论文标题
无监督的机器翻译何时起作用？
When Does Unsupervised Machine Translation Work?
论文作者
论文摘要
尽管据报道无监督的机器翻译（MT）取得了成功，但该领域尚未检查这些方法成功以及它们失败的条件。我们使用不同语言对，不同的域，不同的数据集和正宗的低资源语言对无监督的MT进行了广泛的经验评估。我们发现，当源和目标语料库来自不同域时，性能会迅速恶化，并且随机嵌入初始化会极大地影响下游翻译性能。我们还发现，当源和目标语言使用不同的脚本时，无监督的MT性能会下降，并且在真实的低资源语言对上观察到非常差的性能。我们主张对无监督的MT系统进行广泛的经验评估，以突出失败点并鼓励对最有希望的范式进行继续研究。
Despite the reported success of unsupervised machine translation (MT), the field has yet to examine the conditions under which these methods succeed, and where they fail. We conduct an extensive empirical evaluation of unsupervised MT using dissimilar language pairs, dissimilar domains, diverse datasets, and authentic low-resource languages. We find that performance rapidly deteriorates when source and target corpora are from different domains, and that random word embedding initialization can dramatically affect downstream translation performance. We additionally find that unsupervised MT performance declines when source and target languages use different scripts, and observe very poor performance on authentic low-resource language pairs. We advocate for extensive empirical evaluation of unsupervised MT systems to highlight failure points and encourage continued research on the most promising paradigms.
