Abstract
This paper presents NTT-NAIST SMT systems for English-German and German-English MT tasks of the IWSLT 2013 evaluation campaign. The systems are based on generalized minimum Bayes risk system combination of three SMT systems: forest-to-string, hierarchical phrase-based, phrasebased with pre-ordering. Individual SMT systems include data selection for domain adaptation, rescoring using recurrent neural net language models, interpolated language models, and compound word splitting (only for German-English).- Anthology ID:
- 2013.iwslt-evaluation.12
- Volume:
- Proceedings of the 10th International Workshop on Spoken Language Translation: Evaluation Campaign
- Month:
- December 5-6
- Year:
- 2013
- Address:
- Heidelberg, Germany
- Venue:
- IWSLT
- SIG:
- SIGSLT
- Publisher:
- Note:
- Pages:
- Language:
- URL:
- https://aclanthology.org/2013.iwslt-evaluation.12
- DOI:
- Cite (ACL):
- Katsuhito Sudoh, Graham Neubig, Kevin Duh, and Hajime Tsukada. 2013. NTT-NAIST SMT systems for IWSLT 2013. In Proceedings of the 10th International Workshop on Spoken Language Translation: Evaluation Campaign, Heidelberg, Germany.
- Cite (Informal):
- NTT-NAIST SMT systems for IWSLT 2013 (Sudoh et al., IWSLT 2013)
- PDF:
- https://preview.aclanthology.org/ingestion-script-update/2013.iwslt-evaluation.12.pdf