LINGUIST List 11.361

Fri Feb 18 2000

Calls: Natural Language Generation/3 Workshops

Editor for this issue: Jody Huellmantel <jodylinguistlist.org>

Directory

1. Michael Elhadad, International Natural Lang Generation Conf (INLG-2000) - 3 Workshops

			Call for participation

			 Three workshops
		to be held in parallel just before the

	 International Natural Language Generation Conference
			 INLG-2000
			 Mitzpe Ramon, Israel

		 Workshop date: 12 June 2000
		 Conference dates: 13-16 June 2000

 Conference home page: http://www.cs.bgu.ac.il/~nlg2000/

Workshops:

WS1 : Analysis for generation
 Chair: Svetlana Sheremetieva, New Mexico State University

WS2 : Why care for cognitive modeling when building NLG systems?
 Chair: Ralf Klabunde, University of Heidelberg

WS3 : Coherence in Generated Multimedia
 Co-chairs: Kees van Deemter, University of Brighton
 John Lee, University of Edinburgh


Workshop participants must register for the main conference.

Each workshop has individual submission dates. Please check the
announcement for the appropriate deadline.


======================================================================
		 WS1 : Analysis for generation
======================================================================

		 Analysis for Generation

	 http://crl.nmsu.edu/Events/external.htm

The last decade has seen an explosion in the work done in the field of
NLG with the emphasis on the development of independent NLG
applications rather than generation modules of MT systems. While it
seems natural to consider problems of analysis and generation as two
sides of a coin in such NLP applications as MT, researchers working on
"pure" generation systems sometimes treat problems arising at every
stage of generation -- content specification, sentence planning, and
surface realization -- as independent. Time may be ripe for examining
the mutual utility of analysis and generation in greater detail. The
impetus is, as can be expected, the goal of minimizing system-building
efforts in language engineering.

The workshop proposes to address the issues of

- Analysis as part of generation. 

 A modicum of analysis is, in fact, an essential part of every
 generation system. The input to generation systems such as raw data
 in tables, lists, diagrams, elements of various databases or even
 text snippets that are fed into the system directly by a user still
 must first be somehow processed, that is, analyzed The question
 arises whether it is possible to develop criteria to better choose
 and integrate analysis techniques which could be efficiently applied
 at different stages of generation.

- Reusability and adaptability of analysis techniques and tools for
 generation. 

 While it is not uncommon to believe that generation and analysis are
 not reversible, a number of contributions over the years have
 discussed reversibility of analysis and generation resources,
 especially the grammars and demonstrated how the use of reversible
 grammars may lead to efficient and flexible natural language parsing
 and generation systems. It is worth discussing constraints on
 reversibility.

- Reusability of analysis knowledge and methodology of its acquisition
 for generation. 

 Generation and analysis are closely related in that both processes
 use many similar resources, and often it is less expensive to
 reorganize an existing "analysis" resource (e.g., an analysis
 lexicon) than to acquire one for generation from scratch.It is worth
 discussing how to establish whether a resource built for analysis
 can be used for generation and at what price.

In particular the workshop will seek to address the following issues:

I. Applications of analysis in generation and types of analysis
techniques used in NLG. 

II. Reusability and adaptability of knowledge resources in generation
and analysis 

- knowledge representation
 - lexicon format and indexing
 - rule writing format
- knowledge acquisition and adaptation
 - reversibility of grammars
 - use of microtheories
- architectural issues
 - converting (morphological, syntactic, semantic, etc.) analyzers
 into generator modules
 - architectural peculiarities of systems involving both analysis
 and synthesis and reusability of their modules.

FORMAT FOR SUBMISSION

Paper submissions should consist of full papers (maximum of 12 pages
Including references, 12pt font size). Each submission should include
a separate title page providing the following information: the title,
a short abstract, names and affiliations of all the authors, the full
address of the primary author (or alternate contact person), including
phone, fax, and email. See the INLG main conference page
http://www.cs.bgu.ac.il/~nlg2000/ for details about the use of Latex
(preferred) or Word. Authors are asked to indicate explicitly in their
submission any special requirements that they may have (e.g. use of
VCR, internet access, data projector) beyond an overhead projector.
Please send your electronic submission(PostScript or PDF format) until
March 20 to:

Svetlana Sheremetyeva, 
Computing Research Laboratory New Mexico State University,
 USA Box30001/Dept.3CRL/Las Cruces New Mexico 88003-8001 
505 646 5466 (voice)
 505 646 6218 (fax)
 lanacrl.nmsu.edu

IMPORTANT DATES
Paper submission deadline: March 20 
Notification of acceptance: April 20
Camera ready paper to workshop coordinator: April 28 

Organizing Committee 

Svetlana Sheremetyeva, Chair and contact person
Computing Research Laboratory, New Mexico State University, USA
lanacrl.nmsu.edu

Sergei Nirenburg
Computing Research Laboratory, New Mexico State University, USA
sergeicrl.nmsu.edu

Richard Kittredge
Department of Linguistics and Translation, University of Montreal
kittredgeIRO.UMontreal.CA

Anna Sagvall Hein
Department of linguistics, Uppsala University
Annaling.uu.se

Evelyne Viegas
Microsoft Corporation
evelynevmicrosoft.com

Michael Zock
Language & Cognition LIMSI - CNRS 
zocklimsi.fr

======================================================================
 WS2 : Why care for cognitive modeling when building NLG systems?
======================================================================

 Why care for cognitive modeling when building NLG systems ?

	 http://pc03.idf.uni-heidelberg.de/~ralf/wkshop


WORKSHOP DESCRIPTION

Natural Language Generation (NLG) exists now for many years as a
subdiscipline of computational linguistics. Many systems have been
built with different goals and from different perspectives. While some
approaches are strongly driven by engineering concerns, others are
more concerned with insights in human language production.

By organizing this workshop we pursue three goals: 

(a) to show that cognitive and engineering approaches, rather then being 
 mutually exclusive, are highly complementary; 
(b) to identify some of the loci where the human factor should be taken into 
 account; 
(c) to discuss then what methods could be used in order to enhance current 
 systems or architectures by means of cognitive models of human language 
 generation.

While there is no doubt that cognitive modeling is useful for testing
theories of human behavior, it is probably also quite useful in
cognitive engineering, that is, as a complementary methodology for
building systems. If the engineering point of view is overemphasised,
designers will be more concerned with the machine than with their
final users: people. In order to build truly user-friendly
NL-generators, i.e. systems which adapt themselves to users rather
than the other way around, we need a deeper understanding of the
knowledge and the processes that people use when producing
language. These kinds of insights can profitably be used when building
systems, especially if they are meant to be used by people.

At present, we observe a strong tendency towards an engineering
approach. While many researchers in the field have based their systems
on empirical research, their approach still remains more motivated by
engineering considerations (efficiency) then by psycho-linguistic
factors (the problems people face). Yet we do believe, that
integrating the human factor into the engineering approach would
greatly enhance the overall quality (adequacy, flexibility, scope) of
the existing systems.

To approach these goals, the workshop invites full papers that deal with 
any aspect of the following topics:

- ARCHITECTURE (flexibility, decomposition and control of the process) 
- CONTENT DETERMINATION / CONCEPTUALIZATION 
- NATURE OF THE INPUT (proximity to language) 
- OUTLINE PLANNING 
- NP-GENERATION (the problem of reference) 
- LEXICAL ACCESS 

For more information on the workshop and its topics have a look at the 
workshop homepage.


WORKSHOP ORGANIZERS

Gerard Kempen, University of Leiden, The Netherlands
kempenrulfsw.leidenuniv.nl

Ralf Klabunde, University of Heidelberg, Germany (Chair) 
klabundenovell1.gs.uni-heidelberg.de

Koenraad de Smedt, University of Bergen, Norway
deSmedthf.uib.no	

Michael Zock, LIMSI - CNRS, France
zocklimsi.fr


CONTACT INFORMATION FOR QUESTIONS

If you have any questions, please contact Michael Zock: zocklimsi.fr


FORMAT FOR SUBMISSION

Paper submissions should consist of full papers (maximum of 12 pages
including references, 12pt font size). Each submission should include
a separate title page providing the following information: the title,
a short abstract, names and affiliations of all the authors, the full
address of the primary author (or alternate contact person), including
phone, fax, and email. Electronic submissions are preferred, and
should be sent to Ralf Klabunde until March 20. We strongly advise
standard html for electronic submissions, but PostScript or PDF form
is also possible.

Ralf Klabunde
University of Heidelberg
Center for Computational Linguistics
Karlstr. 2
69117 Heidelberg, Germany 

klabundenovell1.gs.uni-heidelberg.de


IMPORTANT DATES
Paper submission deadline:		March 20
Notification of acceptance:		April 12 
Final paper to workshop coordinator: May 2


======================================================================
	 WS3 : Coherence in Generated Multimedia
======================================================================

		 Coherence in Generated Multimedia

		http://www.hcrc.ed.ac.uk/~john/inlg-mm

Keywords: 

 - Generation of multimedia documents/presentations
 - Cross-media coreference, deixis, and anaphora
 - Media allocation, layout, and synchronization


More and more often, Natural Language Generation is performed as a
component of a larger Multimedia Presentation System, whose output
consists of language/speech combined with graphics, animation,
non-speech audio, etc. Recent years have seen a growing interest in
various issues relevant for the design of such systems, such as the
issue of multimedia system architecture (e.g., Bordegoni et al. 1997)
and media allocation (e.g., ETAI 1997-8). It is gradually becoming
clear that a Multimedia Presentation System forces its designers to
rethink some fundamental issues, at the core of which is a generalized
notion of document coherence, which subsumes the purely linguistic
notion of coherence, and which can take different forms depending on
the type of document.

This workshop, which is associated with the INLG-2000 conference,
invites submission of papers that shed light on the issue of discourse
coherence in relation to the generation of documents/ presentations in
which **natural language + at least one other medium** play a
nontrivial role. Topics of interest include but are not limited to:

- Document/presentation structure. How can theories of discourse 
 structure (e.g. Rhetorical Structure Theory) be enhanced to 
 cover types of discourse that combine natural language/speech 
 and other media?
- Media allocation. How does the system decide what combination 
 of media is used for expressing a given item of information, 
 and how can authors be allowed to influence such decisions?
- Interlinguality. For example, what types of semantic formalism
 are most suitable for expressing the meanings of expressions 
 that use different media (e.g., pictures as well as text)?
- `Fusion' of information from different media. For example,
 how can the referring expressions generated by a Multimedia
 Presentation System be simplified if the system is able to
 use pointing?
- Cross-media coreference, deixis, and anaphora. For example,
 when are expressions like `the method illustrated in figure 5'
 felicitous, and how can a system be enabled to generate
 such expressions?
- Media layout (in the case of a written document) and
 media synchronization (e.g., in the case of a presentation
 by a life-like agent).
- Corpora. Multimedia corpora are an obvious potential source of 
 information for multimedia generation, but how can connections 
 between different media be captured? (E.g., how should pictures, 
 graphs, or gestures be annotated?)
- System architecture. For example, can the requirements of
 generating coherent multimedia be reconciled with the 
 advantages of a pipeline architecture (e.g. Reiter 1994,
 McKeown et al. 1992)?
- Evaluation of the quality (e.g. coherence) of documents
 or presentations generated by a Multimedia Presentation 
 System.

Presentations containing live demonstrations are welcome, but there is
also room for purely theoretical contributions. Each presentation will
be followed by a comment from one of the other participants, who will
have been enabled to see the final version of the paper beforehand.
Submissions format: Submissions (deadline: 31 March) have a preferred
length of about 5 double-spaced pages (not counting title page,
abstract, and references). See the INLG main conference page
http://www.cs.bgu.ac.il/~nlg2000/ for details about the use of Latex
(preferred) or Word. The deadline for camera-ready final versions is
4 May. Authors are asked to indicate explicitly in their submission
any special requirements that they may have (e.g. use of VCR, internet
access, data projector) beyond an overhead projector.

 References:

- AIR (1995). Special Issue "Integration of Natural Language 
 and Vision Processing: Intelligent Multimedia". Artif. 
 Intell. Review 9, Nos.2-3.
- Bordegoni et al (1997). Bordegoni, Faconti, Feiner, Maybury, 
 Rist, Ruggieri, Trahanias, and Wilson (1997): A Standard 
 Reference Model for Intelligent Multimedia Presentation 
 Systems. Computer Standards and Interfaces 18, pp.477-496.
- ETAI (1997, 1998). ETAI News Journal on Intelligent User 
 Interfaces, Vol 1, No's 1 and 2. See especially 
 http://www.dfki.de/etai/statements/reiter-nov-97-responses.html
- Maybury and Wahlster (1998). Readings in Intelligent User
 Interfaces. Morgan Kaufmann, San Francisco.
- McKeown et al. (1992). McKeown, Feiner, Robin, Seligman, 
 Tanenblatt. Generating Cross-references for Multimedia 
 Explanation. In Procs. of Tenth National Conf. on Artif. 
 Intell., p.9-16. Menlo Park.
- Reiter (1994). Has a Consensus NL Generation Architecture 
 Appeared, and is it Psycholinguistically Plausible? In Proc. 
 of 7th Int. Generation Ws. Kennebunkport, Maine.


List of organizers (in alphabetical order)

 Elisabeth Andre (Saarbruecken)
 John Lee (Co-chair, Edinburgh)
 James Lester (Raleigh, NC)
 Johanna Moore (Edinburgh)
 Jon Oberlander (Edinburgh)
 Ivandre Paraboni (Brighton)
 Ehud Reiter (Aberdeen)
 Thomas Rist (Saarbruecken)
 Laurent Romary (Loria) 
 Donia Scott (Brighton)
 Kees van Deemter (Co-chair, Brighton)

Contact for questions: <Kees.van.Deemteritri.brighton.ac.uk>


Submission addresses: Submissions should be sent to *both* of
the co-chairs:

 <johncogsci.ed.ac.uk> (John Lee)
 <Kees.van.Deemteritri.brighton.ac.uk> (Kees van Deemter)


 Please say "INLG Multimedia ws submission" in the subject line 
 of your message.


Deadlines: 

Submissions due by 31 March
Acceptance/rejection notices sent by 21 April
Camera-ready papers due by 4 May



- 
Michael Elhadad
http://www.cs.bgu.ac.il/~elhadad
Dept of Computer Science, Ben Gurion University
Beer Sheva, Israel