CAJ | 학술논문

本文通过对双语语料的分析，从形式和语义两方面探讨计算语言学方法对翻译研究的作用和启示。在语言形式方面，通过文本特征统计信息研究了双语文本的基本形式特征，通过赋码语料研究了双语文本的深层句法信息。在语言意义方面，运用潜语义分析方法计算语句意义闾的相似度，语句计算的结果与人工评判的结果相差无几，表明计算语言学的意义分析方法十分有效，对判断多个译文闻、译文与原文间的“形似”与“神似”有重要意义。
본문통과대쌍어어료적분석，종형식화어의량방면탐토계산어언학방법대번역연구적작용화계시。재어언형식방면，통과문본특정통계신식연구료쌍어문본적기본형식특정，통과부마어료연구료쌍어문본적심층구법신식。재어언의의방면，운용잠어의분석방법계산어구의의려적상사도，어구계산적결과여인공평판적결과상차무궤，표명계산어언학적의의분석방법십분유효，대판단다개역문문、역문여원문간적“형사”여“신사”유중요의의。
This article, focusing on language form and language meaning, introduces the theories and methodologies of computational linguistics into translation studies. In terms of language form, with the help of the statistics extracted from the texts, researches can be done on both superficial formal features and deep syntactic features of the paralld texts. In terms d language meaning, semantic similarities can be confuted with the help of LSA, the result of which bears a very great significance for the judgment of formal and semantic similarities between multiple target texts or between source texts and target texts.