Better Quality Estimation for Low Resource Corpus Mining
This addresses the robustness issue in QE for low-resource corpus mining, enabling more efficient data usage in machine translation, though it is incremental as it builds on existing QE and PCM techniques.
The paper tackled the problem of Quality Estimation (QE) models performing poorly in Parallel Corpus Mining (PCM) due to lack of robustness to out-of-domain examples, and proposed a method combining multitask training, data augmentation, and contrastive learning that increased PCM accuracy by over 0.80 using only 7K parallel sentences, matching SOTA methods that use millions of sentences.
Quality Estimation (QE) models have the potential to change how we evaluate and maybe even train machine translation models. However, these models still lack the robustness to achieve general adoption. We show that State-of-the-art QE models, when tested in a Parallel Corpus Mining (PCM) setting, perform unexpectedly bad due to a lack of robustness to out-of-domain examples. We propose a combination of multitask training, data augmentation and contrastive learning to achieve better and more robust QE performance. We show that our method improves QE performance significantly in the MLQE challenge and the robustness of QE models when tested in the Parallel Corpus Mining setup. We increase the accuracy in PCM by more than 0.80, making it on par with state-of-the-art PCM methods that use millions of sentence pairs to train their models. In comparison, we use a thousand times less data, 7K parallel sentences in total, and propose a novel low resource PCM method.