首页> 外国专利> Apparatus And Method For Word Sense Disambiguation In Machine Translation System

Apparatus And Method For Word Sense Disambiguation In Machine Translation System

机译:机器翻译系统中词义消歧的装置和方法

摘要

PURPOSE: An apparatus and method for solving an ambiguity of a vocabulary in a machinery translation system is provided to solve an ambiguity of a vocabulary in case of a machinery translation by automatically generating and constructing a corpus tagged, and by respectively sampling local syntactic patterns and context information in the form of vocabulary and concept. CONSTITUTION: An input unit(101) inputs a sentence to be translated. A corpus generation unit(102) generates a corpus tagged using a COBALT-J/K(Collocation-Based Language Translator from Japanese to Korean). A sampling unit(103) performs a sampling of local syntactic patterns and context information from the corpus tagged. A processing unit(104) processes a generalization of the patterns and the information. A noun ambiguity solving unit(105) solves an ambiguity of a noun. A conversational meaning restriction database construction unit(106) constructs a database based on conversational meaning restriction information. In addition, the conversational meaning restriction database construction unit(106) provides the information to the noun ambiguity solving unit(105).
机译:目的:提供一种用于解决机器翻译系统中词汇歧义的装置和方法,以通过自动生成和构造标记的语料库,并分别对局部句法模式进行采样,来解决机器翻译中词汇歧义的问题。词汇和概念形式的上下文信息。构成:输入单元(101)输入要翻译的句子。语料生成单元(102)生成使用COBALT-J / K(从日语到韩语的基于位置的语言翻译器)标记的语料。采样单元(103)从被标记的语料库中采样局部句法模式和上下文信息。处理单元(104)处理模式和信息的概括。名词歧义求解单元(105)求解名词的歧义。会话含义限制数据库构建单元(106)基于会话含义限制信息来构建数据库。另外,会话含义限制数据库构建单元(106)将信息提供给名词歧义解决单元(105)。

著录项

  • 公开/公告号KR100574887B1

    专利类型

  • 公开/公告日2006-04-27

    原文格式PDF

  • 申请/专利权人

    申请/专利号KR19990068046

  • 发明设计人 이종혁;문경희;장두성;

    申请日1999-12-31

  • 分类号G06F17/27;

  • 国家 KR

  • 入库时间 2022-08-21 21:23:54

相似文献

  • 专利
  • 外文文献
  • 中文文献
获取专利

客服邮箱:kefu@zhangqiaokeyan.com

京公网安备:11010802029741号 ICP备案号:京ICP备15016152号-6 六维联合信息科技 (北京) 有限公司©版权所有
  • 客服微信

  • 服务号