Guardado en:
Detalles Bibliográficos
Autores principales: Ranathungaa, Surangika, Nayak, Shravan, Huang, Shih-Ting Cindy, Mao, Yanke, Su, Tong, Chan, Yun-Hsiang Ray, Yuan, Songchen, Rinaldi, Anthony, Lee, Annie En-Shiun
Formato: Preprint
Publicado: 2024
Materias:
Acceso en línea:https://arxiv.org/abs/2412.19522
Etiquetas: Agregar Etiqueta
Sin Etiquetas, Sea el primero en etiquetar este registro!
_version_ 1866917361874894848
author Ranathungaa, Surangika
Nayak, Shravan
Huang, Shih-Ting Cindy
Mao, Yanke
Su, Tong
Chan, Yun-Hsiang Ray
Yuan, Songchen
Rinaldi, Anthony
Lee, Annie En-Shiun
author_facet Ranathungaa, Surangika
Nayak, Shravan
Huang, Shih-Ting Cindy
Mao, Yanke
Su, Tong
Chan, Yun-Hsiang Ray
Yuan, Songchen
Rinaldi, Anthony
Lee, Annie En-Shiun
contents Neural Machine Translation (NMT) systems built on multilingual sequence-to-sequence Language Models (msLMs) fail to deliver expected results when the amount of parallel data for a language, as well as the language's representation in the model are limited. This restricts the capabilities of domain-specific NMT systems for low-resource languages (LRLs). As a solution, parallel data from auxiliary domains can be used either to fine-tune or to further pre-train the msLM. We present an evaluation of the effectiveness of these two techniques in the context of domain-specific LRL-NMT. We also explore the impact of domain divergence on NMT model performance. We recommend several strategies for utilizing auxiliary parallel data in building domain-specific NMT models for LRLs.
format Preprint
id arxiv_https___arxiv_org_abs_2412_19522
institution arXiv
publishDate 2024
record_format arxiv
spellingShingle Exploiting Domain-Specific Parallel Data on Multilingual Language Models for Low-resource Language Translation
Ranathungaa, Surangika
Nayak, Shravan
Huang, Shih-Ting Cindy
Mao, Yanke
Su, Tong
Chan, Yun-Hsiang Ray
Yuan, Songchen
Rinaldi, Anthony
Lee, Annie En-Shiun
Computation and Language
Neural Machine Translation (NMT) systems built on multilingual sequence-to-sequence Language Models (msLMs) fail to deliver expected results when the amount of parallel data for a language, as well as the language's representation in the model are limited. This restricts the capabilities of domain-specific NMT systems for low-resource languages (LRLs). As a solution, parallel data from auxiliary domains can be used either to fine-tune or to further pre-train the msLM. We present an evaluation of the effectiveness of these two techniques in the context of domain-specific LRL-NMT. We also explore the impact of domain divergence on NMT model performance. We recommend several strategies for utilizing auxiliary parallel data in building domain-specific NMT models for LRLs.
title Exploiting Domain-Specific Parallel Data on Multilingual Language Models for Low-resource Language Translation
topic Computation and Language
url https://arxiv.org/abs/2412.19522