nl-kr-request@CS.ROCHESTER.EDU (12/16/87)
NL-KR Digest (12/15/87 21:07:46) Volume 3 Number 62
Today's Topics: (seminars and conferences)
Seminar - Composing and Decomposing Universal Plans (SRI)
Seminar - Generating NL Text (Bell Labs)
BBN AI Seminar -- David Chapman
ACM-SIGIR88 : LAST CALL
Seminar - Practical Reasoning (BBN)
From CSLI Calendar, December 10, 3:10
Intl. Conference on Principles of KR&R
CfP - ICSC 88
Seminar - The Psychology of Everyday Things (Suny Buffalo)
Cognitive Science Calendar
----------------------------------------------------------------------
Date: Thu, 3 Dec 87 19:03 EST
From: Amy Lansky <lansky@venice.ai.sri.com>
Subject: Seminar - Composing and Decomposing Universal Plans (SRI)
COMPOSING AND DECOMPOSING UNIVERSAL PLANS
Marcel Schoppers
Advanced Decision Systems (MARCEL@ADS.ARPA)
11:00 AM, MONDAY, December 7
SRI International, Building E, Room EJ228
``Universal plans'' are representations for robot behavior; they are
unique in being both highly reactive and automatically synthesized. As
a consequence of this plan representation, subplans have conditional
effects, and hence there are conditional goal conflicts. When block
promotion (= subplan concatenation) cannot remove an interaction, I
resort not to individual promotion (= subplan interleaving) but to
confinement (falsifying preconditions of the interaction). With
individual promotion out of the way, planning is a fundamentally
different problem: plan structure directly reflects goal structure,
plans can be conveniently composed from subplans, and each goal
conflict needs to be resolved only once during the lifetime of the
problem domain. Conflict analysis is computationally expensive,
however, and interactions may be more easily observed at execution
time than predicted at planning time.
All conflict elimination decisions can be cached as annotated
operators. Hence it is possible to throw away a universal plan, later
reconstructing it from its component operators without doing any
planning. Indeed, an algorithm resembling backchaining mindlessly
reassembles just enough of a universal plan to select an action that
is helpful in the current world state. Since the selected action is
both a situated response and part of a plan, recent rhetoric about
situated action as *opposed* to planning is defeated.
VISITORS: Please arrive 5 minutes early so that you can be escorted up
from the E-building receptionist's desk. Thanks!
------------------------------
Date: Mon, 7 Dec 87 10:36 EST
From: allegra!dlm
Subject: tSeminar - Generating NL Text (Bell Labs)
From Water to Wine: Generating Natural Language Text
From Today's Applications Programs
David McDonald
Brattle Research Corporation
Tues, Dec 8 (definite)
1:00 pm (tentative)
AT&T Bell Laboratories MH-3D-473 (tentative)
Today's AI programs all too often cut corners in the conceptual
models they reason with. To generate natural sounding texts
from these models, one needs to compensate for these semantic
deficits but without compromising the principled grammatical
treatments in the generator. I will talk about how the
interface to our generator, Mumble-86, handles these problems,
using examples from several different knowledge-based programs.
Sponsor: B. Ballard - allegra!bwb
------------------------------
Date: Mon, 7 Dec 87 19:40 EST
From: Marc Vilain <MVILAIN@G.BBN.COM>
Subject: BBN AI Seminar -- David Chapman
BBN Science Development Program
AI Seminar Series Lecture
PENGI: AN IMPLEMENTATION OF A THEORY OF ACTIVITY
David Chapman
MIT AI Laboratory
(ZVONA%OZ@MC.LCS.MIT.EDU)
BBN Labs
10 Moulton Street
2nd floor large conference room
10:30 am, Tuesday December 15
AI has generally interpreted the organized nature of everyday activity
in terms of plan-following. Nobody could doubt that people often make
and follow plans. But the complexity, uncertainty, and immediacy of
the real world require a central role for moment-to-moment
improvisation. Before and beneath any planning ahead, one continually
decides what to do *now*. Investigation of the dynamics of
everyday routine activity reveals important regularities in the
interaction of very simple machinery with its environment. Phil Agre
and I have used these ideas to design a program, called Pengi, that
engages in complex, apparently planful activity without requiring
models of the world.
------------------------------
Date: Tue, 8 Dec 87 09:05 EST
From: Pierre LAFORGUE <mcvax!imag!pierre@uunet.UU.NET>
Subject: ACM-SIGIR88 : LAST CALL
CALL FOR PAPERS
SIGIR 88
in cooperation with the ACM
11th INTERNATIONAL CONFERENCE
ON
RESEARCH AND DEVELOPMENT IN INFORMATION RETRIEVAL
ACM-SIGIR
JUNE 13-15 1988
GRENOBLE (FRANCE)
Conference Chairman : Yves CHIARAMELLA (USTMG - Grenoble, France)
Program Comittee :
M.ADIBA (F) G.KNORZ (Germany)
R.BOUCHE (F) S.MIRANDA (F)
A.BOOKSTEIN (USA) C.D.PAICE (UK)
M.F.BRUANDET (F) F.RABITTI (I)
E.CHOURAQUI (F) V.V.RAGHAVAN (USA)
W.B.CROFT (USA) K.VAN RIJSBERGEN (UK)
T.E.DOSZKOCS (USA) G.SALTON (USA)
A.S. FRAENKEL (Israel) P.WILLETT (UK)
N.FUHR (Germany) S.K.M. WONG (Canada)
Papers are invited on theory, methodology, implementation and applications of
information retrieval.
Communications from areas of prime interest for information retrieval, such
as artificial intelligence, database systems, office automation, hardware
technology, natural language processing, are welcome.
The main topics thus include, but are not limited to:
- retrieval system modelling :
linguistic models
mathematical models
cognitive and semantic models
- information retrieval and artificial intelligence:
knowledge representation
expert systems
thesaurus management
- evaluation techniques:
retrieval and system performances
system development and evaluation
- natural language processing:
parsers
deep understanding
multilingual systems
- information retrieval and database management:
storage and research techniques
multimedia databases
fifth generation databases
deductive databases
document databases for ofice automation
database machines
- user interfaces:
natural language interfaces
graphic interfaces
- advanced applications
INSTRUCTIONS TO AUTHORS :
Full length papers should not exceed 20 or 25 pages. Extented abstracts of
about 10 pages are also accepted. Both must contain a complete author
identification and an abstract of about a hundred words. Four copies of each
paper should be submitted to the Program Committee. Papers from North America
should be sent to G.SALTON; submissions from outside North America should be
sent to E.CHOURAQUI:
Gerard SALTON Eugene CHOURAQUI
CORNELL UNIVERSITY GRTC-CNRS
Dept. of Computer Science 31 chemin J.AIGUIER
4130 UPSON HALL 13402 MARSEILLE
ITHACA Cedex 9
N.Y. 14853 - 7501 USA FRANCE
Important dates :
submission deadline : january 15, 1988
acceptance notification : march 21, 1988
final copy due : may 16, 1988
Communication ways :
electronic address : siri@imag.UUCP
telex address : 98 01 34
telecopy address : 76 51 48 48
==============================================================================
= SIGIR 88 - REPLY MESSAGE : =
= ------------------------- =
= =
= Please return to Y.CHIARAMELLA =
= - electronic address : siri@imag.imag.fr =
= or {uunet.uu.net|mcvax}!imag!siri =
= - mail address : Laboratoire IMAG - Genie Informatique =
= BP 68 - 38402 Saint Martin d'Heres Cedex =
= FRANCE =
= =
= Last Name, First Name : ------------------------------------- =
= Address : --------------------------------------------------- =
= --------------------------------------------------- =
= --------------------------------------------------- =
= Electronic address : ---------------------------------------- =
= - I intend to participate the Conference =
= and want to receive the final program =
= =
= - I intend to submit a paper : - selected topic : =
= - previsional title : =
= =
= =
==============================================================================
Pierre Laforgue pierre@imag.imag.fr {uunet.uu.net|mcvax}!imag!pierre
------------------------------
Date: Wed, 9 Dec 87 08:33 EST
From: Dori Wells <DWELLS@G.BBN.COM>
Subject: Seminar - Practical Reasoning (BBN)
BBN Science Development Program
Language And Cognition Seminar
ISSUES IN THE STUDY OF PRACTICAL REASONING:
DESIGNING COMPUTER SUPPORT FOR
"UNSTRUCTURED WORK"
Constance Perin
Sloan School of Management
BBN Laboratories Inc.
10 Moulton Street
Large Conference Room, 2nd Floor
10:00 a.m., Wednesday, December 9, 1987
Abstract: To develop computer applications that are relevant to
nonroutine, relatively unstructured work processes requires
descriptions of them in terms of the rational, irrational, and
nonrational thought they employ. Deriving structures from the
particularities of these tasks and from the relationships among tasks
is one representational problem which needs to be addressed in
designing computer support for such tasks. Another is how to
acknowledge the influence of contexts on tasks. A third problem is
how to decrease the probability of miscommunication and increase that
of shared interpretations in complex organizations. The perspectives
of discourse analysis, semantic analysis, and figurative language
analysis seem to be appropriate to this set of questions. In this
talk, I will discuss how these types of observation and analysis might
be employed in designing research methods appropriate to knowledge
acquisition for tasks in unstructured work domains.
------------------------------
Date: Thu, 10 Dec 87 11:48 EST
From: emma@russell.stanford.edu
Subject: From CSLI Calendar, December 10, 3:10
[Excerpted from CSLI Calendar]
Representation Strategies in LILOG
Hans Uszkoreit
IBM Germany and University of Stuttgart
LILOG (Linguistic and Logic Methods for Knowledge-based Natural-Language
Understanding) is a basic research project funded by IBM Germany. It is
jointly conducted by the science and technology division of IBM Germany and
partner projects at five German universities. The long-term goal of the
project is to develop a knowledge-based text-understanding system for
German. The methods that are used in the design of the linguistic
components of LILOG share relevant features with ongoing work in the FOG
project at CSLI.
The talk will start with an overview of the objectives, organization, and
status of the project.
The underlying language for the representation of linguistic and extra-
linguistic knowledge is STUF (Stuttgart Type Unification Formalism). The
formalism has been implemented as a data type whose operations can be
utilized by all modules of the system. In addition to standard operations
such as unification, generalization, and subsumption, it provides for a
generalized version of functional application.
Newer developments of the STUF formalism include the integration of free-
arity and fixed-arity types and the introduction of knowledge domains.
Examples will be presented that demonstrate how the uniform formalism is
employed to account for the interaction of syntax and semantics.
--------------
CSLI SEMINAR
Interpretation as Abduction
Jerry R. Hobbs
December 17, 1987
2:15, Redwood Hall G-19
The goal of the TACITUS project at SRI is investigate the use of
commonsense knowledge in the interpretation of discourse. We have
recently developed a new scheme for abductive inference that yields a
dramatic simplification of our characterization of what interpretation
is. I will discuss its use in solving various local pragmatics
problems, such as the resolution of reference, metonymy, and syntactic
ambiguity problems and the interpretation of compound nominals. The
scheme also suggests an elegant way to integrate syntactic and
pragmatic processing, and this will be discussed briefly.
------------------------------
Date: Thu, 10 Dec 87 14:29 EST
From: rjb%research.att.com@RELAY.CS.NET
Subject: Intl. Conference on Principles of KR&R
CALL FOR PAPERS
FIRST INTERNATIONAL CONFERENCE ON
PRINCIPLES OF KNOWLEDGE REPRESENTATION AND REASONING
Royal York Hotel
Toronto, Ontario, CANADA
May 15-18, 1989
Sponsored by the Canadian Society for Computational Studies of Intelligence,
with support from AAAI, IJCAI, the Canadian Institute for Advanced
Research, and the Information Technology Research Centre of Ontario,
in cooperation with AISB and ACM SIGART (pending approval)
The idea of explicit representations of knowledge, manipulated by
general-purpose inference algorithms, underlies much of the work in
artificial intelligence, from natural language to expert systems. A growing
number of researchers are interested in the principles governing systems
based on this idea. This conference will bring together these researchers in
a more intimate setting than that of the general AI conferences. In
particular, all authors will be expected to appear and give presentations of
adequate length to present substantial results. Accepted papers will be
collected in a conference proceedings, to be published by Morgan Kaufmann
Publishers, Inc.
The conference will focus on principles of commonsense reasoning and
representation, as distinct from concerns of engineering and details of
implementation. Thus of direct interest are logical specifications of
reasoning behaviors, comparative analyses of competing algorithms and
theories, and analyses of the correctness and/or the computational complexity
of reasoning algorithms. Papers that attempt to move away from or refute the
knowledge-based paradigm in a principled way are also welcome, so long as
appropriate connections are made to the central body of work in the field.
Submissions are encouraged in at least the following topic areas:
Analogical Reasoning Qualitative Reasoning
Commonsense Reasoning Temporal Reasoning
Deductive Reasoning Planning
Diagnostic and Knowledge Representation Formalisms
Abductive Reasoning Theories of the Commonsense World
Evidential Reasoning Theories of Knowledge and Belief
Inductive Reasoning Belief Management and Revision
Nonmonotonic Reasoning Formal Task and Domain Specifications
REVIEW OF PAPERS
The Program Committee will review extended abstracts (not complete papers).
Each submission will be read by at least two members of the Committee and
judged on clarity, significance, and originality. An important criterion for
acceptance of a paper is that it clearly contribute to principles of
representation and reasoning that are likely to influence current and future
AI practice.
Extended abstracts should contain enough information to enable the Program
Committee to identify the principal contribution of the research and its
importance. It should also be clear from the extended abstract how the work
compares to related work in the field. References to relevant literature must
be included.
Submitted papers must never have been published. Submissions must also be
substantively different from papers currently under review and must not be
submitted elsewhere before the author notification date (December 15, 1988).
SUBMISSION OF PAPERS
Submitted abstracts must be at most eight (8) double-spaced pages. All
abstracts must be submitted on 8-1/2" x 11" paper (or alternatively, a4),
and typed in 12-point font (pica on standard typewriter). Dot matrix
printout is not acceptable.
Each submission should include the names and complete addresses of all
authors. Also, authors should indicate under the title which of the
topic ares listed above best describes their paper (if none is
appropriate, please give a set of keywords that best describe the
topic of the paper).
Abstracts must be received no later than November 1, 1988, at the address
listed immediately below. Authors will be notified of the Program Committee's
decision by December 15, 1988. Final camera-ready copies of the full papers
will be due a short time later, on February 15, 1989. Final papers will be at
most twelve (12) double-column pages in the conference proceedings.
Send five (5) copies of extended abstracts [one copy is acceptable from
countries where access to copiers is limited] to
Ron Brachman and Hector Levesque, Program Co-chairs
First International Conference on Principles of
Knowledge Representation and Reasoning
c/o AT&T Bell Laboratories
600 Mountain Avenue, Room 3C-439
Murray Hill, NJ 07974
USA
Inquiries of a general nature can be addressed to the Conference Chair:
Raymond Reiter, Conference Chair
First International Conference on Principles of
Knowledge Representation and Reasoning
c/o Department of Computer Science
University of Toronto
10 Kings College Road
Toronto, Ontario M5S 1A4
CANADA
electronic mail: reiter@ai.toronto.edu
IMPORTANT DATES
Submission deadline: November 1, 1988
Author notification date: December 15, 1988
Camera-ready copy due
to publisher: February 15, 1989
Conference: May 15-18, 1989
PROGRAM COMMITTEE
James Allen (University of Rochester)
Giuseppe Attardi (Delphi SpA, Italy)
Woody Bledsoe (MCC/University of Texas)
Alan Bundy (Edinburgh University)
Eugene Charniak (Brown University)
Veronica Dahl (Simon Fraser University)
Koichi Furukawa (ICOT)
Johan de Kleer (Xerox PARC)
Herve Gallaire (European Computer Industry Research Center, Munich)
Michael Genesereth (Stanford University)
Michael Georgeff (SRI International)
Pat Hayes (Xerox PARC)
Geoff Hinton (University of Toronto)
Bob Kowalski (Imperial College)
Vladimir Lifschitz (Stanford University)
Alan Mackworth (University of British Columbia)
Drew McDermott (Yale University)
Tom Mitchell (Carnegie-Mellon University)
Robert Moore (SRI International)
Judea Pearl (UCLA)
Stan Rosenschein (SRI International)
Stuart Shapiro (SUNY at Buffalo)
Yoav Shoham (Stanford University)
William Woods (Applied Expert Systems)
------------------------------
Date: Thu, 10 Dec 87 18:27 EST
From: Isaac Balbin <munnari!mulga.oz.au!isaac@uunet.UU.NET>
Subject: CfP - ICSC 88
Call for Papers
International Computer Science Conference '88
Hong Kong, December 19-21, 1988
Artificial Intelligence: Theory and Applications
Sponsored by
THE COMPUTER SOCIETY OF THE IEEE, HONG KONG CHAPTER
International Computer Science Conference '88 is to be the first international
conference in Hong Kong devoted to computer science. The purpose of the
conference is to bring together people from academia and industry of the East
and of the West, who are interested in problems related to computer science.
The main focus of this conference will be on the Theory and Applications of
Artificial Intelligence. Our expectation is that this conference will provide a
forum for the sharing of research advances and practical experiences among
those working in computer science.
Topics of interest include, but are not limited to:
AI Architectures Expert Systems Knowledge Engineering
Logic Programming Machine Learning Natural Languages
Neural Networks Pattern Recognition Robotics
CAD/CAM Chinese Computing Distributed Systems
Information Systems Office Automation Software Engineering
Paper Submissions
Submit four copies of the paper by June 15, 1988 to either of the Program
Co-Chairmen:
Dr. Jean-Louis Lassez Dr. Francis Y.L. Chin
Room H1-A12 Centre of Computer Studies and
IBM Thomas J. Watson Applications
Research Center University of Hong Kong
P.O. Box 218 Pokfulam Road
Yorktown Heights NY Hong Kong
10598 (For papers from Pan-Pacific region
U.S.A. only)
e-mail: JLL@ibm.com e-mail: hkucs!chin@uunet.uu.net
The first page of the paper should contain the author's name, affiliation,
address, electronic address if available, phone number, 100 word abstract, and
key words or phrases. Papers should be no longer than 5000 words (about 20
double-spaced pages). A submission letter that contains a commitment to present
the paper at the conference if accepted should accompany the paper.
Tutorials
The day after the conference will be devoted to tutorials. Proposals for
tutorials on Artificial Intelligence topics, especially advanced topics, are
welcome. Send proposals by June 15, 1988 to the Program Co-Chairmen.
Conference Timetable and Information
Papers due: June 15, 1988
Tutorial proposals due: June 15, 1988
Acceptance letters sent: September 1, 1988
Camera-ready copy due: October 1, 1988
International Program Committee:
J-P Adam (Paris T.Y. Chen (Melbourne & W.F. Clocksin
Scientific Center) HKU) (Cambridge)
A. Despain (Berkeley) J. Gallier Qingshi Gao (Academia
M. Georgeff (SRI) (Pennsylvania) Sinica)
R.C.T. Lee (National D. Hanson (Princeton) R. Hasegawa (ICOT)
Tsin Hua) M. Maher (IBM) Z. Manna (Stanford &
F. Mizoguchi (Science U. Montanari (Pisa) Weizmann)
U. of Tokyo) P.C. Poole (Melbourne) K. Mukai (ICOT)
H.N. Phien (AIT) C.K. Yuen (Singapore) D.S.L. Tung (CUHK)
Organizing Committee Local Arrangements Publicity Chairman:
Chairman: Chairman:
Mr. Wanbil Lee
Dr. K.W. Ng Dr. K.P. Chow Department of Computer
Department of Computer Centre of Computer Studies
Science Studies and Applications City Polytechnic of
The Chinese University University of Hong Kong Hong Kong
of Hong Kong Pokfulam Road Argyle Center, Kowloon
Shatin, N.T. Hong Kong Hong Kong
Hong Kong e-mail:
hkucs!icsc@uunet.uu.net
In Cooperation With:
Center for Computing Studies and Services, Hong Kong Baptist College
Centre of Computer Studies and Applications, University of Hong Kong
Department of Computer Science, The Chinese University of Hong Kong
Department of Computer Studies, City Polytechnic of Hong Kong
Department of Computing Studies, Hong Kong Polytechnic
------------------------------
Date: Tue, 15 Dec 87 14:29 EST
From: William J. Rapaport <rapaport@cs.Buffalo.EDU>
Subject: Seminar - The Psychology of Everyday Things (Suny Buffalo)
STATE UNIVERSITY OF NEW YORK AT BUFFALO
The Steering Committee of the
GRADUATE STUDIES AND RESEARCH INITIATIVE IN
COGNITIVE AND LINGUISTIC SCIENCES
PRESENTS
DONALD A. NORMAN
Institute for Cognitive Science
University of California, San Diego
THE PSYCHOLOGY OF EVERYDAY THINGS
How do we manage the tasks of everyday life? The traditional answer is
that we engage in problem solving, planning, and thought. How do we
know what to do? Again, the traditional answer is that we learn, in
part through experience, in part through instruction. I suggest that
this view is misleading. Less planning and problem solving is required
than is commonly supposed. Many tasks need never be learned: the
proper behavior is obvious from the start. The problem space for most
everyday tasks is shallow or narrow, not wide and deep as the tradi-
tional approach suggests. The minimization of the problem space occurs
because natural and contrived properties of the environment combine to
constrain the set of possible actions. The effect is as if one had put
the knowledge required to do a thing on the thing itself: the knowledge
is in the world.
I show that seven stages are relevant to the performance of an action,
including three stages for execution of an act, three for evaluation,
and a goal stage. Consideration of the rule of each stage, along with
the principles of natural mappings and natural constraints, leads to a
set of psychological principles for design. Couple these principles
with the suggestion that most real tasks are shallow or narrow, and we
start to have a psychology of everyday things and everyday actions.
The talk itself is meant to be light and enjoyable. However, there are
profound implications for the type of theory one develops for simulating
cognitive computation. There are serious implications for massively
parallel structures (what we call Parallel Distributed Processing or
connectionist approaches), for memory storage and retrieval via descrip-
tions or coarse coding, and, in general, for a central role for pattern
matching, constraint satisficing, and nonsymbolic processing mechanisms
in human cognition. But the main implications of the work are for the
design of understandable and usable objects.
Monday, February 1, 1988
4:00 P.M.
Park 280, Amherst Campus
There will also be an informal evening discussion at a place and time to
be announced. Call Bill Rapaport (Computer Science, (716) 636-3193,
3180) for further information.
------------------------------
Date: Tue, 15 Dec 87 16:22 EST
From: Peter de Jong <DEJONG%OZ.AI.MIT.EDU@XX.LCS.MIT.EDU>
Subject: Cognitive Science Calendar
[Excerpted from cog-sci-calendar digest]
Date: Monday, 14 December 1987 11:58-EST
From: Paul Resnick <pr at ht.ai.mit.edu>
Re: AI Revolving Seminar Thursday; Drew McDermott
Thursday 17, December 4:00pm Room: NE43- 8th floor Playroom
The Artificial Intelligence Lab
Revolving Seminar Series
LOGICALC: AN INTERACTIVE THEREOM PROVER
USING SKOLEMIZATION
Drew McDermott, Yale University
We have been developing an interactive first-order
theorem prover with several novel features:
Use of Skolemization and sequent logic
Maintenance of AND/OR graph of proofs
Uniquification of goals
Lambda-calculus extensions
The resulting system allows pattern-directed search
for rules to be used in pursuing goals. Although it is
already a useful tool, we plan to give it more autonomous
search capability in the future.
*******
Date: Monday, 14 December 1987 10:07-EST
From: Rosemary B. Hegg <ROSIE at XX.LCS.MIT.EDU>
Re: Jon Doyle seminar on 12/18 at 2pm
DATE: December 18, 1987
TIME: Refreshments: 1.45 pm
Lecture: 2:00PM
PLACE: NE43-8th floor playroom
ON UNIVERSAL THEORIES OF DEFAULTS
JON DOYLE
Department of Computer Science
Carnegie Mellon University
ABSTRACT
Though unifications of some of the numerous theories of
default reasoning have been found, we bolster doubts about the
existence of universal theories by viewing default reasoning from the
standpoint of decision theory as a case of rational self-government of
inference. Default rules express not only methods for deriving new
conclusions from old, but also preferences among sets of possible
conclusions. Conflicting default rules, which form the central
difficulty in the theories, represent inconsistent preferences about
conclusions. These conflicting rules arise naturally in practice,
especially in databases representing the knowledge of several experts.
We formally compare these theories of rational inference with theories
of group decision making, and develop doubts about universal theories
of the former by considering well-known negative results about the
latter.
HOST: Prof. Peter Szolovits
------------------------------
End of NL-KR Digest
*******************