公开(公告)号：WO2012030053A3

公开(公告)日：2012-04-19

申请号：PCT/KR2011003832

申请日：2011-05-25

Applicant: SK TELECOM CO LTD ,  KIM SANG-BUM ,  YIN CHANG HAO ,  HWANG YOUNG SOOK ,  RIM HAE CHANG ,  LEE HYOUNG GYU

Inventor： KIM SANG-BUM ,  YIN CHANG HAO ,  HWANG YOUNG SOOK ,  RIM HAE CHANG ,  LEE HYOUNG GYU

IPC: G06F17/26 ,  G06F17/27

CPC classification number: G06F17/289 ,  G06F17/2775 ,  G06F17/2818 ,  G06F17/2827

Abstract: The present invention relates to an apparatus and method for recognizing an idiomatic expression using phrase alignment of a parallel corpus, and more particularly, to an apparatus and method extracting an idiom candidate expression using phrase alignment information of a parallel corpus and measuring an idiomatic expression index for each candidate idiomatic expression in order to recognize an idiomatic expression, thereby correcting errors in the measurement of translation entropy and in the extraction of a representative target word, as well as enhancing the accuracy of recognizing an idiomatic expression.

Abstract translation: 本发明涉及一种使用平行语料库的短语比对来识别习惯用语的装置和方法，更具体地，涉及一种使用平行语料库的短语比对信息来提取习语候选词表达并且测量习语表达指数 对于每个候选惯用语表达以便识别惯用表达，从而纠正翻译熵的测量和代表性目标词的提取中的错误，以及增强识别惯用表达的准确性。