Unsupervised Machine Translation Using Monolingual Corpora Only
Guillaume Lample ††\dagger ‡‡\ddagger , Alexis Conneau ††\dagger , Ludovic Denoyer ‡‡\ddagger , Marc’Aurelio Ranzato ††\dagger ††\dagger Facebook AI Research, ‡‡\ddagger Sorbonne Universités, UPMC Univ Paris 06, LIP6 UMR 7606, CNRS
Abstract
Machine translation has recently achieved impressive performance thanks to recent advances in deep learning and the availability of large-scale parallel corpora. There have been numerous attempts to extend these successes to low-resource language pairs, yet requiring tens of thousands of parallel sentences. In this work, we take this research direction to the extreme and investigate whether it is possible to learn to translate even without any parallel data. We propose a model that takes sentences from monolingual corpora in two different languages and maps them into the same latent space. By learning to reconstruct in both languages from this shared feature space, the model effectively learns to translate without using any labeled data. We demonstrate our model on two widely used datasets and two language pairs, reporting BLEU scores of 32.8 and 15.1 on the Multi30k and WMT English-French datasets, without using even a single parallel sentence at training time.
中文速览
不依赖任何平行语料、仅靠两种语言各自的单语数据来训练机器翻译系统，这正是本文要攻克的难题。研究者设计了一个共享编解码器框架：先用无监督方式构建跨语言词典完成逐词初始化，再通过去噪自编码（denoising auto-encoding）迫使模型学会在噪声干扰下重建原句，同时利用回译（back-translation）做跨语言重建训练，并加入对抗正则化让两种语言的句子表示落入同一隐空间，三重目标迭代优化，最终让模型在完全无监督的条件下学会翻译。实验在 Multi30k 和 WMT 英法数据集上分别取得了 32.8 和 15.1 的 BLEU 分数，翻译质量相当于用 10 万句平行语料训练的有监督系统。这项工作证明了零平行数据机器翻译的可行性，为资源匮乏语言对的翻译研究提供了全新的技术路线和性能基准。
原文 arXiv:1711.00043;中英对照 + 大白话阅读 https://aha.fim.ai/paper/1711.00043v2