Appel: FOUR ANLP/NAACL-2000 Workshop
Philippe Blache
pb at lpl.univ-aix.fr
Thu Jan 13 12:27:53 UTC 2000
From: Priscilla Rasmussen <rasmusse at cs.rutgers.edu>
The following four Calls for Papers for workshops associated with the
ACL-sponsored ANLP/NAACL-2000 Conference are included below, separated
by dash lines:
1)Workshop on Conversational Systems
May 4, 2000, following ANLP/NAACL 2000
2)EMBEDDED MACHINE TRANSLATION SYSTEMS WORKSHOP II
Thursday, May 4, 2000
3)Workshop on Applied Interlinguas: Practical Applications of Interlingual
Approaches to NLP
Sunday, April 30, 2000
4)Workshop on Reading Comprehension Tests as Evaluation for
Computer-Based Language Understanding Systems
Thursday, May 4th, 2000, Seattle, Washington, USA
-----------------------------------------------------------------------
Call for Papers
Workshop on Conversational Systems
May 4, 2000, following ANLP/NAACL 2000
The purpose of this workshop is to focus the discourse and dialogue
community on best practices as well as theory of conversational
systems, both speech based and text based. The workshop will also
bring together creators of working conversational systems to discuss their
efforts, both successes and limitations.
In this workshop we encourage papers on either theoretical or applied
research with a focus on results in working systems. We also welcome
papers on working systems that provide a critical appraisal of their
capabilities as well as their limitations; we encourage such papers to
provide the criteria of critique that the authors feel are most
relevant to their work. This workshop will consider in particular:
- How can systems be designed so that it is easier to build
applications in new domains?
- What significant features of dialogue are beyond current working
systems? What proposals show the most promise for capturing these
features?
- What knowledge does a system need to represent about a domain, tasks
and discourse to support intelligent conversational interaction?
- What can be learned from data and what should be learned from data?
Can robust systems be built for domains where there is not a large amount
of data available?
- What is the role of natural language generation in conversational systems?
- What aspects of discourse prosody are now feasible in conversational
systems?
- What aspects of nonverbal behavior are now feasible -- and worthwhile
implementing -- in conversational systems?
- How can the real-world performance of conversational systems be measured
and anticipated? How can the performance of different systems be compared?
In addition to the presentation of papers and the discussions that
will result from them, we plan demonstration sessions and a panel
session. The demonstration sessions will be open to anyone who wishes
to bring their conversational systems for demonstration to other
members of the workshop. Presenters are asked to submit a paper that
is specifically directed at a demonstration of their current systems.
These papers should cover the following topics as well as others the
presenters think are relevant:
-a short system description,
-an example dialogue or dialogues, as space permits,
-discussion of the most important contribution of the work,
-discussion of the most significant limitation of the work.
These papers will be included in the workshop proceedings.
In the panel session we plan to bring together a set of experts to
compare various approaches (including frame-based, finite-state,
plan-based and statistical and logical reasoning-based) to dialogue in
working conversational systems.
A website which will provide additional information on the workshop as
it becomes available is located at:
http://www.cs.umd.edu/users/traum/ConvSys/.
I. IMPORTANT DATES
Paper submission deadline: February 4, 2000
Notification of acceptance for papers: March 1, 2000
Camera ready papers due: March 13, 2000
Workshop date: May 4, 2000
II. FORMAT FOR SUBMISSION
Submissions must use the ACL latex style or ACL Microsoft Word style,
both of which can be found at
http://www.gte.com/AboutGTE/gto/anlp-naacl2000/cfp_submission.html. Paper
submissions should consist of a full paper of 8 pages (including
references). Please send submission questions to Alex
Rudnicky,air at cs.cmu.edu, before, not after, January 31, 2000.
Submission Procedure:
Electronic submission only: send the pdf (preferred), postscript or
MS Word form of your submission to: Alex Rudnicky, air at cs.cmu.edu. The
Subject line should be "ANLP-NAACL2000 WORKSHOP PAPER
SUBMISSION". Because reviewing is blind, no author information is
included as part of the paper. An identification page must be sent in
a separate email with the subject line: "ANLP-NAACL2000 WORKSHOP ID
PAGE" and must include title, all authors, theme area, keywords, word
count, and an abstract of no more than 5 lines. Late submissions will
not be accepted. Notification of receipt will be e-mailed to the first
author shortly after receipt.
The Organizing Committee for this workshop includes:
Candy Sidner, Lotus (Chair)
James Allen, Univ. of Rochester
Harald Aust, Philips Corp.
Phil Cohen, Oregon Graduate Institute
Justine Cassell, Media Lab, MIT
Laila Dybkjaer, University of Southern Denmark
X.D. Huang, Microsoft
Masato Ishizaki, Japan Adv. Institute of Science and Technology
Candace Kamm, AT&T
Lin-Shan Lee, Taiwan University
Susann Luperfoy, Akamai Technologies
Patti Price, SRI International
Owen Rambow, AT&T
Norbert Reithinger, DFKI Saarbruecken
Alex Rudnicky, Carnegie Mellon University
Stephanie Seneff, MIT
Dave Stallard, BBN/GTE
David Traum, University of Maryland
Marilyn Walker, AT&T
Wayne Ward, Univ of Colorado, Boulder
------------------------------------------------------------------------
****************** CALL FOR PAPERS ***************************
EMBEDDED MACHINE TRANSLATION SYSTEMS
WORKSHOP II
held in conjunction with NAACL/ANLP2000
Thursday, May 4, 2000
Seattle, Washington, USA
Embedded MT Systems homepage for this workshop
http://lamp.cfar.umd.edu/Embedded_MT_Systems
WHAT IS AN "EMBEDDED MACHINE TRANSLATION (MT) SYSTEM"?
An "embedded MT system" is a computational system with one
or more MT engines among its components. These systems
accept multilingual, multimodal inputs and create various outputs
that enable the users to access the original information in their
own language. An MT component embedded in an end-to-end system
allows users to perform their specific tasks on foreign language
input that they previously only had been able to perform in
their native language. To date, these tasks have included
summarization, content extraction, filtering and document
retrieval.
BACKGROUND
The first workshop on Embedded MT Systems was held in conjunction
with the biennial meeting of the Association for Machine Translation
in the Americas (AMTA), in October, 1998, in Langhorne, PA. The
Embedded MT Systems Workshop II is a response to the growing
community commitment to translingual information research,
e.g., the DARPA TIDES initiative. By holding the workshop at
the combined NAACL and ANLP conferences this year, there will be
an opportunity for a multi-disciplinary mix of researchers and
to attend, contribute and benefit from the workshop.
WORKSHOP DESCRIPTION
The strengths and weaknesses of machine translation engines
have become better understood and accepted. There has been a marked
increase in the development of a range of computer systems containing
an MT component. This workshop will focus on the system designs, the
associated information access tasks of such end-to-end systems,
and the measures of system effectiveness.
Of particular interest are systems that accept one or another of
various types of input including hard-copy pages, online text
files, and speech (natural or transcribed). These inputs
present real-world, noisy data that challenge MT engine capabilities.
We would like to know the degradation in performance that
these challenges present and the compensation strategies that
system developers have tested or used. We also seek
submissions describing possible channel-specific feedback
processes from other system components that help correct the
noisy input.
Papers describing multiple MT engines and algorithms for selecting
among their outputs are encouraged. It would be interesting
to hear how these complex MT components have been integrated into
specific applications. For example, do certain MT engines produce
results better suited for summarization, retrieval, or online
foreign language tutoring?
The field of MT evaluation currently lacks an adequate methodology.
There are no widely used standards and few statisticians have been
called upon to assess the metrics that have been proposed. We will
look for submissions that include measures for the individual system
components and end-to-end system evaluation. Also of interest are
measures that evaluate user performance on specific tasks.
We expect that the range of papers from both the first and this
second workshop will provide sufficient material for us to pursue
a special journal issue dedicated to Embedded MT Systems.
IMPORTANT DATES
Intent to submit: Friday, Feb. 11, 2000
Paper submission deadline: Wednesday, Feb. 16, 2000
Notification of acceptance of papers: Friday, March 3, 2000
Camera-ready papers due: Monday, March 13, 2000
SUBMISSION PROCEDURE
Electronic submission of Intent to Submit should have the
following subject line:
"NAACL-ANLP2000 WORKSHOP - Intent to submit"
Body of message should include Identification Page information:
- title of submission
- names of all authors
- primary author name and email address, phone and fax
- presentation type preference
(select one or more per system: demo, poster, or paper)
- keywords
Authors may submit short papers, full-length papers, poster
presentations and/or demos.
For electronic submission, include the Identification
Page Information (see above) as a separate page from the paper
itself. Reviewing will be blind. No author information should be
included with the main body of the paper. Full paper submissions
may be up to 5000 words in length, including references.
Submissions for poster presentations and short papers may be up
to 2000 words in length, including references.
Demo presentations are encouraged in conjunction with papers
or posters. For demo-only presentations, submissions up to
two pages long should describe the system design and
capabilities with respect to (ii) above:
an end-to-end process flow covering the system
input, any pre-MT processing, the MT component itself, any
post-MT processing, and the system output.
FORMAT FOR SUBMISSION
Submissions must use the ACL latex style or Microsoft Word style.
Both are available from the ANLP-NAACL2000 Conference web page:
http://www.gte.com/AboutGTE/gto/anlp-naacl2000/
Please send submissions and questions to: voss at arl.mil
Notification of receipt will be sent to the primary author.
ORGANIZING COMMITTEE
Carol Van Ess-Dykema, US Dept. of Defense
Clare R. Voss, US Army Research Lab
Florence Reeder, MITRE Corp.
PROGRAM COMMITTEE
Gary Coen, Boeing Phantom Works
Bob Frederking, LTI, Carnegie Mellon Univ.
Laurie Gerber, SYSTRAN
Inge Gorm Hansen and Henrik Selsoe Sorenson, Copenhagen Business School
Lori Levin, LTI, Carnegie Mellon Univ.
Bill Ogden, CRL, NMSU
Kathi Taylor, Georgetown U.
-------------------------------------------------------------------------
CALL FOR PAPERS
Workshop on Applied Interlinguas:
Practical Applications of Interlingual Approaches to NLP
(pre-conference workshop in conjunction with ANLP-NAACL2000)
Sunday, April 30, 2000
Seattle, Washington, USA
The organizing committee wishes to invite submissions to the Workshop
on the Practical Applications of Interlingual Approaches to NLP which
will be held on Sunday, April 30, 2000, in conjunction with the
ANLP-NAACL2000.
Interlingual approaches to NLP have been developed within the field of
Machine Translation. The central goal is to analyze natural language
expressions in terms of a representation language that will capture
those aspects of a communication which permit the generation of an
equivalent expression in some other language (that is, a
representation of the communicative intent of the utterance). An
interlingual representation of some utterance should ideally represent
what was said by whom and to whom and relevant information about
where, when, why and how it was said. The representations are usually
very rich and extremely knowledge intensive. Many aspects of such
representation systems are unknown or underdeveloped.
Often, though not invariably, the lexicon of an IL representation will
be drawn from the names of nodes in an Ontology, representing classes,
events, or concepts. The syntax of the IL prescribes how these nodes
are combined into an utterance representation. An IL-based approach to
Machine Translation then specifies how a source language sentence can
be analyzed into an IL representation and how this representation can
then generate a natural language output.
This workshop will focus on these latter two aspects of the IL
approach: the syntax of the IL and the techniques used to analyze and
generate natural language. The uses of an Ontology outside of
Knowledge-based Machine Translated is a related, but slightly
different subject.
To date, such approaches have been essentially theoretical (although a
number of limited applications exist). One of the criticisms of such
approaches is that they are impractical -- requiring too much
hand-coding or too deep a knowledge-representation to be
useful. However, several examples of IL specifications are available.
For example, there is
the Text Meaning Representation of the Mikrokosmos Knowledge-based
Machine Translation system at the Computing Research Laboratory
(http://crl.nmsu.edu/Research/Projects/mikro/index.html).
the IL used in ISI's GAZELLE MT project
(http://www.isi.edu/natural-language/mt/interlingua.html)
IL representations of a Spanish text produced by the KANT system at
the Language Technologies Institute
(http://www.lti.cs.cmu.edu/IRW/)
(http://www.lti.cs.cmu.edu/Research/Kant)
IL representation developed for a speech-to-speech system dealing with
travel planning by the Consortium for Speech Translation Advanced
Research (C-STAR)
(http://www.c-star.org)
Interlingual approaches offer powerful semantics-based and
pragmatics-based solutions to any number of NLP problems
(disambiguation, reference resolution, interpretation of figurative
speech to name a few). This workshop will focus on methods for making
interlingual approaches tractable within specific, well-defined tasks
not only for machine translation but for a range of NLP applications.
The goal of the workshop is to stimulate interest in more cognitive
research in NLP while focusing such work on near term, practical
applications. Papers are invited on:
- methods for developing (or extending) underlying knowledge sources,
- techniques for processing in the face of knowledge-poor sources or
gaps in knowledge sources,
- interlingual approaches to particular NLP tasks (reference
resolution, disambiguation, interpretation of ellipsis, etc.),
- interlingual approaches to different NLP applications
(MT--including speech-to-speech translation, Information
Extraction, Summarization, NL generation, Intelligent Tutoring
Systems, etc.).
Since there is limited work on the application of IL approaches to NLP
currently, concept design papers are encouraged. Preference will be
given to actual research projects focusing on actual processing
problems and exploiting extant sources, but any contribution should
clearly focus on one of the topics above.
The workshop will consist of 6 30-minute presentations, each followed
by a half-hour discussion, beginning with two informal 6-minute
critical responses from reviewers followed by a short rebuttal by the
author and open discussion. Ideally, the critical responses will also
be available by the March 1 acceptance date, but in no case later
than March 31. All critiques and rebuttals received by March 13 will
be included in the proceedings. The remainder will be made available
at the workshop itself.
The Journal of Machine Translation will consider the results of the
workshop for publication in a special issue in 2001. In addition, the
contributions will be published as an NAACL workshop proceedings.
The program committee (and initial discussants) includes:
Bonnie Dorr UMIACS-UMd
David Farwell CRL-NMSU
Stephen Helmreich CRL-NMSU
Eduard Hovy ISI-USC
Kevin Knight ISI-USC
Lori Levin LTI-CMU
Teruko Mitamura LTI-CMU
Keith Miller MITRE
Sergei Nirenburg CRL-NMSU
Mari Olsen Microsoft
Boyan Onyshkeyvch DOD
Florence Reeder MITRE
Harry Somers UMIST
Yorick Wilks USheffield
The workshop is sponsored in part by the Special Interest Group on
Interlinguas of the Association for Machine Translation in the
Americas. For further information about this series of workshops see:
http://crl.nmsu.edu/Events/FWOI/index.html.
Dates
Submission of papers: February 4, 2000
Notification of acceptance: March 1, 2000
Submission of Final Copies: March 13, 2000
Critiques of Accepted Papers: March 31, 2000
Author's Rebuttals: April 21, 2000
Workshop: April 30, 2000
The dates related to the preparation of a special issue of the Journal
of Machine Translation will be made public at the workshop.
Paper Requirements
Submissions must use the ACL latex style or Microsoft Word style (both
available from the ANLP-NAACL2000 Conference web page --
http://www.gte.com/AboutGTE/gto/anlp-naacl2000/). Paper submissions
should consist of a full paper (5000 words or less, including
references). Please send papers and submission questions to
shelmrei at crl.nmsu.edu.
Cost
There will be a registration fee of $50.00 per person. A banquet for
the participants and guests will be organized separately for Sunday
evening.
-----------------------------------------------------------------------
Call for Papers
Workshop on Reading Comprehension Tests as Evaluation for
Computer-Based Language Understanding Systems
Thursday, May 4th, 2000, Seattle, Washington, USA
(post-conference workshop in conjunction with ANLP-NAACL2000)
Reading Comprehension tests, such as the one below, are designed to help
evaluate a reader's understanding of a text passage.
How Maple Syrup is Made
Maple syrup comes from sugar maple trees. At one time, maple
syrup was used to make sugar. This is why the tree is called a
"sugar" maple tree.
Sugar maple trees make sap. Farmers collect the sap. The best
time to collect sap is in February and March. The nights must be
cold and the days warm.
The farmer drills a few small holes in each tree. He puts a
spout in each hole. Then he hangs a bucket on the end of each
spout. The bucket has a cover to keep rain and snow out. The sap
drips into the bucket. About 10 gallons of sap come from each
hole.
1. Who collects maple sap? (Farmers)
2. What does the farmer hang from a spout? (A bucket)
3. When is sap collected? (February and March)
4. Where does the maple sap come from? (Sugar maple trees)
5. Why is the bucket covered? (to keep rain and snow out)
Such tests exist in many languages, have human performance benchmarks
associated with them, and come in a variety of types (short-answer,
multiple choice) and levels of difficulty. In addition, they are
generally written to make each story and set of questions
self-contained, in order to require as little outside knowledge as
possible to answer the questions.
The focus of the proposed workshop will be to explore the following
questions:
- Can such exams be used to evaluate computer-based language
understanding effectively and efficiently?
- Would they provide an impetus and test bed for interesting and
useful research?
- Are they too hard for current technology?
- Or are they too easy, such that simple hacks can score high,
although there is clearly no understanding involved?
The most direct method of exploring these questions is to choose a set
of tests and build a system that takes these tests. Some preliminary
results indicate that such tests are tractable, but not trivial and
that linguistic processing is helpful (Hirschman, et al. ACL-99). A
test set, evaluation routines, prototype system, and documentation are
available upon request to light at mitre.org.
We hope that a number of submissions will present results based on
actual reading comprehension systems. In addition, we encourage
submissions that report on other kinds of tests or similar tests in
other languages, or that address our list of questions by other
means. Note that submissions are encouraged that describe work in
progress with preliminary empirical results.
Invited speaker:
Karen Kukich (Educational Testing Service)
"NLP Tools for Analyzing TOEFL Reading Comprehension Passages and
Items"
Format for Submission
Authors are asked to submit previously unpublished papers only; a
workshop proceedings will be published. Our target submission length
is 2000 words but both shorter and longer submissions will also be
considered. Electronic submission of postscript will be accepted.
Hard copy submissions should include 4 copies of the paper. Since the
papers will be reviewed anonymously, please do not place the author
name on the paper. Instead include a separate title page with title,
abstract, author, and e-mail address. Unless requested otherwise,
notification of acceptance will be sent electronically to the first
author. Parallel submission is unproblematic; however if your paper
is accepted to this workshop and you decide to present it here, we
will ask you to withdraw it from any other events.
Important Dates
Deadline for submission: February 11th, 2000
Notification of authors: March 1st, 2000
Final versions due: March 10th, 2000
Address for Submission and Further Information
Marc Light
The MITRE Corporation
202 Burlington Rd.
M/S K329
Bedford, MA 01730
USA
Phone: 1-781-271-5579
light at mitre.org
(The mailing list, read-comp at linus.mitre.org, has been set up to
discuss reading comprehension tests as evaluation for computer-based
language understanding systems. It is open subscription and
unmoderated. To subscribe, send email to majordomo at linus.mitre.org
with 'subscribe read-comp' in the body.)
Program Committee:
Eric Brill
Eugene Charniak
Mary Harper
Marc Light (chair)
Ellen Riloff
Ellen Voorhees
More information about the Ln
mailing list