FRINK Linked Data Fragments server

Matches in SemOpenAlex for { <https://semopenalex.org/work/W2994702971> ?p ?o ?g. }

• W2994702971 abstract "End-to-end neural approaches have achieved state-of-the-art performance in many natural language processing (NLP) tasks. Yet, they often lack transparency of the underlying decision-making process, hindering error analysis and certain model improvements. In this work, we revisit the binary linearization approach to surface realization, which exhibits more interpretable behavior, but was falling short in terms of prediction accuracy. We show how enriching the training data to better capture word order constraints almost doubles the performance of the system. We further demonstrate that encoding both local and global prediction contexts yields another considerable performance boost. With the proposed modifications, the system which ranked low in the latest shared task on multilingual surface realization now achieves best results in five out of ten languages, while being on par with the state-of-the-art approaches in others." @default.
• W2994702971 created "2019-12-26" @default.
• W2994702971 creator A5027062608 @default.
• W2994702971 creator A5027450194 @default.
• W2994702971 creator A5033006575 @default.
• W2994702971 creator A5087459652 @default.
• W2994702971 date "2019-01-01" @default.
• W2994702971 modified "2023-10-16" @default.
• W2994702971 title "Revisiting the Binary Linearization Technique for Surface Realization" @default.
• W2994702971 cites W1493882948 @default.
• W2994702971 cites W1627132856 @default.
• W2994702971 cites W199274192 @default.
• W2994702971 cites W2078861931 @default.
• W2994702971 cites W2101105183 @default.
• W2994702971 cites W2125417976 @default.
• W2994702971 cites W2130942839 @default.
• W2994702971 cites W2157331557 @default.
• W2994702971 cites W2296207831 @default.
• W2994702971 cites W2468120018 @default.
• W2994702971 cites W2625720409 @default.
• W2994702971 cites W2786660442 @default.
• W2994702971 cites W2787644385 @default.
• W2994702971 cites W2798552002 @default.
• W2994702971 cites W2898454195 @default.
• W2994702971 cites W2899771611 @default.
• W2994702971 cites W2902149415 @default.
• W2994702971 cites W2921512818 @default.
• W2994702971 cites W2953961604 @default.
• W2994702971 cites W2962950136 @default.
• W2994702971 cites W2963374482 @default.
• W2994702971 cites W2963506925 @default.
• W2994702971 cites W2963840929 @default.
• W2994702971 cites W2964116568 @default.
• W2994702971 cites W2964121744 @default.
• W2994702971 cites W2964308564 @default.
• W2994702971 doi "https://doi.org/10.18653/v1/w19-8635" @default.
• W2994702971 hasPublicationYear "2019" @default.
• W2994702971 type Work @default.
• W2994702971 sameAs 2994702971 @default.
• W2994702971 citedByCount "1" @default.
• W2994702971 countsByYear W29947029712021 @default.
• W2994702971 crossrefType "proceedings-article" @default.
• W2994702971 hasAuthorship W2994702971A5027062608 @default.
• W2994702971 hasAuthorship W2994702971A5027450194 @default.
• W2994702971 hasAuthorship W2994702971A5033006575 @default.
• W2994702971 hasAuthorship W2994702971A5087459652 @default.
• W2994702971 hasBestOaLocation W29947029711 @default.
• W2994702971 hasConcept C105795698 @default.
• W2994702971 hasConcept C11210021 @default.
• W2994702971 hasConcept C11413529 @default.
• W2994702971 hasConcept C119857082 @default.
• W2994702971 hasConcept C121332964 @default.
• W2994702971 hasConcept C125411270 @default.
• W2994702971 hasConcept C127413603 @default.
• W2994702971 hasConcept C154945302 @default.
• W2994702971 hasConcept C158622935 @default.
• W2994702971 hasConcept C199360897 @default.
• W2994702971 hasConcept C201995342 @default.
• W2994702971 hasConcept C204321447 @default.
• W2994702971 hasConcept C2524010 @default.
• W2994702971 hasConcept C2780233690 @default.
• W2994702971 hasConcept C2780451532 @default.
• W2994702971 hasConcept C2781089630 @default.
• W2994702971 hasConcept C33923547 @default.
• W2994702971 hasConcept C38652104 @default.
• W2994702971 hasConcept C41008148 @default.
• W2994702971 hasConcept C48103436 @default.
• W2994702971 hasConcept C48372109 @default.
• W2994702971 hasConcept C62520636 @default.
• W2994702971 hasConcept C70777604 @default.
• W2994702971 hasConcept C90805587 @default.
• W2994702971 hasConcept C94375191 @default.
• W2994702971 hasConcept C98045186 @default.
• W2994702971 hasConceptScore W2994702971C105795698 @default.
• W2994702971 hasConceptScore W2994702971C11210021 @default.
• W2994702971 hasConceptScore W2994702971C11413529 @default.
• W2994702971 hasConceptScore W2994702971C119857082 @default.
• W2994702971 hasConceptScore W2994702971C121332964 @default.
• W2994702971 hasConceptScore W2994702971C125411270 @default.
• W2994702971 hasConceptScore W2994702971C127413603 @default.
• W2994702971 hasConceptScore W2994702971C154945302 @default.
• W2994702971 hasConceptScore W2994702971C158622935 @default.
• W2994702971 hasConceptScore W2994702971C199360897 @default.
• W2994702971 hasConceptScore W2994702971C201995342 @default.
• W2994702971 hasConceptScore W2994702971C204321447 @default.
• W2994702971 hasConceptScore W2994702971C2524010 @default.
• W2994702971 hasConceptScore W2994702971C2780233690 @default.
• W2994702971 hasConceptScore W2994702971C2780451532 @default.
• W2994702971 hasConceptScore W2994702971C2781089630 @default.
• W2994702971 hasConceptScore W2994702971C33923547 @default.
• W2994702971 hasConceptScore W2994702971C38652104 @default.
• W2994702971 hasConceptScore W2994702971C41008148 @default.
• W2994702971 hasConceptScore W2994702971C48103436 @default.
• W2994702971 hasConceptScore W2994702971C48372109 @default.
• W2994702971 hasConceptScore W2994702971C62520636 @default.
• W2994702971 hasConceptScore W2994702971C70777604 @default.
• W2994702971 hasConceptScore W2994702971C90805587 @default.
• W2994702971 hasConceptScore W2994702971C94375191 @default.
• W2994702971 hasConceptScore W2994702971C98045186 @default.
• W2994702971 hasLocation W29947029711 @default.

• next