Bridging the Data Gap between Training and Inference for Unsupervised Neural Machine Translation
This addresses a specific bottleneck in UNMT by improving translation accuracy, though it is incremental as it builds on existing back-translation methods.
The paper tackled the performance gap in Unsupervised Neural Machine Translation caused by source data discrepancies between training and inference, and proposed an online self-training approach that outperformed strong baselines like XLM and MASS on multiple language pairs.
Back-translation is a critical component of Unsupervised Neural Machine Translation (UNMT), which generates pseudo parallel data from target monolingual data. A UNMT model is trained on the pseudo parallel data with translated source, and translates natural source sentences in inference. The source discrepancy between training and inference hinders the translation performance of UNMT models. By carefully designing experiments, we identify two representative characteristics of the data gap in source: (1) style gap (i.e., translated vs. natural text style) that leads to poor generalization capability; (2) content gap that induces the model to produce hallucination content biased towards the target language. To narrow the data gap, we propose an online self-training approach, which simultaneously uses the pseudo parallel data {natural source, translated target} to mimic the inference scenario. Experimental results on several widely-used language pairs show that our approach outperforms two strong baselines (XLM and MASS) by remedying the style and content gaps.