Learning Translation Quality Evaluation on Low Resource Languages from Large Language Models
This addresses the challenge of expensive data acquisition for machine translation evaluation in low-resource languages, though it is incremental as it builds on existing learned metrics.
The paper tackles the problem of training learned translation quality evaluation metrics for low-resource languages by distilling knowledge from Large Language Models to create synthetic datasets, improving BLEURT-like model performance without human annotation.
Learned metrics such as BLEURT have in recent years become widely employed to evaluate the quality of machine translation systems. Training such metrics requires data which can be expensive and difficult to acquire, particularly for lower-resource languages. We show how knowledge can be distilled from Large Language Models (LLMs) to improve upon such learned metrics without requiring human annotators, by creating synthetic datasets which can be mixed into existing datasets, requiring only a corpus of text in the target language. We show that the performance of a BLEURT-like model on lower resource languages can be improved in this way.