Self-Training for Unsupervised Neural Machine Translation in Unbalanced Training Data Scenarios
This addresses a practical issue for low-resource language translation, but it is incremental as it builds on existing UNMT methods.
The paper tackles the problem of unsupervised neural machine translation (UNMT) performing poorly in unbalanced training data scenarios, where one language lacks adequate monolingual corpora, and proposes self-training mechanisms that substantially outperform conventional UNMT systems in experiments on several language pairs.
Unsupervised neural machine translation (UNMT) that relies solely on massive monolingual corpora has achieved remarkable results in several translation tasks. However, in real-world scenarios, massive monolingual corpora do not exist for some extremely low-resource languages such as Estonian, and UNMT systems usually perform poorly when there is not adequate training corpus for one language. In this paper, we first define and analyze the unbalanced training data scenario for UNMT. Based on this scenario, we propose UNMT self-training mechanisms to train a robust UNMT system and improve its performance in this case. Experimental results on several language pairs show that the proposed methods substantially outperform conventional UNMT systems.