常熟理工学院学报2025，Vol.39Issue(3)：36-42,7.
基于语言模因神经网络的机器翻译质量研究
Research on the Quality of Machine Translation Based on Language Meme Neural Networks
摘要
Abstract
As a fundamental unit of cultural transmission,language meme offers a new perspective toward mechanisms of replication and dissemination for machine translation.This study analyzes an English-Chinese bilingual corpus,extracts representative linguistic memes,and constructs neural network models.The research finds that in the process of linguistic meme dissemination,specific words,phrases,and sentence patterns often have higher transmission power and adaptability.These memes can maintain their core meaning and form when transmitted between different languages and cultures,which could provide translation patterns for machine translation.Experimental results show that the neural network model incorporating linguistic memes significantly outperforms traditional models in translation quality in terms of translation fluency and accuracy.When dealing with expressions with cultural characteristics,the linguistic meme model can better preserve the semantics and style of the original text,enhancing the fluency and accuracy of machine translation.关键词
语言模因/神经网络/机器翻译/汉语俚语英译Key words
language meme/neural network/machine translation/translation of Chinese slang into English分类
计算机与自动化引用本文复制引用
王纯磊,范开芳..基于语言模因神经网络的机器翻译质量研究[J].常熟理工学院学报,2025,39(3):36-42,7.基金项目
江苏省社会科学基金项目"语言'模因'神经网络机器翻译研究"(21YYB014) （21YYB014）
