首页> 外国专利> Method for restoring Korean omitted components based on language model

Method for restoring Korean omitted components based on language model

机译:基于语言模型的韩文遗漏成分恢复方法

摘要

The present invention relates to a method for restoring a Korean omitted component based on a language model. The method comprises: an input step of receiving a sentence; an analysis step of analyzing construction for the received sentence; a determination step of determining whether an omitted word exists through construction analysis; an extraction step of extracting a restoration candidate word in the sentence for a phrase not directly related to a result of the construction analysis if the omitted word exists in the determination step; a language model applying step of generating configurable restoration candidate sentences using components in which omitted components exist; and an omission target determination step of calculating a configuration probability value for each generated restoration candidate sentence and selecting a restoration candidate sentence using the calculated configuration probability value.
机译:本发明涉及一种基于语言模型恢复韩文省略组件的方法。该方法包括:接收句子的输入步骤;以及分析步骤,用于分析所接收句子的构造;确定步骤,通过结构分析确定是否存在遗漏词;提取步骤,如果在确定步骤中存在遗漏的单词,则提取与该结构分析的结果不直接相关的短语的句子中的恢复候选单词;语言模型应用步骤,使用其中存在省略的组件的组件来生成可配置的恢复候选语句;省略目标确定步骤,为每个生成的恢复候选句子计算配置概率值,并使用计算出的配置概率值选择恢复候选句子。

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