FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2995616860> ?p ?o ?g. }

• W2995616860 endingPage "15" @default.
• W2995616860 startingPage "1" @default.
• W2995616860 abstract "Translation into morphologically-rich languages challenges neural machine translation (NMT) models with extremely sparse vocabularies where atomic treatment of surface forms is unrealistic. This problem is typically addressed by either pre-processing words into subword units or performing translation directly at the level of characters. The former is based on word segmentation algorithms optimized using corpus-level statistics with no regard to the translation task. The latter learns directly from translation data but requires rather deep architectures. In this paper, we propose to translate words by modeling word formation through a hierarchical latent variable model which mimics the process of morphological inflection. Our model generates words one character at a time by composing two latent representations: a continuous one, aimed at capturing the lexical semantics, and a set of (approximately) discrete features, aimed at capturing the morphosyntactic function, which are shared among different surface forms. Our model achieves better accuracy in translation into three morphologically-rich languages than conventional open-vocabulary NMT methods, while also demonstrating a better generalization capacity under low to mid-resource settings." @default.
• W2995616860 created "2019-12-26" @default.
• W2995616860 creator A5038456766 @default.
• W2995616860 creator A5074142241 @default.
• W2995616860 creator A5086132343 @default.
• W2995616860 date "2020-04-30" @default.
• W2995616860 modified "2023-09-23" @default.
• W2995616860 title "A Latent Morphology Model for Open-Vocabulary Neural Machine Translation" @default.
• W2995616860 cites W1516111018 @default.
• W2995616860 cites W1902237438 @default.
• W2995616860 cites W1959608418 @default.
• W2995616860 cites W1973137995 @default.
• W2995616860 cites W1994616650 @default.
• W2995616860 cites W2098949613 @default.
• W2995616860 cites W2101105183 @default.
• W2995616860 cites W2102486516 @default.
• W2995616860 cites W2119717200 @default.
• W2995616860 cites W2144600658 @default.
• W2995616860 cites W2184135559 @default.
• W2995616860 cites W2220350356 @default.
• W2995616860 cites W2242818861 @default.
• W2995616860 cites W2250342921 @default.
• W2995616860 cites W2350508517 @default.
• W2995616860 cites W2419539795 @default.
• W2995616860 cites W2525778437 @default.
• W2995616860 cites W2547875792 @default.
• W2995616860 cites W2567691283 @default.
• W2995616860 cites W2579599705 @default.
• W2995616860 cites W2798493043 @default.
• W2995616860 cites W2799074487 @default.
• W2995616860 cites W2896691342 @default.
• W2995616860 cites W2899771611 @default.
• W2995616860 cites W2949227999 @default.
• W2995616860 cites W2952165242 @default.
• W2995616860 cites W2962784628 @default.
• W2995616860 cites W2962897886 @default.
• W2995616860 cites W2963099225 @default.
• W2995616860 cites W2963212250 @default.
• W2995616860 cites W2963631431 @default.
• W2995616860 cites W2963645026 @default.
• W2995616860 cites W2963713328 @default.
• W2995616860 cites W2963828549 @default.
• W2995616860 cites W2963887123 @default.
• W2995616860 cites W2964053711 @default.
• W2995616860 cites W2964121744 @default.
• W2995616860 cites W2964199361 @default.
• W2995616860 cites W2964308564 @default.
• W2995616860 cites W2980969112 @default.
• W2995616860 cites W338621447 @default.
• W2995616860 cites W630532510 @default.
• W2995616860 cites W635530177 @default.
• W2995616860 hasPublicationYear "2020" @default.
• W2995616860 type Work @default.
• W2995616860 sameAs 2995616860 @default.
• W2995616860 citedByCount "2" @default.
• W2995616860 countsByYear W29956168602020 @default.
• W2995616860 countsByYear W29956168602021 @default.
• W2995616860 crossrefType "proceedings-article" @default.
• W2995616860 hasAuthorship W2995616860A5038456766 @default.
• W2995616860 hasAuthorship W2995616860A5074142241 @default.
• W2995616860 hasAuthorship W2995616860A5086132343 @default.
• W2995616860 hasConcept C104317684 @default.
• W2995616860 hasConcept C105580179 @default.
• W2995616860 hasConcept C126706616 @default.
• W2995616860 hasConcept C130597682 @default.
• W2995616860 hasConcept C134306372 @default.
• W2995616860 hasConcept C138885662 @default.
• W2995616860 hasConcept C149364088 @default.
• W2995616860 hasConcept C154945302 @default.
• W2995616860 hasConcept C159403335 @default.
• W2995616860 hasConcept C177148314 @default.
• W2995616860 hasConcept C177264268 @default.
• W2995616860 hasConcept C184337299 @default.
• W2995616860 hasConcept C185592680 @default.
• W2995616860 hasConcept C199360897 @default.
• W2995616860 hasConcept C203005215 @default.
• W2995616860 hasConcept C204321447 @default.
• W2995616860 hasConcept C24687705 @default.
• W2995616860 hasConcept C2777601683 @default.
• W2995616860 hasConcept C33923547 @default.
• W2995616860 hasConcept C41008148 @default.
• W2995616860 hasConcept C41895202 @default.
• W2995616860 hasConcept C55493867 @default.
• W2995616860 hasConcept C90805587 @default.
• W2995616860 hasConcept C98954769 @default.
• W2995616860 hasConceptScore W2995616860C104317684 @default.
• W2995616860 hasConceptScore W2995616860C105580179 @default.
• W2995616860 hasConceptScore W2995616860C126706616 @default.
• W2995616860 hasConceptScore W2995616860C130597682 @default.
• W2995616860 hasConceptScore W2995616860C134306372 @default.
• W2995616860 hasConceptScore W2995616860C138885662 @default.
• W2995616860 hasConceptScore W2995616860C149364088 @default.
• W2995616860 hasConceptScore W2995616860C154945302 @default.
• W2995616860 hasConceptScore W2995616860C159403335 @default.
• W2995616860 hasConceptScore W2995616860C177148314 @default.
• W2995616860 hasConceptScore W2995616860C177264268 @default.
• W2995616860 hasConceptScore W2995616860C184337299 @default.
• W2995616860 hasConceptScore W2995616860C185592680 @default.

• next