A study in greedy oracle improvement of translation hypotheses

Benjamin Marie, Aurélien Max


Abstract
This paper describes a study of translation hypotheses that can be obtained by iterative, greedy oracle improvement from the best hypothesis of a state-of-the-art phrase-based Statistical Machine Translation system. The factors that we consider include the influence of the rewriting operations, target languages, and training data sizes. Analysis of our results provide new insights into some previously unanswered questions, which include the reachability of previously unreachable hypotheses via indirect translation (thanks to the introduction of a rewrite operation on the source text), and the potential translation performance of systems relying on pruned phrase tables.
Anthology ID:
2013.iwslt-papers.16
Volume:
Proceedings of the 10th International Workshop on Spoken Language Translation: Papers
Month:
December 5-6
Year:
2013
Address:
Heidelberg, Germany
Venue:
IWSLT
SIG:
Publisher:
Note:
Pages:
Language:
URL:
https://aclanthology.org/2013.iwslt-papers.16
DOI:
Bibkey:
Cite (ACL):
Benjamin Marie and Aurélien Max. 2013. A study in greedy oracle improvement of translation hypotheses. In Proceedings of the 10th International Workshop on Spoken Language Translation: Papers, Heidelberg, Germany.
Cite (Informal):
A study in greedy oracle improvement of translation hypotheses (Marie & Max, IWSLT 2013)
Copy Citation:
PDF:
https://preview.aclanthology.org/update-css-js/2013.iwslt-papers.16.pdf