摘要
针对无监督神经机器翻译在处理源语言和目标语言之间存在显著语法结构和词汇差异时鲁棒性与泛化能力不足的问题,提出一种基于自适应噪声插入策略的无监督神经机器翻译方法。该方法通过对源语言和目标语言在语法结构、词汇复杂度及句法差异性的分析,能够动态调整噪声的插入位置和强度,从而更好地适应不同语言对的复杂性。在结构简单的句子中,插入较少噪声以保护核心语义,而在复杂句子中,插入较为复杂的噪声以增强模型对复杂语言的学习能力,确保模型在保留重要语义信息的同时,能够增强模型的泛化能力和鲁棒性。实验结果表明,与基线模型相比,该方法在8个基准翻译任务中,显著提升双语评估替补(BLEU)值。
To address the issue of insufficient robustness and generalization ability in unsupervised neural machine translation when handling significant grammatical structure and lexical differences be-tween source and target languages,an unsupervised neural machine translation method based on an adaptive noise insertion strategy is proposed.By analyzing the grammatical structure,lexical complex-ity and syntactic difference between the source language and the target language,the noise insertion position and intensity can be dynamically adjusted,so as to better adapt to the complexity of different language pairs.For simpler sentences,less noise is inserted to preserve core semantics,while for more complex sentences,more intricate noise is introduced to enhance the model’sability to learn complex language structures.This ensures that the model can retain important semantic information while im-proving its generalization ability and robustness.Experimental results show that compared with the baseline model,the bilingual evaluation understudy(BLEU)value is significantly improved by using the proposed method in eight benchmark translation tasks.
作者
张传财
屈丹
李真
都力铭
ZHANG Chuancai;QU Dan;LI Zhen;DU Liming(Information Engineering University,Zhengzhou 450001,China;Zhengzhou University,Zhengzhou 450001,China)
出处
《信息工程大学学报》
2025年第4期431-437,共7页
Journal of Information Engineering University
基金
国家自然科学基金(62171470)
中原科技创新领军人才项目(234200510019)
河南省自然科学基金面上项目(232300421240)。
关键词
无监督神经机器翻译
动态噪声插入
注意力机制
源/目标语言
句子复杂度
unsupervised neural machine translation
dynamic noise insertion
attention mechanism
source/target languages
sentence complexity
