AI Paper Atlas

Research map / Label-efficient and robust learning / Recognition and long-tail

Non-Fluent Synthetic Target-Language Data Improve Neural Machine Translation

Víctor M. Sánchez-Cartagena, Miquel Esplà-Gomis, Juan Antonio Pérez-Ortiz, Felipe Sánchez-Martínez

TPAMI 2023 · Label-efficient and robust learning › Recognition and long-tail · approach: Multilingual language representation pretraining · 8 citations (OpenAlex)

Read the paperCheck your idea against 137,217 papers

AI-extracted. The measured effects below were read by AI from the paper's full text (no summary yet). Numbers and claims can be wrong or missing context, so check the paper before you cite or build on it.

Measured effects reported in the paper

What was changedWhereMeasuredEffect
add MaTiLDA (reverse+swap+replace) synthetic non-fluent data to the baselineTransformer base · low-resource in-domain, 10 directionsBLEU: 1.1 (hsb--de) to 4.1 (de--rm) BLEU points, 2.1 on averageImproved
use MaTiLDA instead of the best reference DA system (SeqMix etc.)Transformer base · low-resource in-domainBLEU: 0.4 to 2.5 BLEU points; 1.1 on averageImproved
add the replace transformation alone (ablation)MaTiLDA · de-rm low-resourceBLEU: 49.3 -> 53.3Improved
add the source (copy source sentence) transformation alone (ablation)MaTiLDA · hsb-de low-resourceBLEU: 49.1 -> 48.4Hurt
add the mono transformation alone (ablation)MaTiLDA · en-de low-resourceBLEU: 25.4 -> 25.2Hurt
combine reverse+swap+replace instead of replace alone (ablation)MaTiLDA · rm-de low-resourceBLEU: 52.7 -> 54.1Improved
combine transformations (rev+sw+repl) instead of single swap (ablation)MaTiLDA · hsb-de low-resourceBLEU: 50.2 -> 49.7Hurt
add MaTiLDA on top of back-translationback-translation baseline · de-en low-resource with back-translated dataBLEU: 31.4 -> 32.8Improved
add MaTiLDA on top of back-translationback-translation baseline · rm-de low-resource with back-translated dataBLEU: 49.1 -> 51.3Improved
add MaTiLDA (swap+unk+replace) in high-resource conditionsTransformer base · en-de high-resourceBLEU: 24.3 -> 25.8Improved
add MaTiLDA (swap+unk+replace) in high-resource conditionsTransformer base · en-ro high-resourceBLEU: 23.3 -> 24.3Improved
add MaTiLDA for out-of-domain translation (low-resource)Transformer base · de-en IT out-of-domain low-resourceBLEU: 4.5 -> 13.3Improved

Closest papers by meaning

Writing a paper in this area? scime draft is a free LaTeX editor with AI included, real-time co-authors and a check of your references and venue rules before you submit. It also imports OpenAI Prism projects.

Open scime draft

Get the next scime Dispatch and a monthly field update

Free. We send a confirmation link first; every email has a one-click unsubscribe. Privacy