Matches in SemOpenAlex for { <https://semopenalex.org/work/W2546861570> ?p ?o ?g. }
- W2546861570 abstract "N-best reranking in statistical machine translation (SMT) aims to rescore the possible translation hypotheses such that the best hypothesis appears on the top of the list. The current models either use the standard features (a.k.a model features) that are derived from the SMT decoder or the manually crafted features for reranking. However, in practice, the results are disappointing which had only very small effect on the overall performance. In this paper, we investigate using additional semantic and syntactic feature representations to the reranking framework, with the goal to better capture the differences between best hypothesis and others. These representations are the sentence embeddings that are learned using the recursive autoencoder (RAE). The proposed features are extensively evaluated with various reranking algorithms on WMT2015 French-to-English translation data. The empirical results reveal that the proposed reranking model is able to yield an additional improvement of 1.23 BLEU points over the baseline SMT system." @default.
- W2546861570 created "2016-11-11" @default.
- W2546861570 creator A5010462322 @default.
- W2546861570 creator A5025832925 @default.
- W2546861570 creator A5047428567 @default.
- W2546861570 date "2016-07-01" @default.
- W2546861570 modified "2023-09-26" @default.
- W2546861570 title "Exploiting rich feature representation for SMT N-best reranking" @default.
- W2546861570 cites W1033981595 @default.
- W2546861570 cites W153914110 @default.
- W2546861570 cites W1678356000 @default.
- W2546861570 cites W179314280 @default.
- W2546861570 cites W1966914257 @default.
- W2546861570 cites W1974259881 @default.
- W2546861570 cites W1990589796 @default.
- W2546861570 cites W1991995555 @default.
- W2546861570 cites W2002658919 @default.
- W2546861570 cites W2005126631 @default.
- W2546861570 cites W2007592492 @default.
- W2546861570 cites W2034790838 @default.
- W2546861570 cites W2065180801 @default.
- W2546861570 cites W2107085272 @default.
- W2546861570 cites W2107890099 @default.
- W2546861570 cites W2118960083 @default.
- W2546861570 cites W2124807415 @default.
- W2546861570 cites W2143331230 @default.
- W2546861570 cites W2146574666 @default.
- W2546861570 cites W2154124206 @default.
- W2546861570 cites W2160001241 @default.
- W2546861570 cites W2165510491 @default.
- W2546861570 cites W2250508490 @default.
- W2546861570 cites W2250876691 @default.
- W2546861570 cites W2251362467 @default.
- W2546861570 cites W2251690405 @default.
- W2546861570 cites W2252213522 @default.
- W2546861570 cites W2252267750 @default.
- W2546861570 cites W3171974722 @default.
- W2546861570 cites W71795751 @default.
- W2546861570 doi "https://doi.org/10.1109/icwapr.2016.7731631" @default.
- W2546861570 hasPublicationYear "2016" @default.
- W2546861570 type Work @default.
- W2546861570 sameAs 2546861570 @default.
- W2546861570 citedByCount "1" @default.
- W2546861570 countsByYear W25468615702019 @default.
- W2546861570 crossrefType "proceedings-article" @default.
- W2546861570 hasAuthorship W2546861570A5010462322 @default.
- W2546861570 hasAuthorship W2546861570A5025832925 @default.
- W2546861570 hasAuthorship W2546861570A5047428567 @default.
- W2546861570 hasConcept C101738243 @default.
- W2546861570 hasConcept C104317684 @default.
- W2546861570 hasConcept C105580179 @default.
- W2546861570 hasConcept C119857082 @default.
- W2546861570 hasConcept C138885662 @default.
- W2546861570 hasConcept C149364088 @default.
- W2546861570 hasConcept C154945302 @default.
- W2546861570 hasConcept C17744445 @default.
- W2546861570 hasConcept C185592680 @default.
- W2546861570 hasConcept C199539241 @default.
- W2546861570 hasConcept C203005215 @default.
- W2546861570 hasConcept C204321447 @default.
- W2546861570 hasConcept C2776359362 @default.
- W2546861570 hasConcept C2776401178 @default.
- W2546861570 hasConcept C2777530160 @default.
- W2546861570 hasConcept C41008148 @default.
- W2546861570 hasConcept C41895202 @default.
- W2546861570 hasConcept C50644808 @default.
- W2546861570 hasConcept C55493867 @default.
- W2546861570 hasConcept C94625758 @default.
- W2546861570 hasConceptScore W2546861570C101738243 @default.
- W2546861570 hasConceptScore W2546861570C104317684 @default.
- W2546861570 hasConceptScore W2546861570C105580179 @default.
- W2546861570 hasConceptScore W2546861570C119857082 @default.
- W2546861570 hasConceptScore W2546861570C138885662 @default.
- W2546861570 hasConceptScore W2546861570C149364088 @default.
- W2546861570 hasConceptScore W2546861570C154945302 @default.
- W2546861570 hasConceptScore W2546861570C17744445 @default.
- W2546861570 hasConceptScore W2546861570C185592680 @default.
- W2546861570 hasConceptScore W2546861570C199539241 @default.
- W2546861570 hasConceptScore W2546861570C203005215 @default.
- W2546861570 hasConceptScore W2546861570C204321447 @default.
- W2546861570 hasConceptScore W2546861570C2776359362 @default.
- W2546861570 hasConceptScore W2546861570C2776401178 @default.
- W2546861570 hasConceptScore W2546861570C2777530160 @default.
- W2546861570 hasConceptScore W2546861570C41008148 @default.
- W2546861570 hasConceptScore W2546861570C41895202 @default.
- W2546861570 hasConceptScore W2546861570C50644808 @default.
- W2546861570 hasConceptScore W2546861570C55493867 @default.
- W2546861570 hasConceptScore W2546861570C94625758 @default.
- W2546861570 hasLocation W25468615701 @default.
- W2546861570 hasOpenAccess W2546861570 @default.
- W2546861570 hasPrimaryLocation W25468615701 @default.
- W2546861570 hasRelatedWork W1531482446 @default.
- W2546861570 hasRelatedWork W1541481035 @default.
- W2546861570 hasRelatedWork W2086464933 @default.
- W2546861570 hasRelatedWork W2088074175 @default.
- W2546861570 hasRelatedWork W2233968761 @default.
- W2546861570 hasRelatedWork W2358509285 @default.
- W2546861570 hasRelatedWork W2538151745 @default.
- W2546861570 hasRelatedWork W2575172791 @default.
- W2546861570 hasRelatedWork W2592831379 @default.