Volctrans Parallel Corpus Filtering System for WMT 2020
This work addresses the challenge of parallel corpus filtering for low-resource languages, which is incremental as it builds on existing methods like word alignment and XLM-based scoring.
The paper tackles the problem of aligning and scoring parallel sentence pairs for low-resource language pairs in the WMT20 shared task, achieving the highest performance among submissions with improvements of 3.x/2.x and 2.x/2.x for km-en and ps-en under From Scratch/Fine-Tune conditions.
In this paper, we describe our submissions to the WMT20 shared task on parallel corpus filtering and alignment for low-resource conditions. The task requires the participants to align potential parallel sentence pairs out of the given document pairs, and score them so that low-quality pairs can be filtered. Our system, Volctrans, is made of two modules, i.e., a mining module and a scoring module. Based on the word alignment model, the mining module adopts an iterative mining strategy to extract latent parallel sentences. In the scoring module, an XLM-based scorer provides scores, followed by reranking mechanisms and ensemble. Our submissions outperform the baseline by 3.x/2.x and 2.x/2.x for km-en and ps-en on From Scratch/Fine-Tune conditions, which is the highest among all submissions.