Crossing the Threshold: Idiomatic Machine Translation through Retrieval Augmentation and Loss Weighting
This work addresses a specific problem in machine translation for languages like French, Finnish, and Japanese, offering incremental improvements to existing models.
The paper tackled the challenge of translating idiomatic expressions in machine translation by introducing retrieval augmentation and loss weighting techniques, which improved translation accuracy on idiomatic sentences by up to 13% in absolute terms.
Idioms are common in everyday language, but often pose a challenge to translators because their meanings do not follow from the meanings of their parts. Despite significant advances, machine translation systems still struggle to translate idiomatic expressions. We provide a simple characterization of idiomatic translation and related issues. This allows us to conduct a synthetic experiment revealing a tipping point at which transformer-based machine translation models correctly default to idiomatic translations. To expand multilingual resources, we compile a dataset of ~4k natural sentences containing idiomatic expressions in French, Finnish, and Japanese. To improve translation of natural idioms, we introduce two straightforward yet effective techniques: the strategic upweighting of training loss on potentially idiomatic sentences, and using retrieval-augmented models. This not only improves the accuracy of a strong pretrained MT model on idiomatic sentences by up to 13% in absolute accuracy, but also holds potential benefits for non-idiomatic sentences.