基于改进机器学习的智能英语翻译机器人在线校准系统研究  

Research on the Online Calibration system of Intelligent English Translation Robot Based on improved Machine Learning

作　　者：陈晋华[1] CHEN Jinhua(Yulin University,Yulin,Shaanxi 719000,China)

机构地区：[1]榆林学院,陕西榆林719000

出　　处：《自动化与仪器仪表》2022年第10期201-206,共6页Automation & Instrumentation

基　　金：榆林高新区科技计划项目《语言经济学视角下的英语教育供给侧改革与榆林市区域经济发展的相关性研究》(CXY-2021-76)。

摘　　要：针对传统英语翻译机器人在线校准系统的翻译质量差、译文质量估计不准确的问题,提出基于BiLSTM算法,构建CUNQE译文质量估计模型;之后在CUNQE模型的基础上进行英语翻译机器人在线校准系统整体结构设计。实验结果表明,在传统译文质量估计方法中加入语境词向量后,译文质量估计的准确性有效提升。相较于UNQE方法,CUNQE方法译文质量和相关性明显增强。且在汉英翻译和英汉翻译中,本方法的皮尔森相关系数均比传统UNQE模型高出了1.6%和2.4%。由此可知,提出的CUNQE方法能够进一步提升译文估计质量,具备一定的有效性。For the problems of poor translation quality and inaccurate translation quality estimation of traditional English translation robot,BiLSTM algorithm,Build the CUNQE translation quality estimation model,and then design the overall struc-ture of the English translation robot online calibration system based on the CUNQE model.Experimental results show that the accuracy of the context word vector is effectively improved in the traditional translation quality estimation method.Compared with UNQE,CUNQE significantly improves the translation quality and correlation.Moreover,the Pearson correlation coeffi-cient is 1.6%and 2.4%higher than the traditional UNQE model.Therefore,the CUNQE method proposed in this paper can further improve the quality of translation estimation and has certain effectiveness.

关 键 词：英语翻译机器人 在线校准 译文质量估计 BiLSTM CUNQE 

分 类 号：TP392[自动化与计算机技术—计算机应用技术]