Separating Grains from the Chaff: Using Data Filtering to Improve Multilingual Translation for Low-Resourced African Languages
Abstract
We participated in the WMT 2022 Large-Scale Machine Translation Evaluation for the African Languages Shared Task. This work describes our approach, which is based on filtering the given noisy data using a sentence-pair classifier that was built by fine-tuning a pre-trained language model. To train the classifier, we obtain positive samples (i.e. high-quality parallel sentences) from a gold-standard curated dataset and extract negative samples (i.e. low-quality parallel sentences) from automatically aligned parallel data by choosing sentences with low alignment scores. Our final machine translation model was then trained on filtered data, instead of the entire noisy dataset. We empirically validate our approach by evaluating on two common datasets and show that data filtering generally improves overall translation quality, in some cases even significantly.
Keywords
Cite
@article{arxiv.2210.10692,
title = {Separating Grains from the Chaff: Using Data Filtering to Improve Multilingual Translation for Low-Resourced African Languages},
author = {Idris Abdulmumin and Michael Beukman and Jesujoba O. Alabi and Chris Emezue and Everlyn Asiko and Tosin Adewumi and Shamsuddeen Hassan Muhammad and Mofetoluwa Adeyemi and Oreen Yousuf and Sahib Singh and Tajuddeen Rabiu Gwadabe},
journal= {arXiv preprint arXiv:2210.10692},
year = {2022}
}
Comments
Accepted at the Seventh Conference on Machine Translation (WMT22)