Improved Data Augmentation for Translation Suggestion
This work addresses data scarcity for translation suggestion systems, which is an incremental improvement in a specific domain.
The paper tackles the problem of limited supervised data for translation suggestion models by introducing synthetic data construction strategies and a multi-phase pre-training approach, achieving second and third place rankings in the WMT'22 shared task for English-German and English-Chinese bidirectional tasks.
Translation suggestion (TS) models are used to automatically provide alternative suggestions for incorrect spans in sentences generated by machine translation. This paper introduces the system used in our submission to the WMT'22 Translation Suggestion shared task. Our system is based on the ensemble of different translation architectures, including Transformer, SA-Transformer, and DynamicConv. We use three strategies to construct synthetic data from parallel corpora to compensate for the lack of supervised data. In addition, we introduce a multi-phase pre-training strategy, adding an additional pre-training phase with in-domain data. We rank second and third on the English-German and English-Chinese bidirectional tasks, respectively.