FOURTH WORKSHOP ON
STATISTICAL MACHINE TRANSLATION
March 30 and 31, 2009
Shared task test data release: December 8, 2008
Paper submission deadline: January 9, 2009
This workshop builds on three previous workshops on statistical machines translation: the NAACL-2006 Workshop on Statistical Machine Translation, the ACL-2007 Workshop on Statistical Machine Translation, and the ACL-2008 Workshop on Statistical Machine Translation. This year's workshop will feature three shared tasks: a shared translation task for 10 pairs of European languages, a shared evaluation task to test automatic evaluation metrics, and a new system combination task combining the output of all the systems entered into the shared translation task.
The workshop will also feature scientific papers on topics related to MT. Topics of interest include, but are not limited to:
* word-based, phrase-based, syntax-based SMT
* using comparable corpora for SMT
* incorporating linguistic information into SMT
* system combination
* error analysis
* manual and automatic method for evaluating MT
* scaling MT to very large data sets
We encourage authors to evaluate their approaches to the above topics using the common data sets created for the shared tasks.
SHARED TRANSLATION TASK
The first shared task which will examine translation between the following language pairs:
* English-German and German-English
* English-French and French-English
* English-Spanish and Spanish-English
* English-Czech and Czech-English
* English-Hungarian and Hungarian-English
Participants may submit translations for any or all of the language directions. In addition to the common test sets the workshop organizers will provide optional training resources, including a newly expanded release of the Europarl corpora and out-of-domain corpora.
All participants who submit entries will have their translations evaluated. We will evaluate translation performance by human judgment. To facilitate the human evaluation we will require participants in the shared tasks to manually judge some of the submitted translations.
A more detailed description of the shared translation task (including information about the test and training corpora, a freely available MT system, and a number of other resources) is available from http://www.statmt.org/wmt09/translation-task.html. We also provide two baseline machine translation systems, whose performance is comparable to the best systems from last year's shared task.
SYSTEM COMBINATION TASK
Participants in the system combination task will be provided with the 1-best translations from each of the systems entered in the shared translation task. We will endeavor to provide a held-out development set for system combination, which will include translations from each of the systems and a reference translation. Any system combination strategy is acceptable, whether it selects the best translation on a per sentence basis or create novel translations by combining the systems' translations.
The quality of the system combinations will be judged alongside the individual systems during the manual evaluation, as well as scored with automatic evaluation metrics.
More details of the system combination task is available from http://www.statmt.org/wmt08/system-combination-task.html.
The evaluation task will assess automatic evaluation metrics' ability to:
* Rank systems on their overall performance on the test set
* Rank systems on a sentence by sentence level
Participants in the shared evaluation task will use their automatic evaluation metrics to score the output from the translation task and the system combination task. They will be provided with the output from the other two shared tasks along with reference translations. We will measure the correlation of automatic evaluation metrics with the human judgments.
More details of the shared evaluation task (including submission formats and the collected manual evaluations from last year's workshop) is available from http://www.statmt.org/wmt08/evaluation-task.html.
PAPER SUBMISSION INFORMATION
Submissions will consist of regular full papers of max. 8 pages, formatted following the EACL 2009 guidelines. In addition, shared task participants will be invited to submit short papers (max. 4 pages) describing their systems or their evaluation metrics. Both submission and review processes will be handled electronically.
We encourage individuals who are submitting research papers to evaluate their approaches using the training resources provided by this workshop and past workshops, so that their experiments can be repeated by others using these publicly available corpora.
Paper submission dates:
January 9, 2009 Paper submission deadline (long and short papers) January 30, 2009 Notification of acceptance February 13, 2009 Camera-ready deadline
System submission and evaluation dates:
December 8, 2008 Test set distributed for translation task December 12, 2008 Submission deadline for translation task December 22, 2008 Translations release for system combination January 5, 2009 System combination deadline January 12, 2009 Translations (and combinations) released for evaluation January 23, 2009 Automatic evaluation scores due January 12, 2009 Start of manual evaluation period January 30, 2009 End of manual evaluation
Subscribe to the announcement list for WMT09 by visiting the Workshop's web site: http://www.statmt.org/wmt09/
This list will be used to announce when the test sets are released, to indicate any corrections to the training sets, and to amend the deadlines as needed.
Chris Callison-Burch (Johns Hopkins University) Philipp Koehn (University of Edinburgh) Christof Monz (University of London) Josh Schroeder (University of Edinburgh)
For questions, comments, etc. please send email to pkoehn at inf.ed.ac.uk