butt-end of managerial humour (and subsequent managerial eating of hats!) I practice what I preach in POPLOG because it is a system of great potential; this potential is only partly understood, even at Sussex. As an example, when Graham Higgins (now at Hewlett Packard UK) wanted a more complete Lisp system than the one available in POPLOG 9.0, we wrote our ZLISP in 10 man-days, using POP-11. The subsequent code was of the order of +1Mb. The system had no serious bugs (apart from design errors, visible in long distance hindsight), was as complete as any other Lisp system either of us had used (PSL/Franz/other GEC-Lisps - but not Common Lisp which is significantly better of course!), ran quicker than any other Lisp on the system (VAX-750), had some novel features (such as tracing individual embedded S-expressions) and was integrated with VED (the POPLOG screen editor). This is not just "blowing my own trumpet" because it was easy. Any software engineer at all could do it in POP-11. Anyhow, my interests led me in a number of distinct main directions in POP11 * formal methods and specification tools * applicative languages/style * software design assistance * reusability of software * programming environments * natural user interfaces. I have been slowly building up a body of software and experience in POPLOG that allows me to tackle a few of these issues. Along the way I have encountered some basic problems in using POPLOG the way I want. Some are to do with the POPLOG-- Operating system interface. Some are design errors in POPLOG. But before I dissect some of these problems for you, it might be proper to touch on the really good points of POPLOG, for my work. * Help files. I think the most inspirational aspect of POPLOG (for me) is the 'help' system. I dislike manuals intensely. They represent all that is wrong with computing. The commitment to online help gets 10/10! I am amazed at the high standard of description and usefulness throughout. Of course, 'help' is not enough, but POPLOG shows up its potential. * Virtual machine. The presence of the VM in terms of its 'code planting procedures' is invaluable. [What is needed here is a full description of the VM]. I often remark that it is easier to write a compiler than an interpreter in POPLOG! * POP-11 A super language. Rich without being especially arbitrary (non-orthogonal). The most important feature is the open stack, leading to beautiful transformations. [Caroline and I hope to write a paper (for ECAI 86?) on why Lisp is a "bad" language, using POP-11 as a contrast.] The other important features are: procedures as first-class citizens, the powerful idea of updaters, partial application rather than closures, and the 'keys' system. * VED Despite its faults, plainly the best full screen editor of its type. EMACS comes a poor second in my view. [VED needs a major overhaul, of course, but it is undoubtedly the best all the same.] Without the powerful VED macros I define in my vedinit.p file I could not manipulate code with the same flexibility or have the same productivity. Anyhow the list of generalisations could go on. Suffice it to say that POPLOG does provide an environment conducive to the development of software and teaching programming. POP-11 is the key ingredient. [As an aside, Prolog is useful but I regard it as a difficult language to use well. It is temptingly declarative, but the left-to-right evaluation order gives far too little control and spoils the declarative flavour.] _______ {Steve's letter goes on with a list of complaints at a level of detail which would not make much sense to people who are not familiar with POPLOG. So I have summarised them, and he has checked my summary, as follows:} * Subsystems POPLOG does not yet have a good way of integrating different subsystems (e.g. Lisp, Prolog, POP-11, user defined sub-systems) with the editor and compilers, help system, teach system, etc. The facilities available are not yet fully documented. * POP-11 should not be a privileged language as it is now. The POPLOG VM should come with a disposable kit of tools for building a range of language compilers. (This is also the Sussex long term philosophy - see the paper by Mellish and Hardy in IJCAI 1983.) * VED - the screen editor is obsolete in terms of modern machines, and needs a thorough overhaul including abandoning the explicit use of files, adding non-ascii I/O, adding a window manager, etc. (I appreciate this is a debatable area.) * Need to be able to decouple compiled programs from the development environment after development and testing. 'Having written my 250 line POP-11 program, how can I throw everything else away...?' * Autoloading is a 'messy idea' and, like files, needs to be re-designed. * Sections (POP-11's substitute for 'packages'). 'I don't think these are quite right yet' * Files and Operating System. The O.S. should be abandoned and replaced with an interactive programming environment. POPLOG needs to move more in this direction, including replacing files with for example persistent variables. * Exception handling. 'Novice programmers find writing exception handling quite hard'. Steve has some proposals for improving exception handling. * POPLOG should not have a fixed user-settable upper memory limit as now. It should deduce upper and lower safe limits dynamically. * Use of the stack and variables should be made traceable by redefining the POPLOG VM. This gives the power of access-orientd programming (cf LOOPS). Sophisticated users can get the same effect with nullary operators, but something simpler is needed for beginners. * Lexical analysis of text is currently too inflexible (Incharitem needs improvements.) * User-definable system routines should be replaced by an explicit list of routines to be executed so that dynamic modifications are easier. >From Steve Knight, via Aaron Sloman, Sussex University. ------------------------------ End of AIList Digest ******************** From comsat@vtcs1 Wed Oct 23 00:12:53 1985 Date: Wed, 23 Oct 85 00:12:50 edt From: comsat@vtcs1.VT To: fox@opus (MILLER,FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a004205; 22 Oct 85 13:15 EDT Date: Tue 22 Oct 1985 08:57-PDT Reply-to: AIList%sri-ai.arpa@CSNET-RELAY US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #153 To: AIList%sri-ai.arpa@CSNET-RELAY Received: from rand-relay by vpi; Tue, 22 Oct 85 23:56 EST AIList Digest Tuesday, 22 Oct 1985 Volume 3 : Issue 153 Today's Topics: Queries - Mike O'Donnell & Expert Systems for Law Enforcement, Literature - AI Book by Jackson, AI Tools - YAPS & LISP Workstations, Logic - Counterfactuals, Opinion - SDI Software & AI Hype ---------------------------------------------------------------------- Date: Monday, 21 Oct 85 11:18:14 PDT From: wm%tekchips%tektronix.csnet@CSNET-RELAY.ARPA Subject: Looking for Mike O'Donnell I'm trying to contact Michael J. O'Donnell. I think he is now at University of Chicago, but I'm pretty sure he reads this list. Wm Leler (arpa) wm%tektronix@csnet-relay (csnet) wm@tektronix (usenet) decvax!tektronix!tekchips!wm ------------------------------ Date: 22 Oct 85 09:37 EDT From: Gunther @ DCA-EMS Subject: Expert Systems and Law Enforcement Need information about others working in the area of Expert Systems and Law Enforcement. We are compiling a bibliography of anything related to this area. Please reply with * Bibliographies * Reports * Names and Phone Numbers to * John Popolizio or Jerry Feinstein * Booz, Allen & Hamilton, Inc. * 301-951-2911 (2912) Thank You. ------------------------------ Date: Mon 21 Oct 85 20:40:50-PDT From: Ken Laws Subject: AI Book by Jackson I saw an ad for an inexpensive AI book in a Dover catalog: "Introduction to Artificial Intelligence" by Philip C. Jackson, Jr., second, enlarged edition, $8.95. Just published, it says. ------------------------------ Date: 21 Oct 85 11:27:28 EDT (Mon) From: Liz Allen Subject: YAPS YAPS may be redistributed but only after the whoever wants YAPS gets a license from the Univ of Maryland first. Contact Hans or me for the agreement which should be signed and sent to us. Once that is done, Hans can send you a copy of his code. -Liz [Liz also sent along a copy of the YAPS distribution agreement. For $100 (4.1 sources) or $250 (4.1 and 4.2) UMd includes their Franz flavors package, a window system, an editor, spreadsheet, Z80 emulator and cross-compiler, etc. Here is a description of YAPS: The YAPS production system written in Franz Lisp. This is similar to OPS5 but more flexible in the kinds of lisp expressions that may appear as facts and patterns (sublists are allowed and flavor objects are treated atomically), the variety of tests that may appear in the left hand sides of rules and the kinds of actions may appear in the right hand sides of rules. In addition, YAPS allows multiple data bases which are flavor objects and may be sent messages such as "fact" and "goal". Contact Liz for more details. -- KIL] ------------------------------ Date: 22 Oct 85 10:33:36 EDT (Tue) From: dndobrin@ATHENA.MIT.EDU Subject: LISP Workstations Re: Cugini's and Tatem's <148> flames. I agree that LISP machines are darned hard to learn; I also agree that they're worth the effort. My interests are twofold: why is it that intelligent, capable people like Cugini aren't willing to make the effort? How can the learning be made easier, or at least, more attractive. There's no easy solution. LISP machines are, in my experience, pretty well designed (at least by comparison with the hodgepodge in UNIX), and their documentation is, in most places, very good. (In my book, any documentation which tells you not to use something is already near the front of the pack.) Their online documentation, in particular, has received many awards; in one case, I was one of the judges, and they didn't even need to bribe me. Their documentation department does not consist of M.I.T. hackers; it has very experienced people in it. Jan Walker, their manager, is one of the best in the business. Sometimes, admittedly, they just slap the hacker's documentation onto the system, but most of the time, it shows real care. Then why is it so hard to learn? I think learning a complex system is very much like learning to play a complex game. No one learns chess in a month; very few people can even become competent at nim or rogue in a month. If you could get good quickly, the game would lose its interest. I think, therefore, that one can learn how to teach people LISP-machineology if one studies the way people learn games. Mostly, they learn from other people. In an informal study I did at M.I.T., I discovered that people learn to play rogue by watching other people play rogue and by asking more experienced players about what they should do whenever a difficult situation came up. People who play at home or who never ask don't play as well. Profound discovery. The analogy, however, is exact. Whenever you have many different things to do and the optimal move is not at all clear (or even calculable), you have to have some way of zeroing in on the close-to-optimal solutions. Documentation doesn't help you zero in, because it doesn't usually discuss that situation, and just finding what the documentation does say requires as much work as doing the original zeroing in. Experience does help you zero in. So if you don't have the experience, the easiest thing is to ask somebody who does. So, I would argue, the solution to Cugini's problem is to get Tatem to hang out over there for about three months. Maybe two. Well-designed systems and good documentation do make learning easier. With a well-designed system, when you are confronted with a puzzling situation, you can, in effect, consult the designer, figuring that he or she already knows the situation and has worked out some solution. Similarly, good documentation can often take you through some of the most common problem places. But even with good documentation and design, at some point, there you are on level 23, a griffin on one side, a dragon, on the other, 88 hit points, strength of 24, a +2, +2 two-handed sword, a wand of cold, and a wand of magic missile. What do you do? The analogy does break down in a funny way. In a game, you rarely get in a situation where no move does anything. But in learning a computer system, you often do. That's why Control-C is one of the first things everybody learns. ------------------------------ Date: Sat, 19 Oct 85 19:54:21 edt From: Dana S. Nau Subject: Counterfactuals > From: Mike Dante > . . . > (0) Suppose a class consists of three people, a 6 ft boy (Tom), a 5 ft girl > (Jane), and a 4 ft boy (John). Do you believe the following statements? > > (1) If the tallest person in the class is a boy, then if the tallest > is not Tom, then the tallest will be John. > (2) A boy is the tallest person in the class. > (3) If the tallest person in the class is not Tom then the tallest > person in the class will be John. > > How many readers believe (1) and (2) imply the truth of (3)? If we accept statement (0) as an axiom, then statement (3) is a statement of the form "A => B" whose antecedent A is false. Thus statement (3) is true, regardless of the truth or falsity of B. Since (3) is true, it is also true that (1) and (2) imply (3). The truth or falsity of (1) and (2) are irrelevant in this case. Although I haven't read the earlier articles in this discussion, I suspect your INTENT was that statement (3) would talk about a world identical to (0) except that Tom would not be present--and in this world statement (3) would be false. The only way I know of to handle this using logic is to state (3) in completely separate logical theory from the one whose axiom is (0). The rules for handling logical implications whose antecedents are counterfactuals don't correspond to our intuitive notions of how counterfactuals work. Although this isn't my area, I have an impression that the problem is a pretty thorny one. Dana S. Nau (dsn@rochester) From U. of Maryland, on sabbatical at U. of Rochester ------------------------------ Date: Tuesday, 15 October 1985, 23:31-EDT From: COWAN@MIT-XX Subject: SDI Software and AI hype I must respond to Prof. Minsky's perplexed comment that he does not understand computer scientists who argue against the computing requirements of SDI. I believe Minsky and those computer scientists are talking about two different things. Arguing for the feasibility of the software portion of SDI, Minsky says: ... aiming and controlling them should not be unusally difficult. The arguments I've seen to the contrary all seem to be political flames. Minsky is obviously talking about the technical software problem. But SDI opponents flame POLITICALLY, and so they MUST, because they are talking about the technical AND POLITICAL software problem. Even though I, like Minsky, believe that computers can eventually be made to think like humans, computers cannot RESPONSIBLY be used in place of humans in certain situations, especially those requiring political judgement. I believe that it is easy to see that a lot of POLITICAL decisions will have to be built into the SDI software. Therefore, the political aspects of the SDI software problem must be considered. A bare-bones, apolitical SDI system is one that doesn't consider the Soviet response (including a Soviet SDI). For such a system, I'm willing to agree that SDI software is feasible -- if not today, at least within my lifetime. But such a "vanilla-SDI" system would be worthless, unless there is a US-Soviet treaty outlawing countermeasures and anti-SDI systems, AND the SDIs of both countries (the USSR would not allow us to have a unilateral SDI edge) COULD NOT BE USED OFFENSIVELY. Unfortunately, it is impossible to imagine an SDI system that could not be easily software-upgraded to knock out the other country's SDI. SDI satellites are sitting ducks compared to missiles. A true SDI system would have to be programmed to react to situations where things go wrong, even if the problems are with the other country's SDI. If one country knocks out the other's SDI, then that country could launch a first strike under a protective unbrella -- an unacceptable situation for the country whose SDI was attacked. Thus, each SDI system would probably retaliate if the other SDI system attacked, even if the attack was a mistake. And each SDI system would probably fire on the opposing SDI system if a missile launch were detected on the other side. If you think about this situation for a while, you realize how serious and unstable it is. Even Charles Zraket, executive vice president of the Mitre Corporation, describes multiple SDIs as "the worst crisis-instability situation. It'd be like having two gunfighters in space armed to the teeth with quick-fire capabilities." The country whose trillion dollar SDI system is destroyed first would be tempted to "use 'em or lose 'em" -- to launch a first strike of its own. (Submarine launched cruise missiles could underfly any imaginable SDI.) Any hostile action (even upgrading the software!) could be perceived as an opening maneuver leading to a first strike. The decision of whether to retaliate, if made by a human being, would undoubtedly consider political circumstances on the ground (even statements in Pravda!). But time requirements would preclude human involvement; the software would have to determine a potentially grave response, using incomplete information, to situations for which it was not tested. To be "safe," each country would need an "SDSDI" to protect its SDI. But then, all the arguments of the previous paragraphs would still apply, at a higher defensive level. Boeing, Rockwell, Lockheed, and McDonnell Douglas might be content to build SDSDSDI's and SDSDSDSDI's, but the result would be decreasing stability, not increasing deterrence. The complexity of retaliatory policy would surpass the capabilities of policy makers, and certainly make "SDI control" an even more difficult problem than arms control is today. Why not solve the easier problem? It's fine to argue that software for a "vanilla-SDI" system is feasible, but it is intellectually dishonest to argue feasibility if you realize the true political nature of the software problem. People who write the hyped proposals aren't being dishonest, though. The SDI organization's requests for proposals don't describe the above scenario; they just ask people to study things like "reasoning under uncertain conditions." That's a fine goal, but it is the responsibility of computer scientists to make clear that such disciplines as "reasoning under uncertainty" are not applicable to problems as political as ballistic missile defense. Rich Cowan (cowan@mit-xx) [I would like to remind readers that SDI hype/feasibility is not necessarily AI hype/feasibility, and that ARMS-D@MIT-MC and POLI-SCI@RUTGERS seem to be the proper Arpanet fora for military/political discussions. I will attempt to screen out (censor, if you will) AIList submissions that do not focus on the AI aspects of the debate. -- KIL] ------------------------------ Date: Sat 19 Oct 85 01:02:09-PDT From: Gary Martins Subject: Grading the Professor In response to some simple questions about hype in "AI" [AIList #132], Prof. Minsky illustrates [AIList #139] rather than clarifies the subject matter. First, with respect to his own role in "AI" hype, the professor is evasive. Let's optimistically give him a grade of "Incomplete", and hope he will make up the work later, after he's had a chance to think about it some more. The only alternative is to conclude that he thinks it's OK to recklessly exaggerate the usefulness of "AI". Next, Prof. Minsky retells one of the most bizarre and colorful "AI" myths: that "AI" has somehow been responsible for all kinds of authentic real-world computing applications, such as: air traffic control, CAT scanners, avionics, industrial automation, radar signal processing, resource allocation, etc. !! But he rewards our patience with an astounding Revelation that pumps new wind into these nostalgic creations: that "AI" also deserves credit for the successes of information theory, pattern recognition, and control theory!!!!! Leapin' lizards !!! Could it really be that a field with this kind of distinguished and fruitful past would have regressed to today's fascination with such computationally trivial pursuits as: the "5th Generation", "blackboards", "knowledge engineering", OPS-5, R1/XCON, AM, BACON, EMYCIN, etc. ? Finally, Prof. Minsky proposes to stretch to 15-years the gestation period for "AI" products (an estimate that seems to grow just about linearly with time). But, 15 years FROM WHEN? Since modern "AI" is at least 30 years old, shouldn't we already have experienced a 15-year bonanza of genuine, concrete, real-world "AI" contributions? Where is it ? Every other area of computing can point to a steady succession of useful contributions, large and small. From "AI" the world seems to get back very little, other than amateurish speculations, wild prophecies, toy programs, unproductive "tools", and chamberpots of monotonous hype. What's wrong ? ------------------------------ End of AIList Digest ******************** From csvpi@vtcs1 Thu Oct 24 16:20:49 1985 Date: Thu, 24 Oct 85 16:20:44 edt From: csvpi@vtcs1.VT To: fox@vtopus (MILLER,FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a021261; 24 Oct 85 2:49 EDT Date: Wed 23 Oct 1985 23:01-PDT Reply-to: AIList%sri-ai.arpa@CSNET-RELAY US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #154 To: AIList%sri-ai.arpa@CSNET-RELAY Received: from rand-relay by vpi; Thu, 24 Oct 85 06:27 EST AIList Digest Thursday, 24 Oct 1985 Volume 3 : Issue 154 Today's Topics: Seminars - Representation of Natural Forms (MIT) & LOGIN: Logic Programming with Inheritance (UPenn) & Database Updates in Prolog (UPenn) & Concurrent Logic Programming (CMU) & Person Schemata (UCB) & NETTALK: Connectionist Speech Learning (UPenn) ---------------------------------------------------------------------- Date: Mon 21 Oct 85 09:55:17-EDT From: "Brian C. Williams" Subject: Seminar - Representation of Natural Forms (MIT) [Forwarded from the MIT bboard by SASW@MIT-MC.] Thursday 24, October 4:00pm Room: NE43- 8th floor Playroom The Artificial Intelligence Lab Revolving Seminar Series "Perceptual Organization And The Representation Of Natural Form" Alex P. Pentland AI Center, SRI Int'l and CSLI, Stanford To understand both perception and commonsense reasoning we need a representation that captures important physical regularities and that correctly describes the people's perceptual organization of the stimulus. Unfortunately, the current representations were originally developed for other purposes (e.g., physics, engineering) and are therefore often unsuitable. We have developed a new representation and used it to make accurate descriptions of an extensive variety of natural forms including people, mountains, clouds and trees. The descriptions are amazingly compact. The approach of this representation is to describe scene structure in a manner similar to people's notion of ``a part,'' using descriptions that reflect a possible formative history of the object, e.g., how the object might have been constructed from lumps of clay. For this representation to be useful it must be possible to recover such descriptions from image data; we show that the primitive elements of such descriptions may be recovered in an overconstrained and therefore reliable manner. An interactive ``real-time'' 3-D graphics modeling system based on this representation will be shown, together with short animated sequences demonstrating the descriptive power of the representation. ------------------------------ Date: Tue, 22 Oct 85 18:21 EDT From: Tim Finin Subject: Seminar - LOGIN: Logic Programming with Inheritance (UPenn) Colloquium 3pm 10-26-85 23 Moore, University of Pennsylvania LOGIN: A LOGIC PROGRAMMING LANGUAGE WITH BUILT-IN INHERITANCE HASSAN AIT-KACI A.I. Program, MCC, Austin, Texas Since the early days of research in Automated Deduction, inheritance has been proposed as a means to capture a special kind of information; viz., taxonomic information. For example, when it is asserted that "whales are mammals", we understand that whatever properties mammals possess should also hold for whales. Naturally, this meaning of inheritance can be well captured in logic by the semantics of logical implication. However, this is not operationally satisfactory. Indeed, in a first-order logic deduction system realizing inheritance as implication, inheritance from "mammal" to "whale" is achieved by an inference step. But this special kind of information somehow does not seem to be meant as a deduction step---thus lengthening proofs. Rather, its purpose seems to be to accelerate, or focus, a deduction process---thus shortening proofs. In this talk, I shall argue that the syntax and operational interpretation of first-order terms can be extended to accommodate for taxonomic ordering relations between constructor symbols. As a result, I shall propose a simple and efficient paradigm of unification which allows the separation of (multiple) inheritance from the logical inference machinery of Prolog. This yields more efficient computations and enhanced language expressiveness. The language thus obtained, called LOGIN, subsumes Prolog, in the sense that conventional Prolog programs are equally well executed by LOGIN. I shall start with motivational examples, introducing the flavor of what I believe to be a more expressive and efficient way of using taxonomic information, as opposed to straight Prolog. Then, I shall give a quick formal summary of how first-order terms may be extended to embody taxonomic information as record-like type structures, together with an efficient type unification algorithm. This will lead to a technical proposal for integrating this notion of terms into the SLD-resolution mechanism of Prolog. With examples, I shall illustrate a LOGIN interpreter. ------------------------------ Date: Tue, 22 Oct 85 11:35 EDT From: Tim Finin Subject: Seminar - Database Updates in Prolog (UPenn) Colloquium - 3pm 10-24-85 216 Moore, University of Pennsylvania A LOGICAL APPROACH TO DATABASE UPDATES IN PROLOG DAVID S. WARREN , SUNY AT STONY BROOK The power of the logic programming paradigm (exemplified by the Prolog programming language) lies in its close relationship to logic. This gives logic programs a clean, simple, and elegant declarative semantics, making them easy to understand and reason about. It has turned out, however, that in order to make Prolog a practical and usable programming language, several computational (and non-logical) extensions must be added. These extensions include the ``not'' operator, the ``setof'' operator, the ``var'' predicate, and the ``assert'' and ``retract'' operators. To the extent that these operators are non-logical, they destroy the declarative semantics of programs that use them. Such a program can only be understood by knowing its computation sequence. Progress has been made in providing a logical semantics for the ``not'' operator in Prolog, and the circumstances under which Prolog's negation as failure rule coincides with logical ``not'' are now reasonably well understood. This has allowed Prolog programs which use the ``not'' operator (meeting the appropriate constraints) to be understood declaratively. This talk describes an approach to providing a logical semantics for the Prolog operator ``assert''. We use a simple modal logic, which leads to a slightly different operational semantics for ``assert'' and suggests ways that the assert operator should be restricted in application. The resulting system has interesting implications for a theory of database updates. ------------------------------ Date: 22 Oct 1985 1109-EDT From: Lydia Defilippo Subject: Thesis Oral - Concurrent Logic Programming (CMU) ABSTRACT OF THESIS PROPOSAL Speaker: Vijay A. Saraswat Date: Friday - 1 November, 1985 Time: 10:00 am Place: 7220 Topic: CONCURRENT LOGIC PROGRAMMING LANGUAGES The domain of logic programming languages, consists, of the most part, of programming languages based on Horn logic which provide modified forms of top-down, SLD-refutation execution engines. A program in these languages consists of a set of definite clause axioms with (perhaps implicit) control information for guiding the underlying engine. Execution is initiated by the presentation of a conjunction of goals or queries and terminates when the engine, following the prescribed control, discovers either a proof of the goals, or the impossibility of such a proof. Concurrent logic programming (CLP) languages provide execution engines capable of pursuing concurrently proofs of each of the goals in a conjunctive system (so-called and-parallelism) and also different possible proof paths for each goal (or-parallelism). Examples of existing concurrent Horn languages are Concurrent Prolog, Parlog, GHC, Delta-Prolog and CP[!,|,&]. In this thesis I propose to lay a sound theoretical foundation for, and explore the paradigm of, CLP languages. Specifically, I propose to investigate the design, semantics, implementation and use of such languages. The thesis is intended to make contributions to each of the following areas: -- programming language design, via -- an understanding of the design space for concurrent programming languages based on annotated Horn logic, -- the design of a paradigmatic CLP language (CP[!,|,&,;]) providing a reasonably complete set of control structures for the parallel exploration of the refutation search space, and, -- an extensive comparison of CLP languages with related computational models outside the realm of logic programming, such as Actors, CSP, data-flow languages (including the systolic computational model) and constraint-based languages -- theoretical computer science via an understanding of the formal (operational and denotational) semantics of, and reasoning systems for, concurrent logic languages, including an understanding of the `logic' in such languages, -- programming language implementation, via a compiler-based implementation of the specific concurrent language CP[!,|,&] targetted to a uniprocessor machine, -- the `correct' design of efficient concurrent algorithms in the framework of unification-based concurrent logic programming languages, -- knowledge representation languages, via the design of a `higher-level' object-oriented, schema-based language featuring multiple inheritance with exceptions, and its implementation in Cp[!,|,&]. ------------------------------ Date: Wed, 23 Oct 85 13:52:25 PDT From: admin@cogsci.Berkeley.EDU (Cognitive Science Program) Subject: Seminar - Person Schemata (UCB) BERKELEY COGNITIVE SCIENCE PROGRAM Fall 1985 Cognitive Science Seminar - IDS 237A Tuesday, October 29, 11:00 - 12:30 240 Bechtel Engineering Center Discussion: 12:30 - 1:30 in 200 Building T-4 ``Person Schemata'' Mardi J. Horowitz M.D. Professor of Psychiatry, U.C.S.F. The speaker directs the recently formed Program on Cons- cious and Unconscious Processes of the John and Catherine T. MacArthur Foundation. Research on person schemata is one of the core agendas of this program. After a brief description of the program, the discussion will focus on clinical phenomena as segmented by different states of mind in a single individual. By examining the confi- guration in each state of mind as it occurs over time, it may be possible to infer what the self schemata and role relation- ship models are that organize thoughts, feelings and action into observed patterns. The theory that forms the basis for such inferences includes the postulate that each person's overall self organization may include a partially nested hierarchy of multiple self-concepts. A frequent set of states of mind in pathological grief reactions will provide a concrete illustration of phenomena, methods of inference, and a theory of person schemata. ------------------------------ Date: Wed, 23 Oct 85 16:43 EDT From: Tim Finin Subject: Seminar - NETTALK: Connectionist Speech Learning (UPenn) Colloquium - University Of Pennsylvania 3pm 10-29-85, 216 Moore NETTALK: TEACHING A MASSIVELY-PARALLEL NETWORK TO TALK Terrence J. Sejnowski Biophysics Department, Johns Hopkins University Text to speech is a difficult problem for rule-based systems because English pronunciation is highly context dependent and there are many exceptions to phonological rules. A more suitable knowledge representation for correspondences between letters and phonemes will be described in which rules and exceptions are treated uniformly and can be determined with a learning algorithm. The architecture is a layered network of several hundred simple processing units with several thousand weights on the connections between the units. The training corpus is continuous informal speech transcribed from tape recordings. Following training on 1000 words from this corpus the network can generalize to novel text. Even though this network was not designed to mimic human learning, the development of the network in some respects resembles the early stages in human language acquisition. It is conjectured that the parallel architecture and learning algorithm will also be effective on other problems which depend on evidential reasoning from previous experience. ------------------------------ End of AIList Digest ******************** From csvpi@vtcs1 Thu Oct 24 16:20:40 1985 Date: Thu, 24 Oct 85 16:20:36 edt From: csvpi@vtcs1.VT To: fox@vtopus (MILLER,FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a022153; 24 Oct 85 4:44 EDT Date: Wed 23 Oct 1985 23:04-PDT Reply-to: AIList%sri-ai.arpa@CSNET-RELAY US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #155 To: AIList%sri-ai.arpa@CSNET-RELAY Received: from rand-relay by vpi; Thu, 24 Oct 85 06:29 EST AIList Digest Thursday, 24 Oct 1985 Volume 3 : Issue 155 Today's Topics: Query - AI and Responsibility Panel, Literature - AI Book by Jackson, Philosophy - MetaPhilosophers Mailing List, News - New Jersey Regional AI Colloquium Series, Logic - Modus Ponens, AI Tools - AI Workstations, Opinion - SDI Software and AI Hype & Problems with Current Knowledge-Based Systems ---------------------------------------------------------------------- Date: Tue, 22 Oct 85 22:02:00 PDT From: Richard K. Jennings Subject: AI & Responsibility After reading comments on this net concerning the responsibility of AI systems, I finally got around to looking into the IJCAI proceedings. There was evidently a pretty lively panel discussion between people (one lawyer) who think that computers are the next group to be franchised as people (following blacks and women) and others (AI researchers) who tended to argue that computers are unreliable and bear close watching. Anybody out there attend the real thing and care to comment on how the oral discussion went? Any other comments on the Proceedings text? (pp1260+ in Vol II, IJCAI '85). Rich. ------------------------------ Date: Wed, 23 Oct 1985 00:36 EDT From: MINSKY%MIT-OZ@MIT-MC.ARPA Subject: AI Book by Jackson I am reading Jackson's AI book. It's very good and particularly in respect to the early decades of AI. I have seen no better way to get a picture of all the ideas of the 1960's, which many students don't know and do not always re-invent either. ------------------------------ Date: Tue 22 Oct 85 22:33-EDT From: Glen Daniels Subject: New mailing list MetaPhilosophers%MIT-OZ@MIT-MC.ARPA Discussion of personal philosophies, cosmologies, and metaphysical things. The place to air your ideas (or see others) on life, why we're here, what Mind is (as opposed to Brain), where our "selves" come from, what the universe is, what God is, any anything else in a metaphysical/philosophical vein. Send mail to MetaPhilosophers-Request%MIT-OZ@MIT-MC for more information or to be added. Everyone is welcome! --Gub (The MetaModerator) ------------------------------ Date: 23 Oct 85 15:59:04 EDT From: DRASTAL@RED.RUTGERS.EDU Subject: New Jersey regional AI colloquium series Dear Colleague, During the last IJCAI, it became clear to me that keeping in touch with other members of the AI community is only getting harder. Networks are not the right communication medium for reporting new work in progress, and the major conferences have grown too large for lively exchange. Yet, there are quite enough of us in the central New Jersey area who have something to say about our work in AI. This is why Dr. Yousry and I have decided to parent an informal colloquium series for researchers in this geographic area, and we invite your participation. Reports are welcome in areas ranging from theoretical foundations to implementation techniques. Anyone wishing to present or host a colloquium should send an abstract to one of us at the address below. We will coordinate the date, location, and distribution of announcements. Since this letter creates the series, it is most important that we hear from you now so that a distribution list can be compiled. Speakers will be recruited once we develop a critical mass of interested people. I know that we can look forward to some stimulating chain reactions among the participants. George A. Drastal Mona A. Yousry RCA AT&T Artificial Intelligence Laboratory Engineering Research Center Route 38 ATL Building P.O. Box 900 Moorestown Corporate Center Princeton, NJ 08540 Moorestown, NJ 08057 609-639-2405 609-866-6653 ihnp4!erc780!may DRASTAL@RUTGERS.ARPA ------------------------------ Date: 23 Oct 85 14:43:06 GMT From: Bob Stine Subject: re: modus ponens Mike Dante writes: > Suppose a class consists of three people, a 6 ft boy (Tom), a 5 ft girl >(Jane), and a 4 ft boy (John). Do you believe the following statements? > > (1) If the tallest person in the class is a boy, then if the tallest is > not Tom, then the tallest will be John. > (2) A boy is the tallest person in the class. > (3) If the tallest person in the class is not Tom then the tallest > person in the class will be John. > > How many readers believe (1) and (2) imply the truth of (3)? In answer to the last question - gosh, I sure do. The way the question is framed, however, blurs the distinctions between several separate issues. It would do to review what it means for a statement (or statements) to imply another, which is just that statement A implies statement B if and only if statement A and the negation of statement B are contradictory. If several statements imply B, then their conjunction is inconsistent with the negation of B. In the above example, whether or not (1) and (2) are true, false, or silly, (1) and (2) imply (3). What we believe about the truth or falsity of an argument's premises is quite another issue from the soundness of the argument. What clouds the issue, I think, is that you have introduced a contra-factual hypothesis in (3) (i.e., "assume that Tom is not tallest"). If we assumed that Tom were not tallest, then to preserve consistency, some or all of the other atomic suppositions (Jane is five feet, etc) would have to go. This would terminate the support for the argument's premises, and get us off the hook for asserting its conclusion. One final point. Note that (1) is equivalent to (1') A boy is not tallest or Tom is tallest or John is tallest. >From Mike's supposition - Tom is 6, Jane is 5, and John is 4 feet tall - we can deduce that Tom is tallest. It would be unusual to ask whether we believe the weaker statement (1') once we have established that Tom is indeed tallest. This points to another area where questions of logic part company from questions of belief - logic holds, even where questions of belief are inappropriate. - Bob Stine ------------------------------ Date: 23 Oct 85 08:57:00 EDT From: "CUGINI, JOHN" Reply-to: "CUGINI, JOHN" Subject: Son of DO you really need an AI machine? Since my original eruption provoked some responses (gratifyingly enough), I thought I'd indulge myself to a few comments on the comments. > I think, therefore, that one can learn how to teach people > LISP-machineology if one studies the way people learn games. > Mostly, they learn from other people. In an informal study I did > at M.I.T., I discovered that people learn to play rogue by watching > other people play rogue and by asking more experienced players about > what they should do whenever a difficult situation came up. People who > play at home or who never ask don't play as well. Profound discovery. Yes, I agree completely - we did not have a local Symbolics wizard, and no doubt this made my life more difficult. The situation is reflected in the fact that I had to *develop* (as I said earlier) about 10-12 pages of densely-packed cheat sheets, rather than *inheriting* them, and then customizing. > I agree that LISP machines are darned hard to learn; I also agree > that they're worth the effort. My interests are twofold: why is it > that intelligent, capable people like Cugini aren't willing to make the > effort? How can the learning be made easier, or at least, more > attractive. The heart of the issue here is I *did* make the effort and did get to the point of feeling reasonably comfortable (though I certainly did not attain wizardom) with the beast - I even knew by heart how to get out of the inspector! - and even with that I never felt I was quite getting my money's worth. I believe there are two factors: 1) My own style of programming leans away from spontaneity - perhaps I "over-design", but usually for me the coding is "merely" (hah!) a realization of an existing design. All the features of an AI-machine are focused on *coding and testing* - but by then in some sense the real work is done. Debugging aids are always helpful, of course, but I never really felt the need for all the exotic editor features. Perhaps also a lot of these features really come into their own only with truly large systems (> 5,000 lines). 2) the issue is always, not: is this AI-machine good?, but: is this AI-machine better than the alternatives? If the alternative is writing an expert system in BASIC with a line-oriented editor, then I too would kill to get on a Symbolics. But in my case (not wholly atypical, I think) the alternative was the use of VAX/VMS Common Lisp. My previous message discussed the costs of moving from a familiar, fully functional and maintained (by someone else, I'm pleased to say - who wants to do tape backup, anyway?) system to a new standalone machine. I should re-emphasize a point made in passing last time: the VAX implementation is very well done - it has a slightly intelligent editor (even blinks matching parens for you!), a good debugger, prettyprinter, etc etc. Now in one sense, the AI-machine advocates can crow: "well, the only reason you like the VAX is that they stole, er, borrowed some of the nifty techniques originally developed on AI machines." True enough, but I'm not giving out prizes for creativity; if I can get "most" of the advantages of an AI machine, together with those of a plain old VAX (FORTRAN, Pascal, SNOBOL4, mail to other people including ailist, laser printer, TEX, a single set of files, my very own terminal, free (to me) maintenance, etc..), isn't this the best deal? John Cugini Institute for Computer Sciences and Technology National Bureau of Standards Bldg 225 Room A-265 Gaithersburg, MD 20899 phone: (301) 921-2431 ------------------------------ Date: Wed, 23 Oct 1985 00:28 EDT From: MINSKY%MIT-OZ@MIT-MC.ARPA Subject: SDI Software and AI hype I agree generally with Cowan's analysis of that SDI debate: that I did not consider the "political software" problem. I don't know about the split-second decision problem, because you can complain that we can't program such things, but I'm not so confident about what the President would do in 30 seconds, either. IN any case, I repeat that I didn't mean to suggest that my opinion on SDI has any value because I haven't studied it. I was only reacting to what I thought were political reasons for dragging weak computer science arguments into the debate. As for SDI itself, my only considered opinion is based on meeting some of its principal supporters, and on the whole, they don't seem to be a suitably thoughful crowd to deserve the influence they've acquired. ------------------------------ Date: Wed 23 Oct 85 15:18:31-EDT From: MCCOWN@RADC-TOPS20.ARPA Subject: Fundamental problems with current knowledge based systems The following are my views of some of the fundamental problems with current engineering of knowledge based systems. Most of this is not new, but perhaps needs restating. These ideas have been stated by others in other forms before, but I would like to make sure this captures what has been said. Primarily, the systems are inflexible. If new information is input to the system which is not explicitly represented in the knowledge base, similar though it may be to previous inputs or existing representations, the system cannot deal with it unless explicitly told how. This lack of generalization and analogy capability causes a great bottleneck in the maintenence of the system, requiring experts and knowledge engineers to continuously update the knowledge base to reflect the current possibilities of input. In the rapidly changing real world this is unacceptable. The lack of ability to generalize and analogize is closely related to the ever present learning problem, and this not only affects the knowledge base maintenance problem, but the problem of knowledge acquisition as well. Currently, an inordinate number of hours of an expert's time are required in the interative process of knowledge acquisition. In addition, the capability of the knowledge engineer to understand the domain and to program such knowledge directly affects the quality of the system. A poor knowledge engineer makes for a poor system, regardless of the quality of the expert. While learning is a very general term, the type of learning referred here is the ability to recognize new and relevant information and its relation to information already known, and the ability to store that information and its relationships. While much work has been done in the representation of knowledge (related work being semantic net variations, frames, scripts, and MOPs for taxonomic and time ordered information, as well as production rules for procedural information, and predicate logic), no effective work has been done in getting information from a source into these representations, except for the method currently used - have a human (knowledge engineer) do it. Automated techniques to implement representations from examples (such as RuleMaster) are heavily domain dependent and are nothing more than complex weighted decision tables which work only for certain types of information. Other generalization work (such as RESEARCHER, IPP) are also heavily domain dependent, and are successful in capturing well only taxonomic information (A is a B, B works for C, etc.), and simple time-ordered information (A happened before B). Ways to recognize new related information, and (more important) new relevant related information are still lacking, as are ways of converting input information to consistent internal formats (consistent with the previous existing related knowledge). Indeed, even in the area of knowledge representation itself, the representations are often difficult to relate to other representations in a general way. Such relationships again depend upon the domain and are rigidily coded, creating difficulty in generalization and analogy. Time dependence, location dependence, non-monotonic reasoning, and uncertainty all require the programmer to jump through hoops to find ways to represent and relate information and procedures, forcing domain-dependent representations as well. Many of the problems in distributed and cooperating expert systems also stem from this apparent requisite to code knowledge in a domain- dependent fashion. Obviously if there are no generic techniques for coding knowledge, then a communication scheme must be developed to transfer information from one knowledge base to another (and as with any communication, often something is lost in the translation). It seems to be apparent that the learning problem is probably the most critical missing element in current knowledge based system technology, and that the knowledge representation issue may be the most critical element in the learning problem. This observation is not new, and this line of thought has been persued many times in the past. What I would like to add to the discussion is my belief that the current methods of knowledge representation are fundamentally incapable of solving the learning problem due to their discreteness. While discrete, cleanly delineated representations are (relatively) easy to work with, program well, and are easy to implement using the discrete representations that binary, Von Neuman machines offer, these representations, due to this very same discreteness, do not and can not represent reality in any generic and flexible way. I have an argument as to why, but I would like to here some criticism of the assertions set forth here first. No sense arguing from a shaky foundation. Solutions along the lines (or at least in the spirit of) coarse coding, distributed representation, etc., seem to be a possible solution to some of these problems. Obviously this type of discussion is related to that of "shallow structure" vs. "deep structure" in natural language processing. We can represent the shallow structure using these current representation an relational techniques. However, I feel that these techniques cannot effectively represent the deep structure owing to the property of their inherent discreteness. All relationships must be explicitly represented in these techniques, and are not implicit in the representation. Some means of content-addressable representation is required for implicit relationships between information. This is not to say that these representations are useless. On the contrary, they are very useful programming techniques for some types of analysis problems. They offer insights into problem areas in AI (such as learning), and they're representative of some real psychologically functional products of the human mind, and are useful in representing these products. However, it's time to ask "products of what", and to approach the "what" (learning) rather than the product (knowledge). Thanks for taking the time. Mike McCown mccown@RADC-TOPS20.ARPA RADC/IRDT Griffiss AFB, NY 13441-5700 ------------------------------ End of AIList Digest ******************** From csvpi@vtcs1 Sat Oct 26 07:40:43 1985 Date: Sat, 26 Oct 85 07:40:40 edt From: csvpi@vtcs1.VT To: fox@vtopus (MILLER,FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a009901; 26 Oct 85 3:11 EDT Date: Fri 25 Oct 1985 23:13-PDT Reply-to: AIList%sri-ai.arpa@CSNET-RELAY US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #156 To: AIList%sri-ai.arpa@CSNET-RELAY Received: from rand-relay by vpi; Sat, 26 Oct 85 07:27 EST AIList Digest Saturday, 26 Oct 1985 Volume 3 : Issue 156 Today's Topics: Query - PSL vs Common Lisp, AI Tools - Micro Lisps, Literature - AI Book by Jackson, Correction - Concurrent Logic Programming Languages, Opinion - AI Hype ---------------------------------------------------------------------- Date: 24 Oct 1985 14:51 EDT (Thu) From: Kimberle Koile Subject: PSL vs Common Lisp I'm interested in finding out about the differences between Portable Standard Lisp and Common Lisp. Specifically, how difficult would it be to take something that runs on a Symbolics machine (in Common Lisp) and make it run in PSL on a Vax/VMS or a Cray? Many thanks, Kimberle Koile ------------------------------ Date: Wed, 23 Oct 85 23:11:04 edt From: osiris!snk (Steve Kahane) Subject: Micro Lisps RE: Dr Blum's (BLUM@sumex) request for information on LISP products that run on micros: A paper comparing three products that run in the IBM series of personal computers (muLISP, IQLISP, GCLISP) will be presented at the 1985 Symposium on Computer Applications in Medical Care (SCAMC). Information will be presented on the following: Memory Addressing Capabilities Development Environment (error handling, debugging facilities, editing, graphics, windowing) Tutoring Tools Benchmarks Compilers (GCLISP (beta-test)) IQC-LISP? SCAMC meeting will be in Baltimore (Convention Ctr) on 11/10 - 11/13. For more info on meeting call (202) 676-4509. Reprints of the paper mentioned above are not yet available, but if anyone has any specific questions I would be glad to try to answer. Stephen N. Kahane (snk@osiris) Operational and Clinical Systems Halsted 124 The Johns Hopkins Medical Institution 600 North Wolfe St Baltimore, MD 21205 ------------------------------ Date: Fri, 25 Oct 1985 21:59 EDT From: MINSKY%MIT-OZ@MIT-MC.ARPA Subject: AI Book by Jackson The Jackson book is INTRODUCTION TO ARTIFICIAL INTELLIGENCE Philip C. Jackson, Jr. Dover Publications, New York, 1985 ------------------------------ Date: 25 Oct 1985 11:10-EDT From: Vijay.Saraswat@K.CS.CMU.EDU Subject: Concurrent Logic programming languages Lest there be any misunderstanding: the presentation on Nov. 1 at CMU is my thesis proposal NOT thesis defence! (The "Thesis Oral" in the Subject field was a secretarial oversight.) ------------------------------ Date: Thu 24 Oct 85 11:51:35-CDT From: David Throop Subject: Hype about Hype A CS professor recently told me that he was worried about the AI hype. He (who is in databases, not AI) fears that so much has been promised that there will be an anti-AI reaction and dissapointment that will hurt all of CS. And I've seen much posted on this list about the dangers of all this hype. The fears seem a bit overblown to me. I've gone through the professional employment adds in the New York Times and Wall Street Journal over the last weeks. I didn't notice ANYBODY advertising for hotshots in AI, Rule Based programming, LISP etc. The Austin American Statesman had one mention, but just the "it would also be nice if the candidate had some experience in..." form, what they were really looking for was UNIX. UT has been aswarm with recruiters recently. I'm not interviewing, but I've been talking with them in the halls and restaurants. Nobody up above seems to have told them to grab some heavy AI talent - most of them think experts systems are inferior to decision tree systems and are not impressed. The average American has > 14,000 commercial messages per WEEK aimed at them. I most people are pretty used to hype -we don't get our hopes up very easily. When I see the strong reactions to some of the blatant BS being said about AI, I'm puzzled. I suspect strongly that we're the only ones giving some of this stuff more than a second glance. Do you believe all the claims they make about your toothpaste? ------------------------------ Date: Fri, 25 Oct 85 12:22 EST From: "Christopher A. Welty" Subject: Contributions of AI I for one am tired of seeing this guy Gary Martins polluting the net with his childish attacks on Dr. Minsky. Did Minsky run away with his wife or something? Whatever the cause, keep your personal problems off the net. This should be for more productive discussions dealing with the field. Spending almost two hours some mornings reading my mail is not rewarding when it invloves sifting through accusations and wild generalizations, and things like this: >Every other area of computing can point to a steady >succession of useful contributions, large and small. From >"AI" the world seems to get back very little, other than >amateurish speculations, wild prophecies, toy programs, >unproductive "tools", and chamberpots of monotonous hype. >What's wrong ? [Gary Martins] Read a few books, Mr Martins. Maybe go for a trip to some research centers, or even to some companies and hospitals. Your messages are the only things I can see that fit into the categories of "amateurish speculations, wild prophecies, ..., and chamberpots of monotonous hype." AI has made significant contributions to Computer Science Research, and to the world. Underneath the "hype" are productive systems that are used to do such "toyish" things as diagnose illnesses, and control processes that were once controlled by humans (some of which were hazardous to those humans). These diagnostic and control Expert Systems come in many forms. The most sophisticated Data Base Systems in use today come from the knowledge-base systems sectors of AI research. Other examples are all around us, and there are too many to discuss, this message is long enough already. -Chris ------------------------------ Date: Fri 25 Oct 85 20:20:54-EDT From: Richard A. Cowan Subject: Causes of AI hype This is a response to Gary Martins' question about why AI is frequently hyped. [AILIST: Volume 3, #126] I thought about this for a while, so please tell me which points are weak, or if there are any factors I missed. Martins asks: - why this [hype] happens ? - is this good or bad for "AI" ? - does this happen in all high-tech fields, or is "AI" unique ? - what can or should be done about it ? by whom ? I offer a simple explanation: a lot of money is being pumped into AI to do things the field is not ready for. There are three different ways AI seems "not ready," depending on the intended application. 1. Some applications being funded are within the state of the art, but too few researchers are close enough to the state of the art to warrant the volume of money spent. 2. Other applications currently being attempted are beyond state of the art AI, but may be possible in 5 to 100 years. 3. Still other applications are forever beyond the capabilities of AI, because they involve responsibilities requiring human judgement. Reason number 1 generates hype because there is a continual stream of people from other fields into AI. They go and take crash courses in AI at various training centers, but what can they REALLY learn in one week? They get an excellent overview, which has to be optimistic about AI in order to justify the $3000 expense for the course. Perhaps the huge expenditure on AI training within industry is needed to rapidly enlarge the "AI labor force." But such expenditure puts a severe strain on engineering faculty supply and salaries at universities (MIT's former provost cited this as a primary cause of large tuition increases). This hurts university education in AI just when the need is most critical. It just might be better to slowly wait for the university AI community to enlarge since dramatic corporate funding increases at this stage also run the risk of damaging academic programs by institutionalizing the hype (i.e. MIT's 6.871 Expert Systems course). Reason number 2 generates hype because a disproportionate effort is devoted to goals which are not achievable. Most engineering fields are composed primarily of people applying well-understood engineering skills. A relatively small number of especially creative people do exploratory work advancing the engineering field itself. But in AI, since little is well-understood, almost everyone works on "novel ideas." Thus the "hype ratio" is very large. I know an AI manager at DEC who (previous to DEC) worked on government AI research contracts but now works on expert systems for industry. He is glad to be working on real problems; by contrast much DOD AI work was extremely detached from reality. When private sector profits are at stake, there must be something real underneath the hype for funding to be continued. For a contrast to this, I posted an inquiry about TIMM a couple of weeks back, which received 6 negative responses, and none positive. While problems with one product does not mean that a company is incapable of doing good work, the company (General Research Corporation) has received over $12 million dollars in software research contracts for the Strategic Defense Initiative (SDI) alone. I suspect the SDI office's budget for corporate research is so large and the talent pool so small that they can't be selective. Sorry to pick on General Research; I expect many other companies are the same. It's perfectly possible that such companies would do excellent work if the government would give them problems with a more immediate use to solve. I believe the Japanese 5th generation project will help Japan more than SCI helps us because it has a more commercial orientation. I also believe that the total US effort in AI (ONR + SDI + SCI) is too large. It's always exciting to attack unsolved problems, but AI initiatives of the recent mission-oriented nature consume an awful lot of resources. Why not devote some of those resources to unsolved problems such as acid rain which have interested professors but scant funds? -Rich ------------------------------ End of AIList Digest ******************** From comsat@vtcs1 Tue Oct 29 06:33:35 1985 Date: Tue, 29 Oct 85 06:33:32 est From: comsat@vtcs1.VT To: fox@vtopus (MILLER,FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a002727; 28 Oct 85 1:51 EST Date: Sun 27 Oct 1985 22:09-PST Reply-to: AIList%sri-ai.arpa@CSNET-RELAY US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #157 To: AIList%sri-ai.arpa@CSNET-RELAY Received: from rand-relay by vpi; Tue, 29 Oct 85 06:20 EST AIList Digest Monday, 28 Oct 1985 Volume 3 : Issue 157 Today's Topics: Reports - Davis Working Papers in Linguistics, Seminars - Iterative Knowledge Aggregation (UPenn) & Limits of Expert Systems (Ames) & Layered Control System for a Robot (MIT) & Machine Learning and Knowledge Representation (MIT) & Possible Histories (SRI) ---------------------------------------------------------------------- Date: Wed, 23 Oct 85 22:20:47 pdt From: ucdavis!harpo!lakhota@UCB-VAX.Berkeley.EDU Subject: Davis Working Papers in Linguistics From: Robert Van Valin, UC Davis ucdavis!harpo!lakhota@Berkeley We have started publishing a working paper series at Davis, and I think there are many linguists on the net who would be interested in it. Could you run the following blurb for the first issue in the series? Thanks. DAVIS WORKING PAPERS IN LINGUISTICS No. 1, 1985 Contents: `A lexical theory of auxiliary selection in Italian' Giulia Centineo, UC Berkeley `Clause linkage and zero anaphora in Mandarin Chinese' Liang Tao, UC Davis & Hunan Teacher's College `Aspects of the interaction of syntax and pragmatics: Discourse coreference mechanisms and the typology of grammatical systems' Robert D. Van Valin, Jr., UC Davis `Notes on Tepehua (Totonacan; Mexico) verbal semantics' James K. Watters, UC Berkeley & SIL Approx. 150 pages [This announcement of the report series is fine, but I have suppressed price and ordering information to comply with Arpanet regulations against commercial use. Contact "ucdavis!harpo!lakhota"@Berkeley for details. -- KIL] ------------------------------ Date: Thu, 24 Oct 85 00:43 EDT From: Tim Finin Subject: Seminar - Iterative Knowledge Aggregation (UPenn) 12:00 Monday Oct. 28th 303 Towne Building, University of Pennsylvania A Survey of Iterative Knowledge Aggregation Methods Robert Hummel Courant Institute, NYU Iterative knowledge aggregation methods are used to choose one of a finite set of labels about each of a set of objects. Relaxation labeling processes are one example; there are now numerous other techniques for combining information which is sometimes supportive of a hypothesis and sometimes mutually contradictory. In this talk, I compare and contrast these methods, including stochastic relaxation algorithms, constrained power methods, and the Dempster/Shafer theory of evidence formulation. For each methods, we review the form of the state space, the type of evidence which can be represented, and the updating and convergence properties of the method. ------------------------------ Date: Thu, 24 Oct 85 21:37:39 pdt From: eugene@AMES-NAS.ARPA (Eugene Miya) Subject: Seminar - Limits of Expert Systems (Ames) Terry Winograd, 11/5, 1030am National Aeronautics and Space Administration Ames Research Center Joint Ames AI Forum/RCR Branch Seminar SPEAKER: Terry Winograd Computer Science Dept. Stanford University TOPIC: Expert systems: How far can they go? ABSTRACT: We are in the midst of a great wave of enthusiasm about the potential for expert systems in every area of human life and work. There is no agreement, however, as to just how much they can achieve, and where they will run into fundamental limits. This talk will address some basic questions as to what expert systems can really be expected to do. I will describe the "blindness" that is inevitable in the process of articulating the "systematic domains" that are needed for computer manipulation, and argue that it leads to important limitations on what we can expect AI techniques to accomplish. DATE: 5 Nov. 1985 TIME: 1030am BLDG: 245 ROOM: Space Sci Aud. Tuesday POINT OF CONTACT: E. Miya PHONE NUMBER: (415)-694-6453 eugene@ames-nas.ARPA VISITORS ARE WELCOME: Register and obtain vehicle pass at Ames Visitor Reception Building (N-253) or the Security Station near Gate 18. See map below. Do not use the Navy Main Gate. Non-citizens (except Permanent Residents) must have prior approval from the Director's Office one week in advance. Submit requests to the point of contact indicated above. Non-citizens must register at the Visitor Reception Building. Permanent Residents are required to show Alien Registration Card at the time of registration. If you are a foreign national E-mailing a request, please include your nationality, and Visa Type and number. ------------------------------ Date: Sat, 26 Oct 85 16:58:36 EDT From: "Steven A. Swernofsky" Subject: Seminar - Layered Control System for a Robot (MIT) Thursday 31, October 4: 00pm Room: NE43- 8th floor Playroom The Artificial Intelligence Lab Revolving Seminar Series "A Layered Robust Control System for a Mobile Robot" Rod Brooks MIT AI Lab The AI Lab Mobile Robot project has built one robot and we are constructing a second. They are intended to autonomously wander around office areas of the lab at the same time as people are occupying those areas. The robots will eventually build maps of their surrondings and the second one will interact with the environment with an onboard manipulator. We describe a new architecture for controlling these mobile robots. Layers of control system are built to let the robot operate at increasing levels of competence. Layers are made up of asynchronous modules which communicate over low bandwidth channels. Each module is an instance of a fairly simple computational machine. Higher level layers can subsume the roles of lower levels by suppressing their outputs. However, lower levels continue to function as higher levels are added. The result is a robust and flexible robot control system. The talk will end with speculations on evolution and brains, and modelling them with the Unconnection Machine. ------------------------------ Date: Sat, 26 Oct 85 17:00:31 EDT From: "Steven A. Swernofsky" Subject: Seminar - Machine Learning and Knowledge Representation (MIT) Wednesday 30, October 4: 00pm Room: 405 Robinson Hall Northeastern University 360 Huntington Ave. Boston MA Northeastern University College of Computer Science Colloquium Brittleness, Tunnel Vision, Machine Learning and Knowledge Representation Prof. Steve Gallant Northeastern University A system is brittle if it fails when presented with slight deviations from expected input. This is a major problem with knowledge representation schemes and particularly with expert systems which use them. This talk defines the notion of Tunnel Vision and shows it to be a major cause of brittleness. As a consequence it will be claimed that commonly used schemes for machine learning and knowledge representation are pre- disposed toward brittle behavior. These include decision trees, frames, and disjunctive normal form expressions. Some systems which are free from tunnel vision will be described. INFO: Carole D Hafner ------------------------------ Date: Sun 27 Oct 85 18:48:47-PST From: LANSKY@SRI-AI.ARPA Subject: Seminar - Possible Histories (SRI) POSSIBLE HISTORIES Pat Hayes Schlumberger Palo Alto Research, AI Lab 11:00 AM, MONDAY, October 28 SRI International, Building E, Room EJ228 (new conference room) A history is a connected piece of space/time with 'natural' boundaries. Using these as a basic ontology for talking about events, processes, etc. has some advantages over some other frameworks, and doesn't have some of the disadvantages which are sometimes attributed to it. However, it does have one major problem, which is the difficulty of talking about alternative possible futures, to allow planning to be done. In this talk I discuss a new way of using histories which looks like it can overcome this problem. ------------------------------ End of AIList Digest ******************** From comsat@vtcs1 Thu Oct 31 07:08:02 1985 Date: Thu, 31 Oct 85 07:07:58 est From: comsat@vtcs1.VT To: fox@vtopus (MILLER,FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a020740; 30 Oct 85 14:36 EST Date: Wed 30 Oct 1985 10:17-PST Reply-to: AIList%sri-ai.arpa@CSNET-RELAY US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #158 To: AIList%sri-ai.arpa@CSNET-RELAY Received: from rand-relay by vpi; Thu, 31 Oct 85 06:53 EST AIList Digest Wednesday, 30 Oct 1985 Volume 3 : Issue 158 Today's Topics: Opinion - AI Hype, AI Tools - LISP Workstations ---------------------------------------------------------------------- Date: Mon 28 Oct 85 08:42:30-EST From: Richard A. Cowan Subject: AI Hype In response to David Throop's comment: The fears seem a bit overblown to me. I've gone through the professional employment adds in the New York Times and Wall Street Journal over the last weeks. I didn't notice ANYBODY advertising for hotshots in AI, Rule Based programming, LISP etc. The Austin American Statesman had one mention, but just the "it would also be nice if the candidate had some experience in..." form, what they were really looking for was UNIX. Experience demonstrates that hype is in greatest abundance when lots of money is involved. Therefore recruiting is not the place where you'd expect to find hype; after all, these ads hope to appeal to intelligent people. There have certainly been plenty of AI-related ads at MIT, though. The reason for this is quite simple: when applying for government research contracts, companies must list people qualified to work on the project. These companies are facing shortages; they often put each AI person on as many as 3 proposals. Sadly, many companies recruit here to buy contract winning power. Regardless of a person's programming ability, his or her MIT affiliation wins contracts. I do believe that you've picked up on a trend, though: the hype is decreasing. There has been an academic reaction against such hype. But the hype persists where no reaction has occurred: where reaction is self-censored by monetary interest in AI. For real hype, go to IJCAI or read some publication that appeals to defense markets (Defense Electronics, etc.); it's the military that has the money. I sympathize with your perspective; I recently would have agreed. As undergraduates, we are often shielded from the realities of military domination of our work, because our professors (rightfully) are uncomfortable with it. Please forward this to your concerned CS professor. Rich (cowan@mit-xx) ------------------------------ Date: Mon, 28 Oct 85 11:44 EST From: "Steven H. Gutfreund" Subject: Damage from AI Hype David Throop in AI-List V3 #156 mentions that he has not seen much damage from the AI hype to other areas of CS work. This may be true if you look at the commercial cs field (as he did with the New York Times and Texas Austin American) But I had some personal experiences in the last year that indicate to me that it has had a significant effect on CS RESEARCH LABS. I went to several research labs in the last year (DEC, MCC, Siemens, Tektronix, HP, NRL) looking for a research appointment. (I wanted to put off my PhD work for a little while). This process started in February, and it was about this time that the Industry down-turn became severe. Naturally the first areas affected was expansion of research groups hiring. This meant that most groups that were able to get open req's, were targeting them to AI slots, since the companies were perceiving that they "had to get into AI" or be left behind. Needless to say, I did not find a satisfactory position, nor did 3 other New PhD's in software engineering here, who decided it would be better to continue as Post-Docs. I am beggining to see a situation that is defining CS research as AI research, especially when one looks at DARPA and NSF funding. This need "to get into AI or be left behind" is out of control. I have a Tektronix Smalltalk machine that I obtained as a donation to do my Thesis work. (smalltalk based). Next week a bunch of executives from the local power company (a small outfit called Western Mass Elec) is coming here to see my work because it is "ON AN AI Machine". You see they have decided that they have to get into AI or be left behind so these executive officers are dragging out their EDP computing staff to see an AI machine and to get them to see the AI LIGHT! (I am going to feel very bad when I tell them I do Smalltalk, not AI). Don't these people have a better use for the scarce programming talent? This all reminds me of the standard IBM hard-sell techniques of the Thomas J. Watson days. Sell IBM to the top executives. Then force those those poor system programmers to learn/use OS-360. If you sell to the top, you can ignore the complaints of the knowlegable programmers below who will complain about the bestiality of the system. Speaking of bestiality, when at DEC I talked to the R1/XCON people. I saw this project when I worked at DEC 4 years ago. It now takes 18 people to maintain it (I was part of an operating system group that kept the whole OS, and utilites going with 6, and we did developement too!) Compare the functionality. - Steven Gutfreund gutfreund@umass.csnet ------------------------------ Date: 24 Oct 85 14:53:06 EDT (Thu) From: Liz Allen Subject: Re: LISP Workstations I'd like to point out a couple things that made lisp machines hard for me to learn -- that didn't quite coincide with the things that dndobrin@athena.mit.edu was suggesting. My background was strong in running Franz under Berkeley Unix and using the vi editor there. I'd already been maintaining the Maryland Franz Lisp environment for a couple years or so -- and that includes a flavors package. Lisp machine lisp was not a big problem. I was using a Symbolics machine in a place where there was only one of them and no one to give me any pointers -- or even much of a lispm-init file... I expected some learning time, but it was *much* harder to learn that I ever expected... Let me give some examples of problems I had. I should preface this by saying that I do like to use the machine now that I've gotten more comfortable with it, but... I was using "system f" to look at the file system (I hadn't found dired yet) and using the menu stuff pretty successfully. I was missing the notion of current working directory, but I was living with that (I hadn't figured out how to avoid typing long names yet). What I really needed was to see the files that were listed off the bottom of the screen... I tried going to the last file on the screen, but couldn't persuade the screen to scroll with a cr or anything. I'd seen scroll bars briefly before and wanted one of those, but there wasn't one on the screen so I wasn't sure I could use one there was for the window. I tried to look up how to scroll by looking under "window" in the documentation, but only found info on creating windows... I lived without scrolling for a while by closing and opening directories a lot and using ^S in emacs. Then, one day, the scroll bar suddenly appeared -- and then disappeared. But I wasn't sure what I'd done to get it. I experimented a little, but since I was neatly keeping the mouse within the window, I couldn't get the scroll bar back. I got it by accident one or two more times before I figured out that running the mouse into the left margin of the window would get it. Now, wouldn't a little documentation in some obvious place about how to *use* existing windows be a great help? The other big problem I had was in using emacs -- I learned about apropos pretty quickly, but it was not a lot of help. My favorite example is when I wanted to pick up some text without modifying the existing buffer. I tried apropos on "pick" which gave me nothing except the all too familiar single line "Done". Then I tried "yank" which told me how to put it back down again... (Yank in vi is how to pick up text and put it in a register...) That did give me the correct idea that I wanted to put something in the kill ring, so I did an apropos on "kill", but that didn't help. "Copy region" (M-W) was the command that I was looking for -- I stumbled across it by accident much later. My problem was vocabulary -- I knew the basic concepts, but it was hard to find out what the names for them were. I didn't have an emacs reference card (we didn't seem to have one with the Symbolics documentation). The emacs manual was too much reading for too little new info; it tends to assume you don't know anything about an editor. And the index is a lot like apropos -- you have to know the vocabulary. I finally decided that an apropos on something like "region" would probably be a good way to learn emacs verbs, but I had already learned enough to get by so never tried it. Some specific responses to dndobrin@ATHENA.MIT.EDU: LISP machines are, in my experience, pretty well designed (at least by comparison with the hodgepodge in UNIX), and their documentation is, in most places, very good. The documentation is good if you either already know the vocabulary or have someone who can tell you the right word for what you want. In UNIX, at least, apropos matches on descriptions of a command as well as the name of the command (though just using the name of the command under UNIX would be useless...). That would have helped me -- the apropos on "kill" would have given me copy region then. And, as I said above, some basic stuff on *using* windows would have been nice... Then why is it so hard to learn? I think learning a complex system is very much like learning to play a complex game. I'm not sure I like your game analogy -- why do you need a challenge to use a tool? It can be fun to learn, but I think you probably have enough of a challenge just debugging your application. It's not supposed to be you against the machine where the machine is trying to keep you from getting your program to run... (I know you didn't mean that.) Mostly, [new users] learn from other people. That's a good way to learn, but documentation ought to be able to stand alone. Whenever you have many different things to do and the optimal move is not at all clear (or even calculable), you have to have some way of zeroing in on the close-to-optimal solutions. That wasn't the issue. I knew exactly what I wanted to do. There are times when a way to do something on one machine isn't even appropriate for the other machine, but I had an appropriate solution in mind in both of my examples up there. I just didn't know how to use the mouse, etc, to accomplish it. So, I would argue, the solution to Cugini's problem is to get Tatem to hang out over there for about three months. Maybe two. I would recommend taking a course -- I haven't taken any (by the time they were offered, I was already more or less comfortable with the machine) so I don't know what levels are available -- or joining a Local Symbolics Users' Group. Otherwise, you can waste a lot of time searching documentation, etc, for rather obvious things. But even with good documentation and design, at some point, there you are on level 23, a griffin on one side, a dragon, on the other, 88 hit points, strength of 24, a +2, +2 two-handed sword, a wand of cold, and a wand of magic missile. What do you do? You didn't say what kind of armor you have! But I think you're an expert by then... Anyway, I'm not quite sure I follow the analogy -- I already knew how to debug programs if that's what you mean. But knowing what options are available (eg that you can wield a sword) and knowing how to do them (press the "w" key and do a menu select) is something the documentation should tell you without you needing to know the word "wield"; you should be able to find the info under "sword"... -Liz Allen liz@tove.umd.edu or liz@maryland.arpa seismo!umcp-cs!tove!liz ------------------------------ End of AIList Digest ******************** From csvpi@vtcs1 Wed Oct 30 23:07:18 1985 Date: Wed, 30 Oct 85 23:07:11 est From: csvpi@vtcs1.VT To: fox@vtopus (MILLER,FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a022440; 30 Oct 85 17:23 EST Date: Wed 30 Oct 1985 10:29-PST Reply-to: AIList%sri-ai.arpa@CSNET-RELAY US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #159 To: AIList%sri-ai.arpa@CSNET-RELAY Received: from rand-relay by vpi; Wed, 30 Oct 85 22:53 EST AIList Digest Wednesday, 30 Oct 1985 Volume 3 : Issue 159 Today's Topics: Queries - AI and Cognitive Psych in India & Statistical Expert Systems, Correction - BUG-LISPM@MIT-MC, Knowledge Representation - Text Understanding, Applications - LISP-Machine Tutors, Project - GUIDON-2 ---------------------------------------------------------------------- Date: Wed, 30 Oct 85 05:33:46 pst From: gluck@SU-PSYCH (Mark Gluck) Subject: Query: Info wanted on AI & Cognitive Psych in India I'm trying to collect information on scientists in India who are interested in--or doing research in--the areas of Artificial Intelligence or Cognitive Psychology. If you know of any such people or of any releveant research centers or Indian AI associations, I would be grateful if you could pass this information on to me at: gluck%su-psych@sumex-aim. To anyone interested in learning more about the Indian "Cognitive Science" community, I'd be happy to forward the collected information. -Mark Gluck Dept. of Psychology-SU Stanford, CA 94305 ------------------------------ Date: Sat, 26 Oct 85 15:16:37 PDT From: Stuart Crawford Subject: Statistical Expert Systems [Forwarded from the AI-ED list by Laws@SRI-AI.] I am interested in obtaining pointers to recent references regarding the known pros and cons of using pure statistical approaches to medical diagnosis (such as the use of classification and regression trees) as opposed to expert systems approaches. In particular, I am interested in any literature discussing the possible use of the combined use of such approaches. For example, using classification trees to help with the fine tuning of production rules, or using classification rules to augment current knowledge bases. I know much more about the statistical approaches than the ai approaches, but it seems that some interdisciplanary technique might be fruitful. Stuart Crawford [Bob Blum's RX/RADIX work at Stanford is the best reference I can suggest. -- KIL] ------------------------------ Date: Mon, 28 Oct 85 10:50 EST From: Christopher Garrigues <7thSon@SCRC-STONY-BROOK.ARPA> Subject: BUG-LISPM@MIT-MC In a recent AILIST digest, a recommendation was made that people interested in information on Symbolics Lisp Machine Information could join both SLUG@UTEXAS.ARPA and BUG-LISPM@MIT-MC.ARPA. The first is true. SLUG is the Symbolics Lisp Users Group and is exactly what it sounds like. The second is not true. BUG-LISPM@MIT-MC is where MIT users send their bug-reports to Symbolics and is exactly what that sounds like. The volume of requests to BUG-LISPM-REQUEST@MIT-MC got to the point that Symbolics Home Office Software Support (HOSS@SCRC-STONY-BROOK) has been added to that list. I believe that the confusion arose in the fact that MIT like most sites, allows LOCAL users to be on the bug mailing lists. In this case, local users are those who are students, faculty or otherwise associated with MIT. To subsume the purpose that BUG-LISPM@MIT-MC has for writing on, Symbolics customers with software contracts should send mail to HOSS@SCRC-STONY-BROOK.ARPA. If non-contract customers send legitimate bug reports to HOSS, we will note the report and forward it to the appropriate developers. As a source to read from, I believe there has been some discussion on the SLUG list of developing a second list for discussion of bugs and/or problems users have encountered along with bugfixes. You'll have to contact SLUG for futher information on that subject. I apologise for taking up the space on AILIST for a non-AI oriented entry, but since the overflow of traffic to BUG-LISPM-REQUEST@MIT-MC started here, I thought I should go to the source. Thanks for the audience, Chris Garrigues Symbolics Home Office Software Support (7thSon@SCRC-STONY-BROOK) ------------------------------ Date: Tue, 27 Aug 85 16:02:23 cst From: "V.J. Raghavan" Subject: Text Understanding [Excerpted from the IRList Digest by Laws@SRI-AI.] [This year's Montreal ACM SIGIR Conference on Information Retrieval had many interesting papers. A file, in "SMART" form, of the abstracts, was typed in at Univ. of Regina and edited at Virginia Tech. Order info for proceedings is: Proc. of the Eighth Annual Int. ACM SIGIR Conf. on R&D in Inf. Ret., ACM Order No. 606850 -- Ed] .I 10 .T Processing Free-Text Input to Obtain a Database of Medical Information .A EMILE C. CHI CAROL FRIEDMAN NAOMI SAGER MARGARET S. LYMAN .W The Linguistic String Project of New York University has developed computer programs that convert the information in free-text documents of a technical specialty into a structured form suitable for mapping into a relational database. The processing is based upon the restrictions on the use of language that are characteristic of the subject matter and the document type. These restrictions are summarized in a "sublanguage grammar" that provides a set of word classes and formulas corresponding to the objects and relations of interest in the domain. The programs are independent of the particular sublanguage grammar employed. The application to narrative patient records will be described and the applicability of the methods to other domains discussed. [...] ------------------------------ Date: Mon, 28 Oct 85 8:24:34 EST From: Bruce Nevin Subject: need for tutors Re: response to Cugini and Tatem by dndobrin@athena.mit.edu (3.153) This discussion reminds me of the LISP tutor developed at CMU. A brief search did not turn up the article that I read about it (John R. Anderson and Brian J. Reiser, The LISP Tutor, BYTE, April 1985), but I did find a piece to which it referred in its bibliography, `Minimalist Training' by John M. Carroll, Datamation, Nov. 1, 1984. Gist is that, of course, the ideal learning situation is me on one end of a log and Socrates on the other, but who can afford that, so the CMU folks set out to design a tutor for LISP that would provide the kind of immediate feedback at arbitrary depth that a human tutor would. And apparently with considerable success. Reference to a similar effort appeared in a talk by Pirolli at UCB last September 24 (according to Digest # 3.127) Since we can't `get Tatem to hang out over there for about three months' for all the `over theres' that there are, cloning software versions of his familiarity with LISP machines would seem to be the right answer. Bruce Nevin bn@bbncch.arpa BBN Communications 33 Moulton Street Cambridge, MA 02238 (617) 497-3992 ------------------------------ Date: Fri 25 Oct 85 13:40:35-PDT From: Mark Richer Subject: GUIDON-2 Project [Forwarded from the AI-ED list by Laws@SRI-AI.] Here's some information on the GUIDON project, including references: Mark Richer, Oct. 25th, 1985 The GUIDON project is an applied AI research project at the Knowledge Systems Laboratory, Computer Science Department, Stanford University. This project is investigating strategies for teaching diagnostic reasoning (specifically, medical diagnosis) using computers and knowledge-based systems technology. Part of the effort in this project has been to extend the capabilities of KB systems technology for the purpose of explanation and instruction. NEOMYCIN, a knowledge-based diagnostic consultation system, has been implemented and is the foundation for a new series of instructional programs, collectively called GUIDON-2. These programs are substantially different in design than the original GUIDON tutoring system that worked in conjunction with EMYCIN (e.g., MYCIN) systems. The director of the project is William J. Clancey, Ph.D., Senior Research Associate, Computer Science Department, Stanford University. There are about a dozen people associated with this project at present including a physician. Below is a list of references that might be of interest to people doing work in computer-based instruction. Papers that are listed as HPP or KSL technical reports are available by writing or calling Knowledge Systems Laboratory, 701 Welch Road, Bldg. C, Palo Alto, CA 94304, (415) 497-3444. STAN-CS papers (I think) are available through the Computer Science Department, Stanford University, Stanford CA 94305. WARNING: Do not send requests for papers to me; I'm afraid I will get swamped. Try to find the reference yourself if it was published, otherwise request it directly by calling or mailing to KSL or Stanford CS. (KSL is part of the CS Dept, but we are housed in a separate building at present and we maintain our series of technical reports.) Thank you. References: [these are not in any particular order] Clancey, W.J. (1979) Transfer of rule-based expertise through a tutorial dialogue. Computer Science Doctoral Dissertation, Stanford University, NOT Available as a tech report. Revised version, MIT Press, in preparation. Clancey, W.J. (1979) Tutoring rules for guiding a case method dialogue. Int J of Man-Machine Studies, 11, 25-49. Also in Intelligent Tutoring Systems, eds. Sleeman and Brown, Academic Press, London, 1982. Clancey, W.J. (1982) Overview of GUIDON. Journal of Computer-Based Instruction, Summer 1983, Volume 10, Numbers 1 & 2, pages 8-15. Also in The Handbook of Artificial Intelligence, Volume 2, eds. Barr and Feigenbaum, Kaufmann, Los Altos. Also STAN-CS-93-997, HPP-83-42. Richer, M. and Clancey, W. J. (1985) GUIDON-WATCH: A graphic interface for browsing and viewing a knowledge-based system. To appear in IEEE Computer Graphics and Applications, November 1985, Also KSL 85-20. Clancey, W.J., Bennett, J., and Cohen, P. (1979) Applications-oriented AI Research: Education. In The Handbook of Artificial Intelligence, Chapter IX, Volume 2, eds. Barr and Feigenbaum, Kaufmann, Los Altos. Also STAN-CS-79-749, HPP-79-17. Clancey, W.J., Shortliffe, E.H., and Buchanan, B.G. (1979) Intelligent computer-aided instruction for medical diagnosis. In Readings in Medical Artificial Intelligence: The First Decade, eds. W.J. Clancey and E.H. Shortliffe, Addison-Wesley, 1984. Also Proceedings of the Third Annual Symposium on Computer Applications in Medical Care, Silver Spring, Maryland, October 1979, pps. 175-183. Also HPP 80-10. Clancey, W.J. and Letsinger, R. (1981) NEOMYCIN: Reconfiguring a rule-based expert system for application to teaching. In Readings in Medical Artificial Intelligence: The First Decade, eds. W.J. Clancey and E.H. Shortliffe, Addison-Wesley, 1984. Proceedings of Seventh IJCAI, 1981, pps. 829-826. Also STAN-CS-82-908, HPP 81-2. Clancey, W.J. (1981) Methodology for Building an Intelligent Tutoring System. In Method and Tactics in Cognitive Science, eds. Kintsch, Miller, and Polson, Lawrence Erlbaum Associates, Hillsdale, New Jersey, 1984. Also STAN-CS-81-894, HPP 81-18. Clancey, W.J. (1984) Acquiring, representing, and evaluating a competence model of diagnosis. In Contributions to the Nature of Expertise, eds. Chi, Glaser, and Farr, in preparation. Also HPP-84-2. Clancey, W.J. (1979) Dialogue Management for Rule-based Tutorials. Proceedings of Sixth IJCAI, 1979, pps. 155-161. London, B. & Clancey W. J. (1982) Plan recognition strategies in student modeling: Prediction and description. Proceedings of AAAI-82, pps. 335-338. Also STAN-CS-82-909, HPP 82-7. Clancey, W.J. (1983) Communication, Simulation, and Intelligent Agents: Implications of Personal Intelligent Machines for Medical Education. Proceedings of AAMSI-83, pps. 556-560. Also HPP-83-3. Many people have influenced our thinking, but in particular the following paper may be helpful to understand our current thinking with regard to computer-based learning: @Inproceedings[BROWN83, key="Brown" ,Author="Brown, J.S." ,title="Process versus product--a perspective on tools for communal and informal electronic learning" ,booktitle="Education in the Electronic Age" ,note="Proceedings of a conference sponsored by the Educational Broadcasting Corporation, WNET/Thirteen Learning Lab, NY, pp. 41-58 ." ,month=July ,year=1983] The work described in this paper has its home at the XEROX Palo Alto Reasearch Center (PARC). ------------------------------ End of AIList Digest ******************** From csvpi@vtcs1 Fri Nov 1 01:29:05 1985 Date: Fri, 1 Nov 85 01:29:01 est From: csvpi@vtcs1.VT To: fox@vtopus (MILLER,FRANCE,JOSLIN,ROACH,FOX) Subject: From: AIList Moderator Kenneth Laws Status: RO Received: from sri-ai.arpa by CSNET-RELAY.ARPA id a003014; 31 Oct 85 17:36 EST Date: Thu 31 Oct 1985 10:18-PST Reply-to: AIList%sri-ai.arpa@CSNET-RELAY US-Mail: SRI Int., 333 Ravenswood Ave., Menlo Park, CA 94025 Phone: (415) 859-6467 Subject: AIList Digest V3 #160 To: AIList%sri-ai.arpa@CSNET-RELAY Received: from rand-relay by vpi; Fri, 1 Nov 85 01:09 EST AIList Digest Thursday, 31 Oct 1985 Volume 3 : Issue 160 Today's Topics: Seminars - Knowledge-Based Language Production (BBN) & Mechanical Verification of Mathematics (BBN) & Levels of Abstraction in Expert Systems (BBN) & Conversational Language System (BBN) & Correcting Misconceptions (BBN), Conferences - Economics and AI & AI Society of New England & Revised Call for Papers: OIS-86 ---------------------------------------------------------------------- Date: Thu, 31 Oct 85 00:56:16 EST From: "Steven A. Swernofsky" Subject: Seminar - Knowledge-Based Language Production (BBN) Friday 1, November 10: 30am Room: BBN Labs, 10 Moulton Street, 3rd floor large conference room BBN Artificial Intelligence Seminar "A Knowledge-Based Approach to Language Production" Paul Jacobs The development of natural language interfaces to Artificial intelligence systems is dependent on the representation of knowledge. A major impediment to building such systems has been the difficulty in adding sufficient linguistic and conceptual knowledge to extend and adapt their capabilities. This difficulty has been apparent in systems which perform the task of language production, i. e. the generation of natural language output to satisfy the communicative requirements of a system. The problem of extending and adapting linguistic capabilities is rooted in the problem of integrating abstract and specialized knowledge and applying this knowledge to the language processing task. Three aspects of a knowledge representation system are highlighted by this problem: hierarchy, or the ability to represent relationships between abstract and specific knowledge structures; explicit referential knowledge, or knowledge about relationships among concepts used in referring to concepts; and informity, the use of a common framework for linguistic and conceptual knowledge. The knowledge based approach to language production addresses the language generation task from within the broader context of the representation and application of conceptual and linguistic knowledge. This knowledge based approach has led to the design and implementation of a knowledge representation framework, called Ace, geared towards facilitating the interaction of linguistic and conceptual knowledge in language processing. Ace is a uniform, hierarchical representation system, which facilitates the use of abstractions in the encoding of specialized knowledge and the representation of the referential and metaphorical relationships among concepts. A general purpose natural language generator, KING (Knowledge INtensive Generator), has been implemented to apply knowledge in the Ace form. The generator is designed for knowledge intensivity and incrementality, to exploit the power of the Ace knowledge in generation. The generator works by applying structured associations, or mappings, from conceptual to linguistic structures, and combining these structures into grammatical utterances. This has proven to be a simple but powerful mechanism which is relatively easy to adapt and extend. ------------------------------ Date: Thu, 31 Oct 85 02:24:17 EST From: "Steven A. Swernofsky" Subject: Seminar - Mechanical Verification of Mathematics (BBN) Thursday 31, October 10: 30am Room: BBN Labs, 10 Moulton Street, 2nd floor large conference room BBN Laboratories Science Development Program AI Seminars Toward the Mechanical Verification of Abstract Mathematics David McAllester MIT AI Laboratory To mechanically verify a mathematical argument one must translate the argument into some formal language. Many mathematicians doubt that it will ever be practical to translate arbitrary mathematical arguments into a completely formal language. This talk will present a formal language called ONTIC which extends set theory in a way that supports an "object oriented" style of mathematical description. Ontic has been used to formally define some basic concepts of modern algebra, real analysis, and homotopy theory. We feel that any branch of modern mathematics can be concisely expressed in ONTIC. Furthermore it seems practical to translate any mathematical proof into a sequence of ONTIC formulas. A theorem- proving system has been constructed for ONTIC and some simple verifications have been done. ------------------------------ Date: 28 Oct 1985 11:01-EST From: Brad Goodman Subject: Seminar - Levels of Abstraction in Expert Systems (BBN) [Forwarded from the MIT bboard by SASW@MIT-MC.] Speaker: Prof. B. Chandrasekaran Laboratory for Artificial Intelligence Research Department of Computer and Information Science The Ohio State University Date: 10:30am, Monday, November 4th Place: BBN Labs, 10 Moulton Street, 3rd floor large conference room Generic Tasks in Knowledge-Based Reasoning: Characterizing and Designing Expert Systems at the "Right" Level of Abstraction We outline the elements of a framework for expert system design that we have been developing in our research group over the last several years. This framework is based on the claim that complex knowledge-based reasoning tasks can often be decomposed into a number of generic tasks each with associated types of knowledge and family of control regimes. At different stages in reasoning, the system will typically engage in one of the tasks, depending upon the knowledge available and the state of problem solving. The advantages of this point of view are manifold: (i) Since typically the generic tasks are at a much higher level of abstraction than those associated with first generation expert system languages, knowledge can be represented directly at the level appropriate to the information processing task. (ii) Since each of the generic tasks has an appropriate control regime, problem solving behavior may be more perspicuously encoded. (iii) Because of a richer generic vocabulary in terms of which knowledge and control are represented, explanation of problem solving behavior is also more perspicuous. We briefly describe six generic tasks that we have found very useful in our work on knowledge-based reasoning: classification, state abstraction, knowledge-directed retrieval, object synthesis by plan selection and refinement, hypothesis matching, and assembly of compound hypotheses for abduction. ------------------------------ Date: 28 Oct 1985 11:01-EST From: Brad Goodman Subject: Seminar - Conversational Language System (BBN) [Forwarded from the MIT bboard by SASW@MIT-MC.] Speaker: Prof. Janet Murray Dept. of Humanities, MIT Date: 10:30am, Tuesday, November 5th Place: BBN Labs, 10 Moulton Street, 2nd floor large conference room The Next Generation of Language Lab Materials: Developing Prototypes at MIT MIT's Athena Language Learning Project is a five-year enterprise whose aim is to develop prototypes of the next generation of language-lab materials, particularly conversation-based exercises using artificial intelligence to analyse and respond to typed input. The exercises are based upon two systematized methods of instruction that are specialties at MIT: discourse theory and simulations. The project is also seeking to incorporate two associated technologies: digital audio and interactive video. The digital audio sub-project is developing exercises for intonation practice, initially focusing on Japanese speakers learning English. The interactive video component of the project consists of preparation of a demonstration disc which features a variety of interactive video approaches including enhancement of the text-based simulations and presentation of dense conversational material in natural settings. The project is being developed on the Athena system at MIT, and is based upon the model of a near-future language lab/classroom environment that will include stations capable of providing interactive video, digital audio, and AI-based exercises. ------------------------------ Date: 28 Oct 1985 11:01-EST From: Brad Goodman Subject: Seminar - Correcting Misconceptions (BBN) [Forwarded from the MIT bboard by SASW@MIT-MC.] Speaker: Prof. Kathleen F. McCoy University of Delaware Date: 10:30am, Friday, November 8th Place: BBN Labs, 10 Moulton Street, 3rd floor large conference room Correcting Object Related Misconceptions Analysis of a corpus of naturally occurring data shows that users conversing with a database or expert system are likely to reveal misconceptions about the objects modelled by the system. Further analysis reveals that the sort of responses given when such misconceptions are encountered depends greatly on the discourse context. This work develops a context-sensitive method for automatically generating responses to object-related misconceptions with the goal of incorporating a correction module in the front-end of a database or expert system. The method is demonstrated through the ROMPER system (Responding to Object-related Misconceptions using PERspective) which is able to generate responses to two classes of object-related misconceptions: misclassifications and misattributions. The transcript analysis reveals a number of specific strategies used by human experts to correct misconceptions, where each different strategy refutes a different kind of support for the misconception. In this work each strategy is paired with a structural specification of the kind of support it refutes. ROMPER uses this specification, and a model of the user, to determine which kind of support is most likely. The corresponding response strategy is then instantiated. The above process is made context sensitive by a proposed addition to standard knowledge-representation systems termed "object perspective." Object perspective is introduced as a method for augmenting a standard knowledge-representation system to reflect the highlighting affects of previous discourse. It is shown how this resulting highlighting can be used to account for the context-sensitive requirements of the correction process. ------------------------------ Date: Wed 30 Oct 85 21:23:18-PST From: Ken Laws Subject: Conference - Economics and AI See Communications of the ACM, September 1985, p. 1008, for an announcement of the 1st Int. Conf. on Economics and AI (including management science, organizational and behavioral sciences, etc.), to be held in Aix-en-Provence, France, on September 2-4, 1986. ------------------------------ Date: Tue 29 Oct 85 20:13:44-EST From: Michael Lebowitz Subject: Conference - AI Society of New England THE SEVENTH ANNUAL CONFERENCE OF THE ARTIFICIAL INTELLIGENCE SOCIETY OF NEW ENGLAND NOVEMBER 1-2, 1985, BRANDEIS UNIVERSITY, WALTHAM, MA NATHAN SEIFER AUDITORIUM, IN FORD HALL Friday, November 1, 1985 8:00 PM Invited talk by Drew McDermott (Yale University) Easy and Hard Problems in Artificial Intelligence Abstract -- AI has not exactly solved everything. In fact, the more we progress the harder problems we uncover. However, some supposedly hard problems look as if they will evaporate completely. In this talk I will discuss: ancient problems that now look easy, like free will and consciousness; modern problems that are hard, like representing spatial knowledge; ancient problems that are still hard, like the nature of explanation and induction. 9:00 PM Traditional AISNE social hour Saturday, November 2, 1985 10:00 AM 15 minute talks Robert McCartney (Brown University) Algorithmic Synthesis Tom Ellman (Columbia University) Explanation Based Generalization of Logic Circuit Designs Dave Glaubman (Northeastern University) A Novice System for Bidding in Bridge Robert S. Rist (Yale University) Plans in Programming Brian Otis (University of New Hampshire) Knowledge-based Guidance for an Autonomous Underwater Vehicle 11:30 AM Panel chaired by John Kender (Columbia University) Are Vision and Robotics AI? 12:30 PM Lunch Break 2:00 PM more 15 minute talks Henry A. Kautz (University of Rochester) Plan Recognition as Theory Formation Mary P. Harper (Brown University) Tense and Time in English Tony Maddox (Brandeis University) A Parallel Approach to Generating Visual Event Descriptions Marie Vaughan (University of Massachusetts) Rewriting and Regeneration: A Computational Model of the Writing Process Ben Moreland (University of Connecticut) Artificial Ingelligence Research at UConn 3:30 PM still more 15 minute talks Marie Bienkowski (Princeton University) Generation of Elaborations: A Goal-Directed Model Steven Hanks (Yale University) Temporal Reasoning and Default Logic Hon Wai Chun (Brandeis University) Progress Towards Massively Parallel Speech Recognition Richard N. Pelavin (University of Rochester) A Formal Logic that Supports Planning with a Partial Description of the Future 4:30 PM AISNE business meeting -- volunteers for organizing next year's conference will be solicited. There is no registration fee for AISNE, but a small donation is requested to cover the costs of the Friday night social hour. Program chairman: Local arrangements: Professor Michael Lebowitz Tony Maddox Department of Computer Science Brandeis University 450 Computer Science Building Computer Science Department Columbia University Ford Hall 3-227 New York, NY 10027 Waltham, MA 02254 212-280-8196 617-647-2119 lebowitz@columbia-20.arpa tony%brandeis@csnet ------------------------------ Date: Tue, 29 Oct 85 11:26 EST From: Hewitt@MIT-MC.ARPA Subject: REVISED call for papers: OIS-86 ******************* C A L L F O R P A P E R S * * ---------------------------------------------- * * Third ACM Conference On * * OFFICE INFORMATION SYSTEMS * OIS-86 * * * October 6-8, 1986 * * Biltmore Plaza Hotel * * Providence, RI ******************* ------------------------------------------------- General Chair: Carl Hewitt, Topics appropriate for this MIT conference include (but are not restricted to) the following as they Program Chair: Stanley Zdonik, relate to OIS: Brown University Technologies including Display, Voice, Treasurer: Gerald Barber, Telecommunications, Print, etc. Gold Hill Computers Human Interfaces Local Arrangements: Andrea Skarra, Brown University Deployment and Evaluation An interdisciplinary conference on System Design and Construction issues relating to office information systems (OIS) sponsored Goals and Values by ACM/SIGOA in cooperation with Brown University and the MIT Distributed Services and Applications Artificial Intelligence Laboratory. Submissions from the following Knowledge Bases and Reasoning fields are solicited: Distributed Services and Applications Anthropology Artificial Intelligence Indicators and Models Cognitive Science Computer Science Needs and Organizational Factors Economics Management Science Impact of Computer Integrated Psychology Manufacturing Sociology The following have confirmed their membership on the program committee: Guiseppe Attardi Ray Panko University of Pisa University of Hawaii James Bair Robert Rosin Hewlett Packard Syntrex Gerald Barber Erik Sandewall Gold Hill Computers Linkoping University Peter de Jong Walt Scacci MIT USC Irene Greif Andrea Skarra MIT Brown University Sidney Harris Susan Leigh Star Georgia State University Tremont Research Institute Carl Hewitt Luc Steels MIT University of Brussels Heinz Klein Sigfried Treu SUNY University of Pittsburgh Fred Lochovsky Dionysis Tsichritzis University of Toronto University of Geneva Fanya Montalvo Eleanor Wynn MIT Brandon Interscience Naja Naffah Aki Yonezawa Bull Transac Tokyo Institute of Technology Margrethe Olson Stanley Zdonik NYU Brown University The invited keynote speaker is Professor J.C.R. Licklider of MIT. Unpublished papers of up to 5000 words (20 double-spaced pages) are sought. The first page of each paper must include the following information: title, the author's name, affiliations, complete mailing address, telephone number and electronic mail address where applicable, a maximum 150-word abstract of the paper, and up to five keywords (important for the correct classification of the paper). If there are multiple authors, please indicate who will present the paper at OIS-86 if the paper is accepted. Proceeedings will be distributed at the conference and will later be available from ACM. Selected papers will be published in the ACM Transactions on Office Information Systems. Please send eight (8) copies of the paper to: Prof. Stan Zdonick OIS-86 Program Chair Computer Science Department Brown University P.O. Box 1910 Providence, RI 02912 DIRECT INQUIRIES TO: Margaret H. Franchi (401) 863-1839. IMPORTANT DATES Deadline for Paper Submission: February 1, 1986 Notification of Acceptance: April 30, 1986 Deadline for Final Camera-Ready Copy: July 1, 1986 Conference Dates: October 6-8, 1986 ------------------------------ End of AIList Digest ********************