Matches in SemOpenAlex for { <https://semopenalex.org/work/W2883904829> ?p ?o ?g. }
- W2883904829 endingPage "538" @default.
- W2883904829 startingPage "529" @default.
- W2883904829 abstract "Neural machine translation (NMT) has shown very promising results when there are large amounts of parallel corpora. However, for low resource domains, vanilla NMT cannot give satisfactory performance due to overfitting on the small size of parallel corpora. Two categories of domain adaptation approaches have been proposed for low resource NMT, i.e., adaptation using out-of-domain parallel corpora and in-domain monolingual corpora. In this paper, we conduct a comprehensive empirical comparison of the methods in both categories. For domain adaptation using out-of-domain parallel corpora, we further propose a novel domain adaptation method named mixed fine tuning, which combines two existing methods namely fine tuning and multi domain NMT. For domain adaptation using in-domain monolingual corpora, we compare two existing methods namely language model fusion and synthetic data generation. In addition, we propose a method that combines these two categories. We empirically compare all the methods and discuss their benefits and shortcomings. To the best of our knowledge, this is the first work on a comprehensive empirical comparison of domain adaptation methods for NMT." @default.
- W2883904829 created "2018-08-03" @default.
- W2883904829 creator A5008316265 @default.
- W2883904829 creator A5028836340 @default.
- W2883904829 creator A5077362985 @default.
- W2883904829 date "2018-01-01" @default.
- W2883904829 modified "2023-10-03" @default.
- W2883904829 title "A Comprehensive Empirical Comparison of Domain Adaptation Methods for Neural Machine Translation" @default.
- W2883904829 cites W2146574666 @default.
- W2883904829 cites W2157331557 @default.
- W2883904829 cites W2467834614 @default.
- W2883904829 cites W2515631395 @default.
- W2883904829 cites W2561274697 @default.
- W2883904829 cites W2744813330 @default.
- W2883904829 cites W2756566411 @default.
- W2883904829 cites W2758310181 @default.
- W2883904829 cites W2962784628 @default.
- W2883904829 cites W2963088995 @default.
- W2883904829 cites W2963122608 @default.
- W2883904829 cites W2963216553 @default.
- W2883904829 cites W2963357083 @default.
- W2883904829 cites W2964352358 @default.
- W2883904829 cites W3102516861 @default.
- W2883904829 doi "https://doi.org/10.2197/ipsjjip.26.529" @default.
- W2883904829 hasPublicationYear "2018" @default.
- W2883904829 type Work @default.
- W2883904829 sameAs 2883904829 @default.
- W2883904829 citedByCount "12" @default.
- W2883904829 countsByYear W28839048292018 @default.
- W2883904829 countsByYear W28839048292019 @default.
- W2883904829 countsByYear W28839048292020 @default.
- W2883904829 countsByYear W28839048292021 @default.
- W2883904829 countsByYear W28839048292023 @default.
- W2883904829 crossrefType "journal-article" @default.
- W2883904829 hasAuthorship W2883904829A5008316265 @default.
- W2883904829 hasAuthorship W2883904829A5028836340 @default.
- W2883904829 hasAuthorship W2883904829A5077362985 @default.
- W2883904829 hasBestOaLocation W28839048291 @default.
- W2883904829 hasConcept C104317684 @default.
- W2883904829 hasConcept C105580179 @default.
- W2883904829 hasConcept C105795698 @default.
- W2883904829 hasConcept C119857082 @default.
- W2883904829 hasConcept C120665830 @default.
- W2883904829 hasConcept C120936955 @default.
- W2883904829 hasConcept C121332964 @default.
- W2883904829 hasConcept C134306372 @default.
- W2883904829 hasConcept C139807058 @default.
- W2883904829 hasConcept C149364088 @default.
- W2883904829 hasConcept C154945302 @default.
- W2883904829 hasConcept C185592680 @default.
- W2883904829 hasConcept C203005215 @default.
- W2883904829 hasConcept C204321447 @default.
- W2883904829 hasConcept C206345919 @default.
- W2883904829 hasConcept C22019652 @default.
- W2883904829 hasConcept C2776434776 @default.
- W2883904829 hasConcept C31258907 @default.
- W2883904829 hasConcept C33923547 @default.
- W2883904829 hasConcept C36503486 @default.
- W2883904829 hasConcept C41008148 @default.
- W2883904829 hasConcept C50644808 @default.
- W2883904829 hasConcept C55493867 @default.
- W2883904829 hasConcept C95623464 @default.
- W2883904829 hasConceptScore W2883904829C104317684 @default.
- W2883904829 hasConceptScore W2883904829C105580179 @default.
- W2883904829 hasConceptScore W2883904829C105795698 @default.
- W2883904829 hasConceptScore W2883904829C119857082 @default.
- W2883904829 hasConceptScore W2883904829C120665830 @default.
- W2883904829 hasConceptScore W2883904829C120936955 @default.
- W2883904829 hasConceptScore W2883904829C121332964 @default.
- W2883904829 hasConceptScore W2883904829C134306372 @default.
- W2883904829 hasConceptScore W2883904829C139807058 @default.
- W2883904829 hasConceptScore W2883904829C149364088 @default.
- W2883904829 hasConceptScore W2883904829C154945302 @default.
- W2883904829 hasConceptScore W2883904829C185592680 @default.
- W2883904829 hasConceptScore W2883904829C203005215 @default.
- W2883904829 hasConceptScore W2883904829C204321447 @default.
- W2883904829 hasConceptScore W2883904829C206345919 @default.
- W2883904829 hasConceptScore W2883904829C22019652 @default.
- W2883904829 hasConceptScore W2883904829C2776434776 @default.
- W2883904829 hasConceptScore W2883904829C31258907 @default.
- W2883904829 hasConceptScore W2883904829C33923547 @default.
- W2883904829 hasConceptScore W2883904829C36503486 @default.
- W2883904829 hasConceptScore W2883904829C41008148 @default.
- W2883904829 hasConceptScore W2883904829C50644808 @default.
- W2883904829 hasConceptScore W2883904829C55493867 @default.
- W2883904829 hasConceptScore W2883904829C95623464 @default.
- W2883904829 hasIssue "0" @default.
- W2883904829 hasLocation W28839048291 @default.
- W2883904829 hasOpenAccess W2883904829 @default.
- W2883904829 hasPrimaryLocation W28839048291 @default.
- W2883904829 hasRelatedWork W2140460368 @default.
- W2883904829 hasRelatedWork W2581863816 @default.
- W2883904829 hasRelatedWork W2805394970 @default.
- W2883904829 hasRelatedWork W2887289680 @default.
- W2883904829 hasRelatedWork W2963897095 @default.
- W2883904829 hasRelatedWork W3130032352 @default.
- W2883904829 hasRelatedWork W3200074555 @default.
- W2883904829 hasRelatedWork W4225101459 @default.