Knowledge Distillation for Multilingual Unsupervised Neural Machine Translation

TL;DR

This paper introduces a simple multilingual UNMT framework with knowledge distillation techniques, enabling translation among thirteen languages simultaneously and improving performance across various language pairs, including low-resource and zero-shot scenarios.

Contribution

It presents a novel approach for multilingual UNMT using a single encoder-decoder and introduces two knowledge distillation methods to enhance translation quality.

Findings

Outperforms strong unsupervised baselines across multiple language pairs.

Achieves promising zero-shot translation results between non-English languages.

Alleviates poor performance in low-resource language pairs.

Abstract

Unsupervised neural machine translation (UNMT) has recently achieved remarkable results for several language pairs. However, it can only translate between a single language pair and cannot produce translation results for multiple language pairs at the same time. That is, research on multilingual UNMT has been limited. In this paper, we empirically introduce a simple method to translate between thirteen languages using a single encoder and a single decoder, making use of multilingual data to improve UNMT for all language pairs. On the basis of the empirical findings, we propose two knowledge distillation methods to further enhance multilingual UNMT performance. Our experiments on a dataset with English translated to and from twelve other languages (including three language families and six language branches) show remarkable results, surpassing strong unsupervised individual baselines while achieving promising performance between non-English language pairs in zero-shot translation scenarios and alleviating poor performance in low-resource language pairs.