Semantic Structural Decomposition for Neural Machine Translation

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review


Building on recent advances in semantic parsing and text simplification, we investigate the use of semantic splitting of the source sentence as preprocessing for machine translation. We experiment with a Transformer model and evaluate using large-scale crowd-sourcing experiments. Results show a significant increase in fluency on long sentences on an English-to- French setting with a training corpus of 5M sentence pairs, while retaining comparable adequacy. We also perform a manual analysis which explores the tradeoff between adequacy and fluency in the case where all sentence lengths are considered.
Original languageEnglish
Title of host publicationProceedings of the Ninth Joint Conference on Lexical and Computational Semantics
EditorsIryna Gurevych, Marianna Apidianaki, Manaal Faruqui
Place of PublicationBarcelona, Spain (Online)
PublisherAssociation for Computational Linguistics (ACL)
Number of pages8
ISBN (Electronic)978-1-952148-32-3
StatePublished - Dec 2020
Event9th Joint Conference on Lexical and Computational Semantics - Barcelona, Spain (Online), Barcelona, Spain
Duration: 12 Nov 202013 Nov 2020
Conference number: 9


Conference9th Joint Conference on Lexical and Computational Semantics
Internet address


  • semantic parsing
  • text simplification
  • semantic splitting
  • crowd-sourcing
  • Neural Machine Translation
  • Lexical Semantics
  • Computational Semantics


Dive into the research topics of 'Semantic Structural Decomposition for Neural Machine Translation'. Together they form a unique fingerprint.

Cite this