Call for Participation

The International Workshop on Spoken Language Translation (IWSLT) is a yearly scientific workshop, associated with an open evaluation campaign on spoken language translation, where both system descriptions and scientific papers are presented The 7th International Workshop on Spoken Language Translation will take place in Paris, France on 2-3 December 2010.

Scientific Papers

The IWSLT invites submissions of scientific papers to be published in the workshop proceedings and presented in dedicated technical sessions of the workshop, either in oral or poster form. The workshop welcomes high quality contributions covering theoretical and practical issues in the field of machine translation. Possible topics include, but are not limited to:

• Speech and text MT
• Integration of ASR and MT
• MT approaches
• MT evaluation
• Language resources for MT
• Open source software for MT
• Pivot-language-based MT
• Adaptation in MT
• Simultaneous speech translation
• Efficiency in MT
• Stream-based algorithms for MT

Submitted manuscripts will be peer-reviewed by at lest three members of the workshop program committee.  Authors of accepted technical papers are requested to present their paper at the workshop.

Evaluation Campaign

IWSLT's evaluations are not competition-oriented, but their goal is to foster cooperative work and scientific exchange. In this respect, IWSLT proposes challenging research tasks and an open experimental infrastructure for the scientific community working on spoken and written language translation. This year, the IWSLT evaluation campaign will offer three spoken language translation tasks:

public speeches on a variety of topics, from English to French (NEW CHALLENGE)
spoken dialogues in travel situations, between Chinese and English
• traveling expressions, from Arabic, Turkish, and French to English.  

For each task, monolingual and bilingual language resources will be provided to participants in order to train their translation systems, as well as sets of manual and automatic speech transcripts (with n-best and lattices) and reference translations.
Moreover, blind test sets will be released and all translation outputs produced by the participants will be evaluated using several automatic translation quality metrics. Human assessment will be carried out for the translation of spoken dialogues and basic travel expressions. 

The goal of this year's new challenge (translation of public speeches) will be to establish reference baselines and appropriate evaluation protocols for future evaluations. As a consequence, altough an evaluation server will be set-up to compute several translation accuracy metrics, there will be no official ranking of participants published by the organizers for this task. 

Each participant in the evaluation campaign is requested to submit a paper describing the MT system, the utilized resources, and results using the provided test data. Contrastive run submissions using only the bilingual resources provided by IWSLT as well as investigations of the contribution of each utilized resource are highly appreciated. Finally, all participants are requested to present their papers at the workshop.