Matches in SemOpenAlex for { <https://semopenalex.org/work/W3038817008> ?p ?o ?g. }
- W3038817008 abstract "We show that Bayes' rule provides an effective mechanism for creating document translation models that can be learned from only parallel sentences and monolingual documents---a compelling benefit as parallel documents are not always available. In our formulation, the posterior of a candidate translation is the product of the unconditional (prior) of the candidate output document and the reverse translation probability of translating the candidate output back into the source language. Our proposed model uses a powerful autoregressive language model as the prior on target language documents, but it assumes that each sentence is translated independently from the target to the source language. Crucially, at test time, when a source document is observed, the document language model prior induces dependencies between the translations of the source sentences in the posterior. The model's independence assumption not only enables efficient use of available data, but it additionally admits a practical left-to-right beam-search algorithm for carrying out inference. Experiments show that our model benefits from using cross-sentence context in the language model, and it outperforms existing document translation approaches." @default.
- W3038817008 created "2020-07-10" @default.
- W3038817008 creator A5014554970 @default.
- W3038817008 creator A5036811894 @default.
- W3038817008 creator A5050190035 @default.
- W3038817008 creator A5050487837 @default.
- W3038817008 creator A5060960340 @default.
- W3038817008 creator A5061131385 @default.
- W3038817008 creator A5067222034 @default.
- W3038817008 date "2019-10-01" @default.
- W3038817008 modified "2023-09-27" @default.
- W3038817008 title "Better Document-Level Machine Translation with Bayes' Rule" @default.
- W3038817008 cites W1489181569 @default.
- W3038817008 cites W1494598192 @default.
- W3038817008 cites W1915251500 @default.
- W3038817008 cites W2006969979 @default.
- W3038817008 cites W2109664771 @default.
- W3038817008 cites W2124807415 @default.
- W3038817008 cites W2149327368 @default.
- W3038817008 cites W2594610113 @default.
- W3038817008 cites W2792376130 @default.
- W3038817008 cites W2799051177 @default.
- W3038817008 cites W2891534142 @default.
- W3038817008 cites W2914120296 @default.
- W3038817008 cites W2915573484 @default.
- W3038817008 cites W2917166349 @default.
- W3038817008 cites W2922709902 @default.
- W3038817008 cites W2927746189 @default.
- W3038817008 cites W2944815030 @default.
- W3038817008 cites W2947813521 @default.
- W3038817008 cites W2950813464 @default.
- W3038817008 cites W2952444318 @default.
- W3038817008 cites W2952446148 @default.
- W3038817008 cites W2962712961 @default.
- W3038817008 cites W2962739339 @default.
- W3038817008 cites W2962784628 @default.
- W3038817008 cites W2962802109 @default.
- W3038817008 cites W2962824887 @default.
- W3038817008 cites W2962832505 @default.
- W3038817008 cites W2962943802 @default.
- W3038817008 cites W2963216553 @default.
- W3038817008 cites W2963341956 @default.
- W3038817008 cites W2963403868 @default.
- W3038817008 cites W2963443683 @default.
- W3038817008 cites W2963532001 @default.
- W3038817008 cites W2963602293 @default.
- W3038817008 cites W2963736842 @default.
- W3038817008 cites W2963756346 @default.
- W3038817008 cites W2963833497 @default.
- W3038817008 cites W2963842551 @default.
- W3038817008 cites W2964093087 @default.
- W3038817008 cites W2964110616 @default.
- W3038817008 cites W2964121744 @default.
- W3038817008 cites W2964202354 @default.
- W3038817008 cites W2964260331 @default.
- W3038817008 cites W2964289193 @default.
- W3038817008 cites W2964291396 @default.
- W3038817008 cites W2967853777 @default.
- W3038817008 cites W2970295111 @default.
- W3038817008 cites W2970529093 @default.
- W3038817008 cites W2970694516 @default.
- W3038817008 cites W2970752831 @default.
- W3038817008 cites W2970947975 @default.
- W3038817008 cites W2971274815 @default.
- W3038817008 cites W2971347700 @default.
- W3038817008 cites W2971760773 @default.
- W3038817008 cites W2993383518 @default.
- W3038817008 hasPublicationYear "2019" @default.
- W3038817008 type Work @default.
- W3038817008 sameAs 3038817008 @default.
- W3038817008 citedByCount "2" @default.
- W3038817008 countsByYear W30388170082019 @default.
- W3038817008 countsByYear W30388170082020 @default.
- W3038817008 crossrefType "posted-content" @default.
- W3038817008 hasAuthorship W3038817008A5014554970 @default.
- W3038817008 hasAuthorship W3038817008A5036811894 @default.
- W3038817008 hasAuthorship W3038817008A5050190035 @default.
- W3038817008 hasAuthorship W3038817008A5050487837 @default.
- W3038817008 hasAuthorship W3038817008A5060960340 @default.
- W3038817008 hasAuthorship W3038817008A5061131385 @default.
- W3038817008 hasAuthorship W3038817008A5067222034 @default.
- W3038817008 hasConcept C104317684 @default.
- W3038817008 hasConcept C105580179 @default.
- W3038817008 hasConcept C105795698 @default.
- W3038817008 hasConcept C107673813 @default.
- W3038817008 hasConcept C120012220 @default.
- W3038817008 hasConcept C137293760 @default.
- W3038817008 hasConcept C149364088 @default.
- W3038817008 hasConcept C151730666 @default.
- W3038817008 hasConcept C154945302 @default.
- W3038817008 hasConcept C185592680 @default.
- W3038817008 hasConcept C203005215 @default.
- W3038817008 hasConcept C204321447 @default.
- W3038817008 hasConcept C207201462 @default.
- W3038817008 hasConcept C2776214188 @default.
- W3038817008 hasConcept C2777530160 @default.
- W3038817008 hasConcept C2779343474 @default.
- W3038817008 hasConcept C33923547 @default.
- W3038817008 hasConcept C35651441 @default.
- W3038817008 hasConcept C41008148 @default.