Multilingual Communication in the Asylum Context: Evaluating LLM-Based Machine Translation with Fuzzy Match Augmentation and Adaptive NMT across Resource Conditions under Low-Data Constraints
Year: 2026
Type: conference-paper
Abstract: Effective communication in asylum reception settings requires reliable machine translation (MT) across many languages, including low-resource ones. Using data from the MaTIAS project (Machine Translation to Inform Asylum Seekers), we compare retrieval-augmented LLM translation with adaptive Neural MT across 14 target languages with varying resource levels. Working with a very small translation memory of only 358 sentences, we evaluate fuzzy-match (FM) augmentation as an in-context learning strategy for open-source and commercial LLMs and benchmark these against ModernMT with and without domain adaptation. In the LLM setting, FM-based example selection consistently outperforms random selection and zero-shot prompting, with the largest gains for low-resource languages. Adaptive NMT retains an overall advantage, although Gemini Pro approaches its performance and outperforms it on 6 of 14 languages, highlighting a trade-off between translation quality and data sovereignty in privacy-sensitive contexts. These findings show that FM augmentation remains effective under severe data constraints and emphasise the importance of language-specific evaluation in multilingual MT.
Authors: Moerman, Thomas, Tezcan, Arda, Macken, Lieve
Institution: Ghent University
Language: English
Subfield: Artificial Intelligence
Field: Computer Science
Domain: Physical Sciences
SDG: Reduced inequalities
Open Access status: green
OpenAlex ID: W7204741067
FWCI:
Cites:
Cited by:
Related to: