Impacts in Natural Language Generation
NLG Between Technology and Applications
July 24-28, 2000
Schloss Dagstuhl, Saarland, Germany
- immediately preceding COLING 2000 -
Homepage: http://impacts.dfki.de
The tension between theoretical work and its implementation has often
been considered fruitful. In the field of Natural Language Generation,
it is now complemented by another tension, the one between technologies
and applications:
"I have invented a new technique for NLG!" - "What is its impact on
applications?"
"I have built a new NLG application!" - "What is its impact on the
technology?"
Much of NLG technology is based on a theoretical understanding of the
process of language generation, whereas the applications strongly rely
on practical requirements. There are not many theoretically
well-motivated technologies that can straightforwardly be employed
within specific applications. For this workshop, we adopt a broad
notion of application by including pieces of software containing NLG
technology that currently are used by others in order to solve
real-world tasks.
Some NLG application developers find it preferable to not reuse existing
technology. This is often due to the lack of solutions for the knowledge
bottleneck and for the input formation bottleneck: NLG technology lacks
the power of dealing with external conceptual lexical knowledge bases,
and it also lacks standards of representing inputs at a suitable
specificity.
The "IMPACTS" workshop aims at studying, discussing, and clarifying the
relationships between NLG technologies and applications. The workshop
addresses
- researchers and developers in NLG,
- current and potential users of NLG applications,
- providers of large conceptual lexicons usable by NLG, and
- developers of systems that deal with the input specificity problem.
We invite original and unpublished contributions from all areas of NL
generation, either from the technological side or from the applications
point of view. They must state clearly how they relate to the respective
counterpart, hence addressing one of the above questions.
"IMPACTS" will take place from July 24-28, 2000 at International
Conference and Research Center for Computer Science at Schloss Dagstuhl.
It immediately precedes the events of COLING 2000 taking place in
Luxembourg, Saarbrücken, and Nancy. Schloss Dagstuhl is situated in the
Saarland, allowing participants to reach the COLING tutorials on July 29
conveniently.
Dates:
24.-28. July 2000: Workshop at Schloss Dagstuhl
15. June 2000: Camera-ready copies
20. May 2000: Notification of acceptance
01. April 2000: Deadline for submissions
Programme Committee
John Bateman, University of Bremen, Germany
Tilman Becker, DFKI Saarbrücken, Germany (Program Co-Chair)
Stephan Busemann, DFKI Saarbrücken, Germany (Program Co-Chair)
Robert Dale, Microsoft Research Institute, Macquarie University,
Australia
Laurence Danlos, LORIA, France
Michael Elhadad, Ben-Gurion University, Israel
Eduard Hovy, ISI, University of Southern California, USA
Richard Kittredge, CoGenTex Inc., USA
Inderjeet Mani, Mitre Corporation, USA
David D. McDonald, Gensym Corporation, USA
Cecile Paris, CSIRO Mathematical and Information Sciences, Macquarie
University, Australia
Owen Rambow, AT&T, USA
Ehud Reiter, University of Aberdeen, UK
Donia Scott, ITRI, University of Brighton, UK
More information can be found at the Workshop homepage, which is located
at http://impacts.dfki.de and which will be extended during the next
months.
-- Dr. Stephan Busemann, DFKI GmbH Stuhlsatzenhausweg 3, D-66123 Saarbruecken phone: (+49 681) 302-5286, fax: (+49 681) 302-5338 web: http://www.dfki.de/~busemann