arXiv 2412.195222026-03-26cs.CL

利用多语言语言模型中的领域特定平行数据进行低资源语言翻译

Exploiting Domain-Specific Parallel Data on Multilingual Language Models for Low-resource Language Translation

School of Mathematical and Computational Sciences, Massey University（马歇尔大学数学与计算科学学院）；Mila – Quebec AI Institutes（魁北克人工智能研究院）；Computer Science, Ontario Technology University and University of Toronto（安大略技术大学与多伦多大学计算机科学系）

Surangika Ranathungaa, Shravan Nayak, Shih-Ting Cindy Huang, Yanke Mao, Tong Su, Yun-Hsiang Ray Chan, Songchen Yuan, Anthony Rinaldi, Annie En-Shiun Lee

阅读 PDF

arXiv 页面 ↗HTML 原文 ↗

AI总结：

本文研究了如何通过领域特定平行数据提升低资源语言翻译效果，探讨了预训练和微调技术对模型性能的影响，并提出优化策略。

摘要：

基于多语言序列到序列语言模型（msLM）的神经机器翻译（NMT）系统在平行数据有限或语言表示受限时无法达到预期效果。这限制了领域特定NMT系统在低资源语言（LRLs）中的能力。为此，本文提出利用辅助领域平行数据进行微调或进一步预训练msLM。我们评估了这两种技术在领域特定LRL-NMT中的有效性，并探讨了领域差异对NMT模型性能的影响。本文还推荐了几种利用辅助平行数据构建领域特定NMT模型的策略。

英文摘要：

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.

机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。