Skip to Main content Skip to Navigation
Conference papers

One Source, Two Targets: Challenges and Rewards of Dual Decoding

Jitao Xu 1 François Yvon 1
1 TLP - Traitement du Langage Parlé
LISN - Laboratoire Interdisciplinaire des Sciences du Numérique, STL - Sciences et Technologies des Langues
Abstract : Machine translation is generally understood as generating one target text from an input source document. In this paper, we consider a stronger requirement: to jointly generate two texts so that each output side effectively depends on the other. As we discuss, such a device serves several practical purposes, from multi-target machine translation to the generation of controlled variations of the target text. We present an analysis of possible implementations of dual decoding, and experiment with four applications. Viewing the problem from multiple angles allows us to better highlight the challenges of dual decoding and to also thoroughly analyze the benefits of generating matched, rather than independent, translations.
Complete list of metadata
Contributor : Jitao Xu Connect in order to contact the contributor
Submitted on : Wednesday, September 15, 2021 - 3:47:23 PM
Last modification on : Tuesday, January 4, 2022 - 6:04:57 AM
Long-term archiving on: : Thursday, December 16, 2021 - 7:06:31 PM


Files produced by the author(s)


  • HAL Id : hal-03345478, version 1


Jitao Xu, François Yvon. One Source, Two Targets: Challenges and Rewards of Dual Decoding. Conference on Empirical Methods in Natural Language Processing, Nov 2021, Online and Punta Cana, Dominican Republic. ⟨hal-03345478⟩



Les métriques sont temporairement indisponibles