首页> 外国专利> NATURAL LANGUAGE SENTENCE RELATION DECIDING DEVICE, NATURAL LANGUAGE SENTENCE RETRIEVING DEVICE, NATURAL LANGUAGE SENTENCE GENERATING DEVICE, FRAMED EXPRESSION OUTPUT DEVICE TO BE USED FOR THE SAME, METHOD THEREFOR AND RECORDING MEDIUM

NATURAL LANGUAGE SENTENCE RELATION DECIDING DEVICE, NATURAL LANGUAGE SENTENCE RETRIEVING DEVICE, NATURAL LANGUAGE SENTENCE GENERATING DEVICE, FRAMED EXPRESSION OUTPUT DEVICE TO BE USED FOR THE SAME, METHOD THEREFOR AND RECORDING MEDIUM

机译:决定装置的自然语言关联,自然语言检索装置,自然语言产生装置,用于其的框架式表达输出装置,方法和记录介质

摘要

PROBLEM TO BE SOLVED: To improve the accuracy of data retrieval using a key sentence by deciding similarity/non-similarity between sentences after considering a difference between the meaning of both the sentences. SOLUTION: A similarity retrieval engine 10A extracts words from an input character string inputted from an external device 90 and the other character string while classifying them into conceptual word, affix, relational expression or postpositional expression (step S130) and discriminates the similarity between the extracted words (inter-word similarity) while referring to a conceptual word thesaurus 36a or functional expression thesaurus 36b (step S140). While using the value of this inter-word similarity between respective words, similarity between each word string of the input character string and the word string of the comparative character string (inter-word string similarity) is successively found (step S150) and the similarity between the word string including all words in the input character string and the word string including all words in the comparative character string is decided as similarity between the input character string and the comparative character string.
机译:解决的问题:在考虑两个句子含义之间的差异之后,通过确定句子之间的相似性/非相似性,来提高使用关键句子的数据检索的准确性。解决方案:相似度检索引擎10A从外部设备90输入的输入字符串中提取单词和其他字符串,同时将它们分类为概念性单词,词缀,关系表达或后置表达(步骤S130),并区分所提取的单词之间的相似性在参考概念性词库36a或功能性表达词库36b的同时使用词(词间相似度)(步骤S140)。在使用各个单词之间的该单词间相似度的值的同时,连续地找到输入字符串的每个单词列与比较字符串的单词列之间的相似度(单词间相似度)(步骤S150),并且相似度将包括输入字符串中的所有单词的单词串和包括比较字符串中的所有单词的单词串之间的相似性确定为输入字符串和比较字符串之间的相似性。

著录项

  • 公开/公告号JP2000322449A

    专利类型

  • 公开/公告日2000-11-24

    原文格式PDF

  • 申请/专利权人 A I SOFT INC;

    申请/专利号JP20000062878

  • 发明设计人 SHUDO KIMIAKI;KOYAMA YASUO;

    申请日2000-03-08

  • 分类号G06F17/30;G06F17/28;

  • 国家 JP

  • 入库时间 2022-08-22 01:28:25

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