Crossing the Threshold: Idiomatic Machine Translation through Retrieval Augmentation and Loss Weighting
Year: 2023
Type: conference-paper
Abstract: 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 retrievalaugmented 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.1
Authors: Emmy Liu, Aditi Chaudhary, Graham Neubig
Institution: Carnegie Mellon University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
Open Access status: hybrid
OpenAlex ID: W4389524191