TULUN: Transparent and Adaptable Low-resource Machine Translation

Fuente: arXiv
Salvato in:
Dettagli Bibliografici
Autori principali: Merx, Raphaël, Suominen, Hanna, Hong, Lois, Thieberger, Nick, Cohn, Trevor, Vylomova, Ekaterina
Natura: Preprint
Pubblicazione: 2025
Soggetti:
Accesso online:
Tags: Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
_version_ 1866916756654653440
author Merx, Raphaël
Suominen, Hanna
Hong, Lois
Thieberger, Nick
Cohn, Trevor
Vylomova, Ekaterina
author_facet Merx, Raphaël
Suominen, Hanna
Hong, Lois
Thieberger, Nick
Cohn, Trevor
Vylomova, Ekaterina
contents Machine translation (MT) systems that support low-resource languages often struggle on specialized domains. While researchers have proposed various techniques for domain adaptation, these approaches typically require model fine-tuning, making them impractical for non-technical users and small organizations. To address this gap, we propose Tulun, a versatile solution for terminology-aware translation, combining neural MT with large language model (LLM)-based post-editing guided by existing glossaries and translation memories. Our open-source web-based platform enables users to easily create, edit, and leverage terminology resources, fostering a collaborative human-machine translation process that respects and incorporates domain expertise while increasing MT accuracy. Evaluations show effectiveness in both real-world and benchmark scenarios: on medical and disaster relief translation tasks for Tetun and Bislama, our system achieves improvements of 16.90-22.41 ChrF++ points over baseline MT systems. Across six low-resource languages on the FLORES dataset, Tulun outperforms both standalone MT and LLM approaches, achieving an average improvement of 2.8 ChrF points over NLLB-54B.
format Preprint
id arxiv_https___arxiv_org_abs_2505_18683
institution arXiv
publishDate 2025
record_format arxiv
spellingShingle TULUN: Transparent and Adaptable Low-resource Machine Translation
Merx, Raphaël
Suominen, Hanna
Hong, Lois
Thieberger, Nick
Cohn, Trevor
Vylomova, Ekaterina
Computation and Language
Machine translation (MT) systems that support low-resource languages often struggle on specialized domains. While researchers have proposed various techniques for domain adaptation, these approaches typically require model fine-tuning, making them impractical for non-technical users and small organizations. To address this gap, we propose Tulun, a versatile solution for terminology-aware translation, combining neural MT with large language model (LLM)-based post-editing guided by existing glossaries and translation memories. Our open-source web-based platform enables users to easily create, edit, and leverage terminology resources, fostering a collaborative human-machine translation process that respects and incorporates domain expertise while increasing MT accuracy. Evaluations show effectiveness in both real-world and benchmark scenarios: on medical and disaster relief translation tasks for Tetun and Bislama, our system achieves improvements of 16.90-22.41 ChrF++ points over baseline MT systems. Across six low-resource languages on the FLORES dataset, Tulun outperforms both standalone MT and LLM approaches, achieving an average improvement of 2.8 ChrF points over NLLB-54B.
title TULUN: Transparent and Adaptable Low-resource Machine Translation
topic Computation and Language
url https://arxiv.org/abs/2505.18683