摘要
汉英机器翻译的能力一直是应用语言学界探讨的焦点。本文通过对直译和意译在传统翻译理论中的界定,提出计算语言学中对直译和意译的形式化和识别的方法,并把其应用到统计机器翻译(SMT)对直译和意译句对的识别中,最后在机器翻译实验中对直译、意译理论进行验证。
The performance of machine translation has been a significant focus of applied linguistics.This paper defines literal translation and free translation in traditional translation theory,and brings forward the method of formalization and cognition of literal and free translation sentence pairs,including lexical compatibility and grammatical compatibility,then applies them in statistical machine translation（SMT）,validating the theory of literal and free translation in machine translation.
出处
《外语学刊》
CSSCI
北大核心
2011年第4期113-116,共4页
Foreign Language Research
基金
国家自然科学基金项目"英汉动词次范畴化对应关系自动获取"(60770069)
国家自然科学基金项目"面向句法分析的动词次范畴化应用技术"(60873169)的阶段性成果
关键词
直译
意译
统计机器翻译
应用
literal translation
free translation
SMT
application
