| 注册
首页|期刊导航|信息工程大学学报|基于多语言预训练的越南语机器翻译优化技术

基于多语言预训练的越南语机器翻译优化技术

刘俊鸥 朱世平

信息工程大学学报2023,Vol.24Issue(6):679-684,6.
信息工程大学学报2023,Vol.24Issue(6):679-684,6.DOI:10.3969/j.issn.1671-0673.2023.06.007

基于多语言预训练的越南语机器翻译优化技术

Vietnamese Machine Translation Optimization Technology Based on Multilingual Pre-training

刘俊鸥 1朱世平1

作者信息

  • 1. 75776部队,广西南宁 530007
  • 折叠

摘要

Abstract

Traditional Vietnamese machine translation systems only build an independent bilingual translation model between the source language and the target language,and cannot learn information in other languages.This paper introduces a multilingual machine translation model.Due to the posi-tive transfer of resource rich language knowledge,Vietnamese machine translation can obtain rich knowledge information from other resource rich languages,and the translation quality has been sig-nificantly improved.On this basis,the translation effect of Vietnamese can be further improved by bilingual optimization,knowledge distillation,model compression and other methods.

关键词

预训练/模型优化/越南语/机器翻译

Key words

pre-training/model optimization/Vietnamese/machine translation

分类

信息技术与安全科学

引用本文复制引用

刘俊鸥,朱世平..基于多语言预训练的越南语机器翻译优化技术[J].信息工程大学学报,2023,24(6):679-684,6.

信息工程大学学报

1671-0673

访问量0
|
下载量0
段落导航相关论文