Fine-Tuned Machine Translation Metrics Struggle in Unseen Domains
This highlights a critical limitation for MT evaluation in specialized fields like biomedicine, indicating incremental insights into metric generalization.
The study investigated the robustness of fine-tuned machine translation metrics to domain shifts, finding that they experience a substantial performance drop in unseen domains compared to surface-form and pre-trained metrics.
We introduce a new, extensive multidimensional quality metrics (MQM) annotated dataset covering 11 language pairs in the biomedical domain. We use this dataset to investigate whether machine translation (MT) metrics which are fine-tuned on human-generated MT quality judgements are robust to domain shifts between training and inference. We find that fine-tuned metrics exhibit a substantial performance drop in the unseen domain scenario relative to metrics that rely on the surface form, as well as pre-trained metrics which are not fine-tuned on MT quality judgments.