Matches in SemOpenAlex for { <https://semopenalex.org/work/W2892036039> ?p ?o ?g. }
Showing items 1 to 70 of
70
with 100 items per page.
- W2892036039 abstract "Although the Transformer translation model (Vaswani et al., 2017) has achieved state-of-the-art performance in a variety of translation tasks, how to use document-level context to deal with discourse phenomena problematic for Transformer still remains a challenge. In this work, we extend the Transformer model with a new context encoder to represent document-level context, which is then incorporated into the original encoder and decoder. As large-scale document-level parallel corpora are usually not available, we introduce a two-step training method to take full advantage of abundant sentence-level parallel corpora and limited document-level parallel corpora. Experiments on the NIST Chinese-English datasets and the IWSLT French-English datasets show that our approach improves over Transformer significantly." @default.
- W2892036039 created "2018-09-27" @default.
- W2892036039 creator A5012727269 @default.
- W2892036039 creator A5019899786 @default.
- W2892036039 creator A5024810143 @default.
- W2892036039 creator A5046448314 @default.
- W2892036039 creator A5059266417 @default.
- W2892036039 creator A5086000078 @default.
- W2892036039 creator A5091587552 @default.
- W2892036039 date "2018-10-08" @default.
- W2892036039 modified "2023-09-27" @default.
- W2892036039 title "Improving the Transformer Translation Model with Document-Level Context" @default.
- W2892036039 doi "https://doi.org/10.48550/arxiv.1810.03581" @default.
- W2892036039 hasPublicationYear "2018" @default.
- W2892036039 type Work @default.
- W2892036039 sameAs 2892036039 @default.
- W2892036039 citedByCount "3" @default.
- W2892036039 countsByYear W28920360392019 @default.
- W2892036039 countsByYear W28920360392020 @default.
- W2892036039 countsByYear W28920360392021 @default.
- W2892036039 crossrefType "posted-content" @default.
- W2892036039 hasAuthorship W2892036039A5012727269 @default.
- W2892036039 hasAuthorship W2892036039A5019899786 @default.
- W2892036039 hasAuthorship W2892036039A5024810143 @default.
- W2892036039 hasAuthorship W2892036039A5046448314 @default.
- W2892036039 hasAuthorship W2892036039A5059266417 @default.
- W2892036039 hasAuthorship W2892036039A5086000078 @default.
- W2892036039 hasAuthorship W2892036039A5091587552 @default.
- W2892036039 hasBestOaLocation W28920360391 @default.
- W2892036039 hasConcept C111219384 @default.
- W2892036039 hasConcept C111919701 @default.
- W2892036039 hasConcept C118505674 @default.
- W2892036039 hasConcept C119599485 @default.
- W2892036039 hasConcept C127413603 @default.
- W2892036039 hasConcept C154945302 @default.
- W2892036039 hasConcept C165801399 @default.
- W2892036039 hasConcept C203005215 @default.
- W2892036039 hasConcept C204321447 @default.
- W2892036039 hasConcept C2777530160 @default.
- W2892036039 hasConcept C41008148 @default.
- W2892036039 hasConcept C66322947 @default.
- W2892036039 hasConceptScore W2892036039C111219384 @default.
- W2892036039 hasConceptScore W2892036039C111919701 @default.
- W2892036039 hasConceptScore W2892036039C118505674 @default.
- W2892036039 hasConceptScore W2892036039C119599485 @default.
- W2892036039 hasConceptScore W2892036039C127413603 @default.
- W2892036039 hasConceptScore W2892036039C154945302 @default.
- W2892036039 hasConceptScore W2892036039C165801399 @default.
- W2892036039 hasConceptScore W2892036039C203005215 @default.
- W2892036039 hasConceptScore W2892036039C204321447 @default.
- W2892036039 hasConceptScore W2892036039C2777530160 @default.
- W2892036039 hasConceptScore W2892036039C41008148 @default.
- W2892036039 hasConceptScore W2892036039C66322947 @default.
- W2892036039 hasLocation W28920360391 @default.
- W2892036039 hasOpenAccess W2892036039 @default.
- W2892036039 hasPrimaryLocation W28920360391 @default.
- W2892036039 hasRelatedWork W2558155472 @default.
- W2892036039 hasRelatedWork W2753086417 @default.
- W2892036039 hasRelatedWork W2808592175 @default.
- W2892036039 hasRelatedWork W2956301977 @default.
- W2892036039 hasRelatedWork W2961915345 @default.
- W2892036039 hasRelatedWork W2963693355 @default.
- W2892036039 hasRelatedWork W3038033387 @default.
- W2892036039 hasRelatedWork W3171376244 @default.
- W2892036039 hasRelatedWork W4229016505 @default.
- W2892036039 hasRelatedWork W4280626242 @default.
- W2892036039 isParatext "false" @default.
- W2892036039 isRetracted "false" @default.
- W2892036039 magId "2892036039" @default.
- W2892036039 workType "article" @default.