首页> 外国专利> DEVICE AND METHOD FOR PROVIDING CPC BASED TECHNOLOGY CLASSIFICATION OF PATENT DOCUMENT USING MACHINE LEARNING

DEVICE AND METHOD FOR PROVIDING CPC BASED TECHNOLOGY CLASSIFICATION OF PATENT DOCUMENT USING MACHINE LEARNING

机译:使用机器学习提供基于CPC技术分类的设备和方法

摘要

A patent document automatic classification device using machine learning is provided, and a transaction is constructed by extracting the contents contained in at least one identification item corresponding to each of the extraction ranges set in the similar patent classification and classification schema classification from the classification target patent document, , a text mining module that extracts keywords in a transaction through text mining, a classification schema index, CPC class unit index, and publication unit index database for similar patent classification and classification schema classification, and cloud notes composed of characteristic words for each CPC classification code A comparison target having a preset minimum matching rate with the extracted keyword by giving a preset keyword weight to each field including the database engine and the extracted keyword, and giving a preset field weight to each comparison target in the database engine is included in the similar candidate group, the similarity of the comparison target included in the similar candidate group is calculated based on the vector space model, and the degree of similarity is determined through ranking correction using additional information for the CPC classification code in the comparison target. It includes a similarity measurement module that sorts and outputs in high order.
机译:提供了一种使用机器学习的专利文献自动分类装置,并且通过从分类目标专利的类似专利分类和分类模式分类中设置的每个提取范围内容的至少一个识别项目中包含的至少一个识别项目中包含的内容来构建事务文档,是通过文本挖掘,分类架构索引,CPC类单元索引和发布单元索引数据库,用于类似专利分类和分类模式分类的文本挖掘模块,用于用于类似的专利分类和分类架构分类,以及由每个特征字组成的云笔记CPC分类代码通过向包括数据库引擎和提取的关键字的每个字段向包括数据库引擎和提取的关键字提供预设的关键字权重,包括将预设的关键字具有预设的最小匹配速率,并包括数据库引擎中的每个比较目标的每个字段在类似的c中andidate组,基于矢量空间模型计算包括在类似候选组中的比较目标的相似性,并且通过使用比较目标中的CPC分类码的附加信息进行排序来确定相似度的程度。它包括一个相似性测量模块,其按大阶排序和输出。

著录项

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