首页 | 本学科首页   官方微博 | 高级检索  
     检索      

一种用于专利主题词抽取的模板自动生成方法
引用本文:王裴岩,张桂平,蔡东风,白宇,叶娜.一种用于专利主题词抽取的模板自动生成方法[J].沈阳航空工业学院学报,2010,27(3):46-49.
作者姓名:王裴岩  张桂平  蔡东风  白宇  叶娜
作者单位:沈阳航空航天大学知识工程研究中心,辽宁沈阳,110136
摘    要:专利主题词是用以表述发明或实用新型名称的技术关键词,专利主题词的抽取是专利技术方案信息抽取的第一步,是填充信息抽取结果模板的有效依据和填充子之一。将信息抽取技术应用于中文专利摘要文本,在充分分析了专利摘要文本和专利标题特点的基础上,采用无指导的方法构建信息抽取模板,进而完成专利主题词的抽取。实验表明,该方法获得了较好的抽取效果。

关 键 词:中文专利摘要文本  专利主题词  无指导  模板

An automation pattern generation method for patent topic keyword extraction
WANG Pei-yan,ZHANG Gui-ping,CAI Dong-feng,BAI Yu,YE Na.An automation pattern generation method for patent topic keyword extraction[J].Journal of Shenyang Institute of Aeronautical Engineering,2010,27(3):46-49.
Authors:WANG Pei-yan  ZHANG Gui-ping  CAI Dong-feng  BAI Yu  YE Na
Institution:(Knowledge Engineering Research Center,Shenyang Aerospace University,Liaoning Shenyang 110136)
Abstract:The patent topic keyword is the word which describes the title of an invention or an utility model.The extraction of patent topic keyword is first step of a patent technical solution extraction.The patent topic keyword is one of the slot filler which could be an efficient evidence to fill the result templates.This paper applies information extraction technology to extract the patent topic keyword form Chinese patent Abstract texts,using unsupervised method to build information extraction patterns based on deeply analysis method of patent Abstract texts,titles and the other characteristic of patent information.The experimental results show that the method has better performance.
Keywords:Chinese patent Abstract text  patent topic keyword  unsupervised  pattern
本文献已被 维普 万方数据 等数据库收录!
设为首页 | 免责声明 | 关于勤云 | 加入收藏

Copyright©北京勤云科技发展有限公司  京ICP备09084417号