首页> 外国专利> LANGUAGE KNOWLEDGE ACQUISITION DEVICE AND LANGUAGE KNOWLEDGE ACQUISITION PROGRAM

LANGUAGE KNOWLEDGE ACQUISITION DEVICE AND LANGUAGE KNOWLEDGE ACQUISITION PROGRAM

机译:语言知识获取设备和语言知识获取程序

摘要

PPROBLEM TO BE SOLVED: To solve the problem that it is impossible to compensate any case relation and case element missing in a case frame with respect to the sentence of a title from the case frame of the sentence of the other title in a conventional manner, and that it is impossible to extract the relations of case elements each other or declinable words each other, and acquire the IS-A relation of substantives or the IS-A relation of declinable words. PSOLUTION: This language knowledge acquisition device is configured to extract the hierarchial relation of the titles of texts or nodes from an input document having a document structure by a title hierarchical relation extraction part, and to edit an input document by a document edition part for extracting the hierarchical relation of the titles again by a title hierarchy relation extraction part when the hierarchical relation of the titles is inappropriate, and to, when the hierarchical relation of the titles is appropriate, generate the case frame of each title by analyzing the structure of the sentence of the title with a sentence analyzing part by analyzing the structure of the sentence hierarchical relation of the titles, and to acquire the case frame and IS-A relation of the case frame from which the missing case relation and case element are compensated by merging the case element and case relation of the same declinable words with respect to the case frame of every title by a case merge part based on the title hierarchical relation. PCOPYRIGHT: (C)2009,JPO&INPIT
机译:

要解决的问题:解决一个问题,即不可能从一个标题中另一个标题的句子的格中补偿一个标题句子在格框架中丢失的任何格关系和格元素。按照传统方式,不可能提取大小写元素之间的关系或可否定词之间的关系,而无法获得实体词的IS-A关系或可否定词的IS-A关系。

解决方案:该语言知识获取设备被配置为通过标题层次关系提取部分从具有文档结构的输入文档中提取文本或节点的标题的层次关系,并通过文档版本来编辑输入文档。在标题的层次关系不合适的情况下,通过标题层次关系提取部分再次提取标题的层次关系的部分,并且当标题的层次关系适当时,通过分析通过分析标题的句子层次关系的结构,具有句子分析部分的标题的句子的结构,并获得缺少的案例关系和案例要素的案例框架和IS-A关系通过合并大小写元素和相同可拒绝词的大小写关系相对于每个标题的大小写框架来补偿基于标题层次关系的案例合并部分。

版权:(C)2009,日本特许厅&INPIT

著录项

  • 公开/公告号JP2009140056A

    专利类型

  • 公开/公告日2009-06-25

    原文格式PDF

  • 申请/专利权人 MITSUBISHI ELECTRIC CORP;

    申请/专利号JP20070313348

  • 发明设计人 IMAMURA MAKOTO;TAKAYAMA YASUHIRO;

    申请日2007-12-04

  • 分类号G06F17/27;G06F17/21;

  • 国家 JP

  • 入库时间 2022-08-21 19:44:22

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