Document-Wide Decoding for Phrase-Based Statistical Machine Translation

Christian Hardmeier; Joakim Nivre; Jörg Tiedemann

Document-Wide Decoding for Phrase-Based Statistical Machine Translation

Christian Hardmeier, Joakim Nivre, Jörg Tiedemann

Publikation: Konference artikel i Proceeding eller bog/rapport kapitel › Konferencebidrag i proceedings › Forskning › peer review

Abstract

Independence between sentences is an assumption deeply entrenched in the models and algorithms used for statistical machine translation (SMT), particularly in the popular dynamic programming beam search decoding algorithm. This restriction is an obstacle to research on more sophisticated discourse-level models for SMT. We propose a stochastic local search decoding method for phrase-based SMT, which permits free document-wide dependencies in the models. We explore the stability and the search parameters of this method and demonstrate that it can be successfully used to optimise a document-level semantic language model.

Originalsprog	Engelsk
Titel	Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning
Publikationsdato	14 jul. 2012
ISBN (Trykt)	978-1-937284-43-5
Status	Udgivet - 14 jul. 2012
Udgivet eksternt	Ja

Adgang til dokumentet

https://aclanthology.org/D12-1108.pdf

Citationsformater

@inproceedings{dc98dfc88d4c48d2b85b160c3a9ae037,

title = "Document-Wide Decoding for Phrase-Based Statistical Machine Translation",

abstract = "Independence between sentences is an assumption deeply entrenched in the models and algorithms used for statistical machine translation (SMT), particularly in the popular dynamic programming beam search decoding algorithm. This restriction is an obstacle to research on more sophisticated discourse-level models for SMT. We propose a stochastic local search decoding method for phrase-based SMT, which permits free document-wide dependencies in the models. We explore the stability and the search parameters of this method and demonstrate that it can be successfully used to optimise a document-level semantic language model.",

author = "Christian Hardmeier and Joakim Nivre and J{\"o}rg Tiedemann",

year = "2012",

month = jul,

day = "14",

language = "English",

isbn = "978-1-937284-43-5",

booktitle = "Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning",

}

Document-Wide Decoding for Phrase-Based Statistical Machine Translation. / Hardmeier, Christian; Nivre, Joakim; Tiedemann, Jörg.
Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning. 2012.

Publikation: Konference artikel i Proceeding eller bog/rapport kapitel › Konferencebidrag i proceedings › Forskning › peer review

TY - GEN

T1 - Document-Wide Decoding for Phrase-Based Statistical Machine Translation

AU - Hardmeier, Christian

AU - Nivre, Joakim

AU - Tiedemann, Jörg

PY - 2012/7/14

Y1 - 2012/7/14

N2 - Independence between sentences is an assumption deeply entrenched in the models and algorithms used for statistical machine translation (SMT), particularly in the popular dynamic programming beam search decoding algorithm. This restriction is an obstacle to research on more sophisticated discourse-level models for SMT. We propose a stochastic local search decoding method for phrase-based SMT, which permits free document-wide dependencies in the models. We explore the stability and the search parameters of this method and demonstrate that it can be successfully used to optimise a document-level semantic language model.

AB - Independence between sentences is an assumption deeply entrenched in the models and algorithms used for statistical machine translation (SMT), particularly in the popular dynamic programming beam search decoding algorithm. This restriction is an obstacle to research on more sophisticated discourse-level models for SMT. We propose a stochastic local search decoding method for phrase-based SMT, which permits free document-wide dependencies in the models. We explore the stability and the search parameters of this method and demonstrate that it can be successfully used to optimise a document-level semantic language model.

M3 - Article in proceedings

SN - 978-1-937284-43-5

BT - Proceedings of the 2012 Joint Conference on Empirical Methods in Natural Language Processing and Computational Natural Language Learning

ER -

Document-Wide Decoding for Phrase-Based Statistical Machine Translation

Abstract

Adgang til dokumentet

Fingeraftryk

Citationsformater