首页> 外国专利> ARGUMENT LOG EXTRACTION DEVICE, ARGUMENT LOG EXTRACTION METHOD, AND PROGRAM

ARGUMENT LOG EXTRACTION DEVICE, ARGUMENT LOG EXTRACTION METHOD, AND PROGRAM

机译：自变量日志提取设备，自变量日志提取方法和程序

页面导航

摘要
著录项
相似文献

摘要

PPROBLEM TO BE SOLVED: To quickly and highly precisely find a log under semantic argument from among logs whose online community is huge. PSOLUTION: A feature value of page text content in each online page is calculated, and the feature value of the AC act content generated on the page is calculated, and an AC act similarity score indicating the similarity of the calculated feature value of the page text content with the calculated feature value of the AC act content is calculated, and the feature value of the background knowledge of a page creator is calculated, and the feature value of the background knowledge of the AC actor is calculated, and an AC actor similarity score indicating the similarity of the calculated background knowledge of the page creator with the calculated feature value of the background knowledge of the AC actor is calculated, and an argument score which becomes higher according as the calculated AC act similarity score and the calculated AC actor similarity score become higher is calculated for each online page, and the page group whose argument score is equal to or more than a prescribed value is extracted. PCOPYRIGHT: (C)2010,JPO&INPIT

机译：

要解决的问题：要从网上社区非常庞大的日志中快速，高精度地找到一个基于语义论证的日志。

解决方案：计算每个在线页面中页面文本内容的特征值，并计算在页面上生成的AC行为内容的特征值，并且AC行为相似性得分指示计算出的AC内容的相似度计算具有所计算的AC行为内容的特征值的页面文本内容，计算页面创建者的背景知识的特征值，并计算AC行为者的背景知识的特征值，计算表示页面创建者的计算出的背景知识与AC参与者的背景知识的计算出的特征值的相似度的actor相似度分数，并且根据计算出的AC行为相似性分数和计算出的AC，变元分数变高针对每个在线页面计算演员相似度得分更高，并且论点得分等于或大于规定值的页面组被提取。

版权：（C）2010，日本特许厅＆INPIT 展开▼

著录项

公开/公告号JP2010066909A

专利类型
公开/公告日2010-03-25

原文格式PDF
申请/专利权人 NIPPON TELEGR & TELEPH CORP NTT;
展开▼

申请/专利号JP20080231343
发明设计人 MIYATA AKIHIRO;KAWASHIMA HARUMI;
展开▼

申请日2008-09-09
分类号G06F17/30;G06F17/21;G06Q50;
国家 JP
入库时间 2022-08-21 19:03:46

相似文献

专利
外文文献
中文文献