首页> 外国专利> METHOD AND DEVICE FOR ANALYZING MORPHEME OF SENTENCE IN WHICH KANJI AND KANA ARE MIXED

METHOD AND DEVICE FOR ANALYZING MORPHEME OF SENTENCE IN WHICH KANJI AND KANA ARE MIXED

机译:混合汉字和假名的句子形态分析方法和装置

摘要

PURPOSE:To decrease the processing quantity and the memory quantity by determining a pause of all sentence clause candidates by using an inter-character kind connecting and deciding dictionary. CONSTITUTION:A KANJI (Chinese character)/KANA (Japanese syllabary) mixed sentence inputted to an input register 1 passes through an inter-character kind detector 2 in order from the sentence head and sent to an intermediate register 3. In the detector 2, when the kind of character is different by two characters each from the sentence head, its two characters are sent to a comparator 4. In the comparator 4, the two characters are compared with the contents of an inter-character kind connecting and deciding dictionary 5, and when two characters sent from the detector 2 do not exist in the dictionary 5, they are decided to be the head of an independent word and a character end symbol determined in advance is inputted to the register 3. Subsequently, a character-string in the register 3 is sent to a minimum sentence clause number processing and deciding device 6, decomposed to combination of an independent word and an appended word whose sentence clause number is minimized by using an independent word dictionary 7 and an appended word dictionary 8, and its result is sent to an output register 9.
机译:目的:通过使用字符间种类的连接和确定字典来确定所有候选句子从句的暂停,以减少处理量和存储量。组成:输入到输入寄存器1的汉字(汉字)/假名(日语音节)混合语句从字符头依次经过字符间类型检测器2,并发送到中间寄存器3。在检测器2中,当字符的种类与句子首部的每个字符相差两个字符时,将其两个字符发送到比较器4。在比较器4中,将这两个字符与字符间类型的连接和确定字典5的内容进行比较。并且,当从检测器2发送的两个字符在词典5中不存在时,它们被确定为独立单词的开头,并且预先确定的字符结束符号被输入到寄存器3。寄存器3中的句号被发送到最小句子子句号处理和判定装置6,分解为独立词和附加词的组合,其附加词通过使用inde被最小化。下垂词词典7和附加词词典8,其结果发送到输出寄存器9。

著录项

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