首页> 外国专利> METHOD FOR SYNTAX ANALYSIS IN NATURAL LANGUAGE PROCESSOR

METHOD FOR SYNTAX ANALYSIS IN NATURAL LANGUAGE PROCESSOR

机译:自然语言处理中的语法分析方法

摘要

PURPOSE:To restrain a part of speech immediately after a predicate and to improve the accuracy of syntax analysis by writing the name of a part of speech restrained immediately after the predicate in the argument of the predicate. CONSTITUTION:An input sentence supplying device in a natural language processor 1 segments sentences supplied from a data input means into pieces of sentence based upon data indicating a punctuation mark or period and a morpheme analysis executing part divides each segmented input sentence into words and morphemes each of which is smaller than each word and sends these words and morphemes to a syntax analysis part, which executes the syntax analysis of each input sentence and sends the data of tree structure obtained as the result of syntax analysis to a semantic analysis part. Thus a part 2 obtained by writing the name of a part of speech restrained immediately after a predicate in the argument of the predicate during the notation of a difinitive clause grammar is connected to the processor 1. The name of the part of speech can be restrained by writing the name of the part of speech in the argument of the predicate.
机译:目的:通过在谓词的自变量中写紧在谓词之后的词性的名称,以限制在谓词之后的词性,并提高语法分析的准确性。构成:自然语言处理器1中的输入句子提供装置,根据表示标点符号或句点的数据,将从数据输入装置提供的句子分为句子,词素分析执行部分将每个分割的输入句子分为单词和词素其中每个词小于一个单词,然后将这些单词和词素发送到语法分析部分,语法分析部分执行每个输入句子的语法分析,并将作为语法分析结果获得的树结构数据发送到语义分析部分。因此,通过在限定性从句语法的记法期间在谓词的自变量中写紧在谓词之后被限制的词性的名称而获得的部分2被连接到处理器1。可以抑制词性的名称。通过在谓词的论点中写出词性的名称。

著录项

  • 公开/公告号JPH0778162A

    专利类型

  • 公开/公告日1995-03-20

    原文格式PDF

  • 申请/专利权人 MEIDENSHA CORP;

    申请/专利号JP19930161106

  • 发明设计人 OSADA MIHO;

    申请日1993-06-30

  • 分类号G06F17/27;

  • 国家 JP

  • 入库时间 2022-08-22 04:23:53

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