NTT’s Neural Machine Translation Systems for WMT 2018

Morishita, Makoto; Suzuki, Jun; Nagata, Michio

doi:10.18653/v1/w18-6421

Cited by 13 publications

(21 citation statements)

References 12 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Further details on training are not available. (Morishita et al, 2018) NTT combine Transformer "big" model, corpus cleaning technique for provided and synthetic parallel corpora, and right-to-left n-best re-ranking techniques. Through their experiments, NTT found filtering of noisy training sentences and right-to-left re-ranking as the keys to better accuracy.…”

Section: Njunmtmentioning

confidence: 99%

Findings of the 2018 Conference on Machine Translation (WMT18)

Bojar¹,

Federmann²,

Fishel³

et al. 2018

Proceedings of the Third Conference on Machine Translation: Shared Task Papers

269

223

View full text Add to dashboard Cite

This paper presents the results of the premier shared task organized alongside the Conference on Machine Translation (WMT) 2018. Participants were asked to build machine translation systems for any of 7 language pairs in both directions, to be evaluated on a test set of news stories. The main metric for this task is human judgment of translation quality. This year, we also opened up the task to additional test sets to probe specific aspects of translation. Language Sources (Number of Documents) English ABC News (1), BBC (4), Brisbane Times (1), CBS News (1), Daily Mail (4), Euronews (3), Globe and Mail (1), Guardian (4), Independent (4), Los Angeles Times (4), MSNBC (3), Novinte (2), New York Times (2), Reuters (3), Russia Today (2), Scotsman (2), Sydney Morning Herald (2), Telegraph (2), The Local (2), Time Magazine (2), UPI (1), Washington Post (3) Czech blesk.cz (16), deník.cz (5), Deník Referendum (1), DNES.cz (7), lidovky.cz (6), Novinky.cz (3), Reflex (2), tyden.cz (12), ZDN (2) German Aachener Nachrichten (1), Abendzeitung Nürnberg (2), Braunschweiger Zeitung (1

show abstract

Section: Njunmtmentioning

confidence: 99%

Findings of the 2018 Conference on Machine Translation (WMT18)

Bojar¹,

Federmann²,

Fishel³

et al. 2018

Proceedings of the Third Conference on Machine Translation: Shared Task Papers

269

223

View full text Add to dashboard Cite

show abstract

“…They already released earlier versions of the corpora and they were used on the WMT 2018 news shared translation tasks (Bojar et al, 2018). The WMT shared task participants reported that this corpora boosted translation accuracy when used with careful corpus cleaning (Junczys-Dowmunt, 2018;Morishita et al, 2018).…”

Section: Jparacrawl: Unconstrained Settingmentioning

confidence: 99%

NTT Neural Machine Translation Systems at WAT 2019

Morishita

Suzuki

Nagata

2019

Proceedings of the 6th Workshop on Asian Translation

Self Cite

View full text Add to dashboard Cite

In this paper, we describe our systems that were submitted to the translation shared tasks at WAT 2019. This year, we participated in two distinct types of subtasks, a scientific paper subtask and a timely disclosure subtask, where we only considered English-to-Japanese and Japanese-to-English translation directions. We submitted two systems (En-Ja and Ja-En) for the scientific paper subtask and two systems (Ja-En, texts, items) for the timely disclosure subtask. Three of our four systems obtained the best human evaluation performances. We also confirmed that our new additional web-crawled parallel corpus improves the performance in unconstrained settings.

show abstract

“…Effectiveness of filtering noisy data in neural machine translation. Researchers in the field of neural machine translation (NMT) have recognized that collecting high-quality training data to be equally or even more important than exploring sophisticated model architectures (Koehn et al, 2018;Morishita et al, 2018). Techniques used in neural response generation and NMT are nearly identical; e.g., sequenceto-sequence models (Sutskever et al, 2014) and Transformers (Vaswani et al, 2017) are often used as base model architectures.…”

Section: Introductionmentioning

confidence: 99%

Filtering Noisy Dialogue Corpora by Connectivity and Content Relatedness

Akama¹,

Yokoi

Suzuki

et al. 2020

Proceedings of the 2020 Conference on Empirical Methods in Natural Language Processing (EMNLP)

Self Cite

View full text Add to dashboard Cite

Large-scale dialogue datasets have recently become available for training neural dialogue agents. However, these datasets have been reported to contain a non-negligible number of unacceptable utterance pairs. In this paper, we propose a method for scoring the quality of utterance pairs in terms of their connectivity and relatedness. The proposed scoring method is designed based on findings widely shared in the dialogue and linguistics research communities. We demonstrate that it has a relatively good correlation with the human judgment of dialogue quality. Furthermore, the method is applied to filter out potentially unacceptable utterance pairs from a large-scale noisy dialogue corpus to ensure its quality. We experimentally confirm that training data filtered by the proposed method improves the quality of neural dialogue agents in response generation. 1

show abstract

NTT’s Neural Machine Translation Systems for WMT 2018

Cited by 13 publications

References 12 publications

Findings of the 2018 Conference on Machine Translation (WMT18)

Findings of the 2018 Conference on Machine Translation (WMT18)

NTT Neural Machine Translation Systems at WAT 2019

Filtering Noisy Dialogue Corpora by Connectivity and Content Relatedness

Contact Info

Product

Resources

About