Copy Link
Add to Bookmark
Report
NL-KR Digest Volume 09 No. 19
NL-KR Digest (Wed Apr 22 09:36:59 1992) Volume 9 No. 19
Today's Topics:
Announcement: proceedings available (Machine Learning of NL)
Announcement: Philosophy and AI Summer Schools (Italy)
Talk: Anthony Maida at Tulane University
Talk: Stuart Shieber at Brandeis University
CFP: 9th IEEE Conference on AI for Applications
Submissions: nl-kr@cs.rpi.edu
Requests, policy: nl-kr-request@cs.rpi.edu
Back issues are available from host archive.cssage.rpi.edu [128.113.53.18] in
the files nl-kr/Vxx/Nyy (ie nl-kr/V01/N01 for V1#1), mail requests will
not be promptly satisfied. If you can't reach `cs.rpi.edu' you may want
to use `turing.cs.rpi.edu' instead.
BITNET subscribers: we now have a LISTSERVer for nl-kr.
You may send submissions to NL-KR@RPIECS
and any listserv-style administrative requests to LISTSERV@RPIECS.
[ As a few have noticed, I have changed the summary portion above to be
a little more informative, in hopes of being able to easily create an
index. I anyone has any general ideas on how this Digest could be
made more useful I would appreciate it - CW ]
-----------------------------------------------------------------
To: nl-kr@cs.rpi.edu
Date: Tue, 21 Apr 92 17:11:02 +0200
From: walter@kub.nl (Walter Daelemans)
Subject: Announcement: proceedings available (Machine Learning of NL)
ANNOUNCEMENT
For those who missed the workshop, available soon:
BACKGROUND AND EXPERIMENTS IN MACHINE LEARNING OF NATURAL LANGUAGE
(Proceedings First International SHOE Workshop)
Walter Daelemans and David Powers (eds.)
ITK
Institute for Language Technology and AI
Tilburg University
ITK Proceedings 92/1, 282 pages, ISBN: 90-74029-02-7, May 1992.
To obtain a copy, send request (preferably by email) to Walter
Daelemans, ITK, P.O.Box 90153, 5000 LE Tilburg, The Netherlands, Tel:
+31 13 663070, Fax: +31 13 663110, (Email: walter@kub.nl).
SHOE is a Europe-wide informal consortium focussing on the Extraction
of Hierarchical Structure for machine learning of the lower levels of
natural language.
o - E xtraction
/
o - O f
/
o - H ierarchical
/
o - S tructure
for Machine Learning of Natural Language
CONTENTS
PART I: BACKGROUND
The Logic of Learning: An introduction to concept learning and
Inductive Logic Programming.
(Peter Flach)
Topics in `Natural' Natural Language Acquisition: An Introduction.
(Steven Gillis)
The Genetic Algorithm.
(Bernard Manderick)
Probabilistic Performance Analysis of Heuristic Search using
Parallel Hash Tables.
(Giovanni Manzini and Marco Somalvico)
SHOE: The Extraction of Hierarchical Structure for Machine
Learning of Natural Language: Project Summary.
(David Powers and Walter Daelemans)
SHOE related project research in SATUS, SCIPS, and LLAMA.
(Burghard Rieger)
Aims and Perspectives of Quantitative and Synergetic Linguistics.
(Juergen Schrepp)
Language Learning, Cognition and Computing: A Summary.
(Gerry Wolff)
PART II: IMPLEMENTATION
Linguistic Pattern Matching Capabilities of Connectionist Networks.
(Antal van den Bosch and Walter Daelemans)
Analogical Modelling of Main Stress Assignment in
Dutch Simplex Words.
(Gert Durieux)
Learning Vowel Harmony.
(Mark Ellison)
Bootstrapping Syntactic Categories Using Statistical Methods.
(Steven Finch and Nick Chater)
Inductive Learning of Reversible Grammars.
(Sven Naumann and Jurgen Schrepp)
On the Significance of Closed Classes and Boundary Conditions:
Experiments in Lexical and Syntactic Learning.
(David Powers)
Filtering the Pravda with a Self-Organizing Neural Net.
(Jan Scholtes)
Resolving Linguistic Ambiguities with a Neural Data-Oriented
Parsing System.
(Jan Scholtes)
------------------------------
To: nl-kr@cs.rpi.edu
From: perini@.UUCP (Anna Perini)
Newsgroups: comp.ai.nlang-know-rep
Subject: Announcement: Philosophy and AI Summer Schools (Italy)
Keywords: Philosophy of language, Temporal Reasoning
Date: 22 Apr 92 07:57:39 GMT
INTERNATIONAL SUMMER SCHOOLS
IN PHILOSOPHY AND ARTIFICIAL INTELLIGENCE
BOLZANO/BOZEN - ITALY
==============================================================================
July 6-10, 1992
Philosophy of Language
Since the beginning of the century, philosophy of language has played a
major role in philosophical analysis. The two most important
philosophical movements - analytic philosophy and phenomenology - share
a common interest in questions concerning language, meaning, reference,
context and situation. In order to analyse such concepts, logic and
formal semantics are some of the tools mainly used form the analytic
side. On the other side, analysis of the genesis of concepts and of the
different layers that compose them are fundamental aspects of the
phenomenological perspective. The aim of this school is to examine, in
connection with the analysis of language, both the affinities and the
incompatible features, if any, of these two philosophical movements.
ANDREA BONOMI (Milano)
The Role of Context in Formal Semantics
JOACHIM ESCHBACH (Bochum)
B\"uhler's Philosophy of Language
DIEGO MARCONI (Torino)
Reference and Semantic Competence
EDGAR MORSCHER (Salzburg)
States of Affairs and Situations: From Marty to Situations Semantics
Scientific Programme coordinated by Liliana Albertazzi
Centro Studi per la Filosofia Mitteleuropea - Trento
- ---------------------------------------------------------------------------
July 13-17, 1992
Temporal Reasoning
The school will be devoted to advanced aspects of temporal reasoning, a
theme that is receiving increased attention in the Artificial
Intelligence community. Many practical problems require an artificial
system to be able to reason about events, duration, actions collocated
in time and so on. On the other hand the problems related to our
understandig of time are typical cases where Artificial Intelligence
connects with the world of formal philosophy (e.g. first order temporal
logics, non-monotonic logics, modal logics). The present school is
intended to give an indepth view of several different approaches to
temporal reasoning, presented by some of the most respected researchers
in this area, and will give an opportunity for AI scholars and students
to discuss with them.
JAMES ALLEN (Rochester)
JOHAN VAN BENTHEM (Amsterdam)
DREW McDERMOTT (Yale)
ERIK SANDEWALL (Link\"oping)
Scientific Programme coordinated by Oliviero Stock
Istituto per la Ricerca Scientifica e Tecnologica, Trento
Sponsored by
ECCAI - European Coordinating Committee for Artificial Intelligence
AIIA - Associazione Italiana di Intelligenza Artificiale
Notes for applicants:
People wishing to participate should send a short statement of
interest (about 1 page) to the address below.
Applications should arrive by May 31th.
Notification of acceptance will be given by June 10th.
==============================================================================
General information:
1. Attendance to each School will be limited to about 30 participants.
No registration costs.
A hotel list will be sent upon notification of acceptance.
2. Each speaker will give 5 lectures, with ample time for
discussion.
3. All lectures will be in English.
4. A small number of boursaries are available to qualified
students to meet the costs of participation. Applicants should
submit a curriculum vitae.
If you wish to take part in either of the two summer schools
please write to
Roberto Poli
Dipartimento di Teoria storia e ricerca sociale,
via Verdi 26,
38100 Trento
Italy.
tel. (Italy) -461-881403
fax. (Italy -461-881440
e-mail: POLI@ITNCISTI.BITNET.
------------------------------
To: nl-kr@cs.rpi.edu
From: Robert Goldman <rpg@cs.tulane.edu>
Date: Tue, 21 Apr 92 21:20:59 CDT
Subject: Talk: Anthony Maida at Tulane University
Reply-To: rpg@cs.tulane.edu (Robert Goldman)
Modeling the Belief State of Observed Agents
Anthony S. Maida
Center for Advanced Computer Studies
University of Southwestern Louisiana
Presented by the Computer Science Department
Thursday, April 23rd, 2:00 P.M. Room 241, Boggs Building
This talk surveys the author's research on modeling the belief state
of observed agents. An agent is any kind of knowledge-based system
that interacts with the world. As background, we will discuss
ontologies for belief representation focusing on the syntactic
ontology. We will also review issues associated with the interactions
of belief, equality, and quantification, highlighting unsolved
problems. We shall call these beq problems.
We will link beq problems to the author's research on causal
connections. Causal connections are concerned with the relationship
between the data structures or symbolic structures in an agent's
knowledge base and the external objects these structures represent.
We show that beq problems occur when an observed agent has existential
misconceptions -- that is, misconceptions about the existence of
objects. Existential misconceptions can be detected, classified, and
corrected by modeling the agent's causal connections. Because of the
relationship between existential misconceptions and beq problems, the
latter can be solved by systematically modeling the agent's causal
connections.
This talk will also discuss a hypothesis that explains how one agent
may acquire the ability to model the belief state of a second agent.
Specifically, we propose that an agent who can introspect and reason
by analogy should be able to acquire the ability to model the belief
state of other agents. The long-term significance of this is that it
may not be necessary to specifically design computer programs to
reason about the belief state of other agents. If they can introspect
and reason by analogy, then they should be able to acquire this
ability on their own.
------------------------------
To: nl-kr@cs.rpi.edu
From: Dan Cerys <cerys@BBN.COM>
Subject: Talk: Stuart Shieber at Brandeis University
Date: Wed, 22 Apr 92 9:32:48 EDT
**** BRANDEIS COMPUTER SCIENCE DEPARTMENT COLLOQUIUM SERIES ****
TALK BY STUART SHIEBER
LOCATION: Brandeis University
Waltham, Massachusetts 02254
Computer Science Seminar
Room Ford 229
TIME:
Thursday, April 23, 1992 2:00 p.m.
(refreshments at 1:30 p.m.)
ABSTRACT:
New Varieties of Tree-Adjoining Grammars
Dr. Stuart M. Shieber
Harvard University
Like any natural class of formal languages, the tree-adjoining
languages can be characterized from many different perspectives:
grammatical, language-theoretic, automato-theoretic, linguistic. It
is in part the position of the class of tree-adjoining languages at
the juncture of these various characterizations that accounts for the
attractiveness of tree-adjoining grammars (TAG) as a natural language
processing tool. In this talk, we will present two new variants of
TAGs that make intrinsic use of certain of these properties to allow
for the application of TAGs to problems beyond the statement of
grammatical regularities. Synchronous tree-adjoining grammars extend
the use of TAGs to transduction relationships, and so can be used for
describing semantic interpretation or language translation.
Stochastic tree-adjoining grammars generalize TAGs to allow for
statements of a probabilistic nature, and can be used as a basis for
statistical modeling of natural-language. The talk will be
self-contained, and will assume no prior knowledge of tree-adjoining
grammars.
Host: James Pustejovsky
Information: jamesp@cs.brandeis.edu
phone: (617)-736-2709
fax: (617)-736-2741
------------------------------
To: nl-kr@cs.rpi.edu
Date: Mon, 20 Apr 92 11:33:05 EDT
From: finin@algol.cs.umbc.edu (Timothy Finin)
Subject: CFP: 9th IEEE Conference on AI for Applications
PRELIMINARY CALL FOR PARTICIPATION
CAIA-93
The Ninth IEEE Conference on
Artificial Intelligence for Applications
Disneyworld Hilton -- Orlando, Florida
March 1-5, 1993
The conference is devoted to advancing the application of artificial
intelligence techniques to real world problems. Two kinds of papers are
appropriate: case studies of AI applications that solve significant
problems and stimulate the development of useful techniques, and papers on
novel AI techniques and principles that enable more ambitious real-world
applications. This conference provides a forum for such synergy between
applications and AI techniques. Emphasis at this year's conference will
be on new AI paradigms that can or have had an impact on applications.
Papers describing significant unpublished results are solicited in two
areas:
* Applications Papers. Contributions stemming from the general areas of
industry, science and engineering, business, government, law, etc.
Application papers must: (1) Justify the use of the AI technique, based on
the problem definition and an analysis of the application's requirements;
(2) Explain how AI technology was used to solve a significant problem; (3)
Describe the status of the implementation; (4) Evaluate the effectiveness
of the application and the technique used.
Short papers describing systems in use (up to 1000 words, extended
abstract) will also be accepted for presentation in these application
tracks.
* Enabling Technology Papers. Contributions focusing on techniques and
principles that facilitate the development of practical AI applications
that can be scaled to handle increasing problem complexity. Topics
include, but are not limited to: knowledge representation, reasoning,
search, knowledge acquisition, learning, constraint programming, planning,
(including artificial neural nets, genetic algorithms, nearest neighbors,
etc.), validation and verification, project management, natural language
processing, speech, robotics, data discovery and database mining,
multimedia and virtual reality applications, intelligent interfaces,
integration, problem-solving architectures, programming environments,
exploitation of parallelism, and general tools.
Papers should be limited to 5000 words. Papers significantly longer that
this will not be reviewed. The first page of the paper must contain the
following information (where applicable) in the order shown:
* Title.
* Author's name and affiliation (specify student status).
* Contact information (name, postal address, phone and email address).
* Abstract: A 200 word abstract that includes a clear statement
describing the paper's original contributions and what new lesson is
imparted.
* AI topic: One or more terms describing the relevant AI areas, e.g.
knowledge acquisition, explanation, diagnosis, etc.
* Domain area: One or more terms describing the problem domain area, e.g.
mechanical design, factory scheduling, education, medicine, etc.
* Language/Tool: Underlying programming languages, systems and tools used.
* Status: Development and deployment status, as appropriate.
* Effort: Person-years of effort put into developing the particular aspect
of the project being described.
* Impact: A 20 word description of estimated or measured (specify) benefit
of the application developed.
Papers will be accepted in two forms: long papers and short papers. Papers
accepted for publication will be allotted seven pages (long papers) or
four papers (short papers) in the conference proceedings. The best papers
accepted will be considered for a special issue of IEEE EXPERT to appear
late in 1993. Awards will be presented to best paper and best student
paper at the conference.
In addition to papers, we will be accepting the following types of
submissions:
* Proposals for Panel Discussions. Provide a brief description of the
topic (1000 words or less). Indicate the membership of the panel and
interest in organizing/moderating the discussion.
* Proposals for Tutorial Presentations. Proposals for the three hour
tutorials of both an introductory and advanced nature are requested.
Topics should relate to the management and technical development of useful
artificial intelligence applications. Tutorials which analyze classes of
applications in depth or examine techniques appropriate for a particular
class of applications are of particular interest.
Each tutorial should include the following:
* Detailed topic outline and extended abstract (about 3 pages).
* Intended audience and assumed background knowledge.
* Half-page synopsis of focus, topics, and benefits to audience.
* Full professional vita (including lecture/tutorial experience and a
one-paragraph summary.
* Proposals for Workshops. Proposals are sought for one day workshops to
be held in conjunction with the conference. These workshops can focus on a
specific application domain (e.g. aerospace applications) or on a
technical subarea (e.g. intelligent real time problem solving). Workshop
organization and attendance will be governed by the organizers. Submit
proposals to the Workshop Chair.
Important Dates
* August 31, 1992: Four copies of papers, and three copies of all the
proposals are due. Submissions not received by that date will be returned
unopened. Electronically transmitted materials will not be accepted.
* October 30, 1992: Author notifications mailed.
* December 14, 1992: Accepted papers due to IEEE. Accepted tutorial
notes due to Tutorial Chair.
* March 1-2, 1993: Conference tutorial program.
* March 3-5, 1993: Conference technical program.
Submit Papers and Panels to:
David Waltz
Thinking Machines Corporation
245 First Street
Cambridge, MA 02142-1214
Phone: 617-234-2050
Fax: 617-234-4444
Email: waltz@think.com
Submit Tutorial or Workshop Proposals to:
Peter Selfridge
AT&T Bell Labs
2B-425
600 Mountain Avenue
P.O. Box 636
Murray Hill, NJ 07974
Phone: 908-582-6801
Email: pgs@research.att.com
For registration and additional conference information, contact:
CAIA-93
IEEE Computer Society
1730 Massachusetts Avenue, NW
Washington, DC 20036-1903
Phone: 202-371-1013
General Chair:
Jan Aikins, Aion Corporation
Program Chair:
David Waltz,
Thinking Machines Corporation,
and Brandeis University
Tutorial/Workshop Co-Chairs:
Donald McKay, Paramax Systems Corporation
Peter Selfridge, AT&T Bell Laboratories
Publicity Chair:
Curt Hall, Intelligent Software Strategies
Local Arrangements Chair:
Doug Dankel, University of Florida
Program Committee:
Chid Apte, IBM
Ralph Barletta, Cognitive Systems
Madeleine Bates, Bolt, Beranek & Newman
Vasant Dhar, NYU
Bob Engelmore, Stanford University
Tim Finin, University of Maryland Baltimore County
Brad Goodman, Mitre Corporation
Mark Goodman, Cognitive Systems, Inc. and Brandeis University
Paul Haley, The Haley Enterprise
Larry Harris, AI Corporation
Phil Hayes, Carnegie Group, Inc.
Se June Hong, IBM
Anthony Maddox, Northeastern University
Patti Maes, MIT Media Lab
Rob Milne, Intelligent Applications (Scotland)
Bernadette Kowalski Minton, Aion Corporation
Steve Minton, NASA Ames Research Center
Fumio Mizoguchi, Tokyo Science University
Peter Norvig, Sun Microsystems Laboratories
Dan O'Leary, University of Southern California
James Pustejovsky, Brandeis University
Elaine Rich, MCC
Edwina Rissland, University of Massachusetts
Oliver Selfridge, GTE Laboratories
Howard E. Schrobe, Symbolics, Inc.
Candy Sidner, DEC Cambridge Research Laboratory
Craig Stanfill, Thinking Machines Corporation
Oliver Vadas, Pulp and Paper Research Insititute of Canada
Partial list of invited speakers:
Patrick Winston, MIT and Ascent Technology
"Learning and Database Mining"
Wendy Lehnert, University of Massachusetts
"What We've Learned from the DARPA Natural Language Initiative"
------------------------------
End of NL-KR Digest
*******************
