首页> 外国专利> WORD BASE GENERATION DEVICE, METHOD AND PROGRAM, AND COMPUTER-READABLE RECORDING MEDIUM

WORD BASE GENERATION DEVICE, METHOD AND PROGRAM, AND COMPUTER-READABLE RECORDING MEDIUM

机译:字库生成装置,方法和程序以及计算机可读记录介质

摘要

PPROBLEM TO BE SOLVED: To derive a word of an upper concept or a lower concept of various levels, or a sibling concept of a word from the word. PSOLUTION: In this word base generation device, when a corpus is inputted, a word vector wherein a value of each component is a relative value of co-occurrence frequency between an optional word A, and a word corresponding to the component or a word semantic attribute is generated to the word A inside the corpus and is stored in a word vector storage means, a divergence distance under a certain parameter between a word vector of a word B stored in the word vector storage means and a word vector of an optional word C stored in the word vector storage means is calculated to the optional word B, and an inter-word relevance degree between the word B and the word C is calculated and is stored in an inter-word relevance degree storage means. PCOPYRIGHT: (C)2010,JPO&INPIT
机译:

要解决的问题:从各个级别派生一个词的上层概念或下层概念,或一个词的同级概念。

解决方案:在该词库生成设备中,当输入语料库时,一个词向量,其中每个分量的值是可选词A与对应于该分量或词的词之间的共现频率的相对值。对语料库中的单词A生成单词语义属性,并将其存储在单词向量存储装置中,该参数在存储于单词向量存储装置中的单词B的单词向量与将存储在单词向量存储装置中的可选单词C计算为可选单词B,并且计算单词B与单词C之间的单词间相关度并将其存储在单词间相关度存储装置中。

版权:(C)2010,日本特许厅&INPIT

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