摘要
针对当前的翻译系统由于缺乏对语义的充分理解,导致翻译质量不高的问题。文中设计一种基于语义分析的交互式口语机译翻译系统。该系统先通过语音采集,将源语音转换为文本形式,然后对源语言文本进行语义分析,并结合标准语言与逻辑顺序进行语句重组,然后按照文本翻译方式将其翻译为目标语言文本,再将文本转换为目标语言语音形式播放出去,实现口语跨语言交互。实验结果表明:利用所设计系统对两组示例(英译汉、汉译英)进行交互翻译后,通过专家评定,认为所设计系统翻译的语句质量较高,基本满足了人们日常口语交流对话所需。
Since the translation quality of the current translation systems is not high due to the lack of full understanding of semantics.an interactive oral machine translation system based on semantic analysis is designed.Firstly,the system converts the source speech into text form by means of voice acquisition,and then carries out semantic analysis on the source language text.The sentence reorganization is executed in combination with the standard language and logical order,and then the reorganized sentences are translated into the target language text according to the text translation mode.After that,the text is converted into the target language voice form to play out.So far,the oral cross language interaction is realized.In the test verification,the interactive translation(English-Chinese and Chinese-English)of the two sets of examples was performed with the designed system.The expert evaluation results show that the sentence quality of the designed system is high,which basically meets the needs of people′s daily oral communication.
作者
边铀铀
BIAN Youyou(Zhengzhou University of Light Industry,Zhengzhou 450002,China)
出处
《现代电子技术》
2021年第14期75-80,共6页
Modern Electronics Technique
基金
国家自然科学基金项目(61603346)
国家社会科学基金项目(16BYY157)
教育部产学合作协同育人项目(201902166020)。
关键词
口语机译
语义分析
语音采集
语音处理
口语交互
语句重组
spoken language machine translation
semantic analysis
speech acquisition
speech processing
oral language interaction
sentence reorganization