Matches in SemOpenAlex for { <https://semopenalex.org/work/W4311897722> ?p ?o ?g. }
- W4311897722 endingPage "512" @default.
- W4311897722 startingPage "500" @default.
- W4311897722 abstract "Neural machine translation uses a decoder to generate target words auto-regressively by predicting the next target word conditioned on a given source sentence and its previously predicted target words, i.e, its translation history, which suffers from two limitations: 1) the prediction of next word depends heavily on the quality of its history information. Moreover, the discrepancy between training and inference exacerbates this limitation; 2) this left-to-right decoding way cannot make full use of the target-side future information, which leads to the issue of unbalanced outputs. On the one hand, we alleviate the first limitation with a history-refining module, which learns to examine the quality of each history word by assigning it a confidence score. The confidence score is further used as a gate to control the amount of its word embedding flowing to the decoder. On the other hand, we attack the second limitation with a future-foreseeing module, which learns the distribution of future translation at each decoding time step. More importantly, we further propose refining history for future-aware NMT since the two modules can be closely incorporated as they focus on different kinds of context. Experimental results on various translation tasks with different scaled datasets, including WMT English <inline-formula xmlns:mml=http://www.w3.org/1998/Math/MathML xmlns:xlink=http://www.w3.org/1999/xlink><tex-math notation=LaTeX>$leftrightarrow$</tex-math></inline-formula> {German, French, Romanian}, show that our proposed approach achieves significant improvements over strong Transformer-based NMT baselines." @default.
- W4311897722 created "2023-01-02" @default.
- W4311897722 creator A5001889488 @default.
- W4311897722 creator A5021667085 @default.
- W4311897722 creator A5031493092 @default.
- W4311897722 creator A5038566155 @default.
- W4311897722 creator A5053204257 @default.
- W4311897722 creator A5077286641 @default.
- W4311897722 date "2023-01-01" @default.
- W4311897722 modified "2023-10-16" @default.
- W4311897722 title "Refining History for Future-Aware Neural Machine Translation" @default.
- W4311897722 cites W1902237438 @default.
- W4311897722 cites W2101105183 @default.
- W4311897722 cites W2124807415 @default.
- W4311897722 cites W2268617045 @default.
- W4311897722 cites W2769298630 @default.
- W4311897722 cites W2803369080 @default.
- W4311897722 cites W2962712961 @default.
- W4311897722 cites W2962717182 @default.
- W4311897722 cites W2962784628 @default.
- W4311897722 cites W2962997665 @default.
- W4311897722 cites W2963091079 @default.
- W4311897722 cites W2963212250 @default.
- W4311897722 cites W2963223306 @default.
- W4311897722 cites W2963232029 @default.
- W4311897722 cites W2963311117 @default.
- W4311897722 cites W2963713328 @default.
- W4311897722 cites W2963714898 @default.
- W4311897722 cites W2963876447 @default.
- W4311897722 cites W2963991316 @default.
- W4311897722 cites W2964345285 @default.
- W4311897722 cites W2964669873 @default.
- W4311897722 cites W2970646865 @default.
- W4311897722 cites W2970849705 @default.
- W4311897722 cites W2979303251 @default.
- W4311897722 cites W3035511085 @default.
- W4311897722 cites W3122073869 @default.
- W4311897722 cites W3173680274 @default.
- W4311897722 cites W3211978535 @default.
- W4311897722 doi "https://doi.org/10.1109/taslp.2022.3226332" @default.
- W4311897722 hasPublicationYear "2023" @default.
- W4311897722 type Work @default.
- W4311897722 citedByCount "0" @default.
- W4311897722 crossrefType "journal-article" @default.
- W4311897722 hasAuthorship W4311897722A5001889488 @default.
- W4311897722 hasAuthorship W4311897722A5021667085 @default.
- W4311897722 hasAuthorship W4311897722A5031493092 @default.
- W4311897722 hasAuthorship W4311897722A5038566155 @default.
- W4311897722 hasAuthorship W4311897722A5053204257 @default.
- W4311897722 hasAuthorship W4311897722A5077286641 @default.
- W4311897722 hasConcept C104317684 @default.
- W4311897722 hasConcept C105580179 @default.
- W4311897722 hasConcept C11413529 @default.
- W4311897722 hasConcept C119857082 @default.
- W4311897722 hasConcept C121332964 @default.
- W4311897722 hasConcept C149364088 @default.
- W4311897722 hasConcept C151730666 @default.
- W4311897722 hasConcept C154945302 @default.
- W4311897722 hasConcept C165801399 @default.
- W4311897722 hasConcept C185592680 @default.
- W4311897722 hasConcept C203005215 @default.
- W4311897722 hasConcept C204321447 @default.
- W4311897722 hasConcept C2524010 @default.
- W4311897722 hasConcept C2776214188 @default.
- W4311897722 hasConcept C2777530160 @default.
- W4311897722 hasConcept C2779343474 @default.
- W4311897722 hasConcept C28490314 @default.
- W4311897722 hasConcept C33923547 @default.
- W4311897722 hasConcept C41008148 @default.
- W4311897722 hasConcept C55493867 @default.
- W4311897722 hasConcept C57273362 @default.
- W4311897722 hasConcept C62520636 @default.
- W4311897722 hasConcept C66322947 @default.
- W4311897722 hasConcept C86803240 @default.
- W4311897722 hasConcept C90805587 @default.
- W4311897722 hasConceptScore W4311897722C104317684 @default.
- W4311897722 hasConceptScore W4311897722C105580179 @default.
- W4311897722 hasConceptScore W4311897722C11413529 @default.
- W4311897722 hasConceptScore W4311897722C119857082 @default.
- W4311897722 hasConceptScore W4311897722C121332964 @default.
- W4311897722 hasConceptScore W4311897722C149364088 @default.
- W4311897722 hasConceptScore W4311897722C151730666 @default.
- W4311897722 hasConceptScore W4311897722C154945302 @default.
- W4311897722 hasConceptScore W4311897722C165801399 @default.
- W4311897722 hasConceptScore W4311897722C185592680 @default.
- W4311897722 hasConceptScore W4311897722C203005215 @default.
- W4311897722 hasConceptScore W4311897722C204321447 @default.
- W4311897722 hasConceptScore W4311897722C2524010 @default.
- W4311897722 hasConceptScore W4311897722C2776214188 @default.
- W4311897722 hasConceptScore W4311897722C2777530160 @default.
- W4311897722 hasConceptScore W4311897722C2779343474 @default.
- W4311897722 hasConceptScore W4311897722C28490314 @default.
- W4311897722 hasConceptScore W4311897722C33923547 @default.
- W4311897722 hasConceptScore W4311897722C41008148 @default.
- W4311897722 hasConceptScore W4311897722C55493867 @default.
- W4311897722 hasConceptScore W4311897722C57273362 @default.
- W4311897722 hasConceptScore W4311897722C62520636 @default.
- W4311897722 hasConceptScore W4311897722C66322947 @default.