1st Call for Papers

Workshop on Grammar Engineering Across Frameworks - Geaf 2009

on August 6 at the ACL/IJCNLP 2009 Conference, Singapore

http://www-csli.stanford.edu/~thking/GEAF09/


Workshop Topics

This workshop aims to bring together grammar engineers from
different frameworks to compare research and methodologies,
particularly around the themes of evaluation, modularity,
maintainability, relevance to theoretical and computational
linguistics, and applications of "deep" grammars to real-world domains
and NLP tasks.

Recent years have seen the development of techniques and
resources to support robust, deep grammatical analysis of natural
language in real-world domains and applications. The demands of these
types of tasks have resulted in significant advances in areas such as
parser efficiency, hybrid statistical/symbolic approaches to
disambiguation, and the acquisition of large-scale lexicons. The
effective acquisition, development, maintenance and enhancement of
grammars is a central issue in such efforts, and the size and
complexity of realistic grammars makes these tasks extremely
challenging; indeed, these tasks are often tackled in ways that have
much in common with software engineering. This workshop aims to bring
together grammar engineers from different frameworks --- for example
LFG, HPSG, TAG, CCG, dependency grammar --- to compare their research
and methodologies.

The workshop will solicit submissions for papers on the following themes:

1. Evaluation: Proposals concerning evaluation methodologies and
metrics which can capture the added benefits of deep linguistic
analysis; evaluation techniques which can compare grammars across
varieties/languages;

2. Modularity: Reflections on which aspects of linguistic structure
can most easily be separated out from each other, why and how the
analyses of separate linguistic phenomena are
interconnected/interdependent, and the role of frameworks in promoting
or inhibiting modularity;

3. Maintainability: Techniques for improving long-term and
multi-developer maintainability of grammars; impacts of considerations
of maintainability on choices of linguistic analysis;

4. Relevance to theoretical and computational linguistics:
Reflections on how to present grammar engineering work to other
research communities;

5. Regression testing: Evaluation for internal purposes;
methodologies and techniques for test suite construction, role of test
suites in day-to-day progress on grammars.

6. Multilingualism: Approaches, techniques and tools improving
development and maintainability of grammars for multiple system
languages.


Important Dates

Paper submission deadline: 1 May, 2009
Notification of acceptance of papers: 1 June, 2009
Camera-ready copy of papers due: 7 June, 2009
Demo session requests due: 8 May, 2009
ACL-IJCNLP 2009 Workshops: Aug 6, 2009


Paper Submission

All papers must be submitted via the START system at:

https://www.softconf.com/acl-ijcnlp09/GEAF/submit.html

Style files for submission are available at:

http://www.acl-ijcnlp-2009.org/main/authors/stylefiles/

Submissions are 8 pages with one extra page for references.
Please submit pdf (not Word).

Demo session requests will not be published in the proceedings
but will be listed in the program. The requests should contain
a one page description of the demo.


Programme Committee

* Emily Bender, Washington
* Miriam Butt, Konstanz
* John Carroll, Sussex
* Stephen Clark, Oxford
* Ann Copestake, Cambridge
* Berthold Crysmann, Bonn
* Mary Dalrymple, Oxford
* Stefanie Dipper, Bochum
* Dan Flickinger, Stanford
* Josef van Genabith, Dublin
* Julia Hockenmaier, Illinois
* Ron Kaplan, Powerset
* Montserrat Marimon, Barcelona
* Yusuke Miyao, Tokyo
* Jun'ichi Tsujii, Tokyo / Manchester

Organizers and Contacts for Inquiries

* Tracy Holloway King, Microsoft / Powerset: Tracy [dot] King [at]
microsoft [dot] com
* Marianne Santaholma, Geneva University: marianne [dot]
santaholma [at] unige [dot] ch

Proceedings

Accepted papers will form part of the workshop proceeedings and
will be available on the conference CD+hard copy.