摘要
以计算机作为工具,在语言学理论研究的基础上,开发以自然语言为介质的信息系统是语言工程的使命。上述理念正是张柏然教授提出的翻译的语言工程愿景。本文基于张教授所设想的声学语块辅助消歧、语料库助力术语翻译理念,深入探究语音识别过程并揭示其存在的问题,引入声学语块概念,以便提高语音识别的准确率并探讨语义消歧问题。本文还提出建设术语语料库的方案,并探讨术语语料库辅助翻译路径。从语言学视角介入人工智能机器翻译,旨在为其语言工程路径研究提供启示,为人工智能口笔译研究提供方向和思路。
On the basis of the research of linguistic theory,it is the mission of language engineering to develop an information system taking natural language as medium,which is exactly the language engineering vision of translation put forward by Professor Zhang Boran.First,we deeply explore the speech recognition and discuss the semantic disambiguation by introducing phonetic chunking to improve the accuracy of speech recognition.Second,we propose a scheme to construct a terminology corpus and discuss the path of terminology corpus assisting translation.From the perspective of linguistics,AI machine translation is explored to provide enlightenment for the study of language engineering path and directions for the study of AI interpretation and translation.
作者
张笛
程璐璐
ZHANG Di;CHENG Lulu
出处
《当代外语研究》
2023年第2期74-81,共8页
Contemporary Foreign Language Studies
基金
山东省高等学校青年创新团队人才引育计划“多语种语言教育与服务创新团队”与“人工智能与口译研究团队”
山东省优质专业学位教学案例库:《笔译工作坊》仿真实践教学案例库建设(编号SDYAL2022034)
山东省研究生教育优质课程:基础英语(编号SDYKC21212)的阶段性成果。
关键词
张柏然翻译思想
语言工程路径
声学语块
消歧
术语语料库
Translation concept of Professor Zhang Boran
language engineering path
phonetic chunking
disambiguation
terminology corpus