Unsupervised Translation of German--Lower Sorbian: Exploring Training and Novel Transfer Methods on a Low-Resource Language
It addresses translation for a low-resource language, but the methods are incremental adaptations of existing unsupervised techniques.
This paper tackles unsupervised machine translation from German to low-resource Lower Sorbian by introducing a novel vocabulary initialization method and optimizing back-translation order, resulting in BLEU score improvements of up to 4.0 points and achieving top rankings in a WMT 2021 task.
This paper describes the methods behind the systems submitted by the University of Groningen for the WMT 2021 Unsupervised Machine Translation task for German--Lower Sorbian (DE--DSB): a high-resource language to a low-resource one. Our system uses a transformer encoder-decoder architecture in which we make three changes to the standard training procedure. First, our training focuses on two languages at a time, contrasting with a wealth of research on multilingual systems. Second, we introduce a novel method for initializing the vocabulary of an unseen language, achieving improvements of 3.2 BLEU for DE$\rightarrow$DSB and 4.0 BLEU for DSB$\rightarrow$DE. Lastly, we experiment with the order in which offline and online back-translation are used to train an unsupervised system, finding that using online back-translation first works better for DE$\rightarrow$DSB by 2.76 BLEU. Our submissions ranked first (tied with another team) for DSB$\rightarrow$DE and third for DE$\rightarrow$DSB.