图书情报工作 ›› 2017, Vol. 61 ›› Issue (1): 117-124.DOI: 10.13266/j.issn.0252-3116.2017.01.014

• 知识组织 • 上一篇    下一篇

面向TRIZ的领域专利技术挖掘系统设计与实践

胡正银1, 刘春江1, 隗玲1,2,3, 杨宁1, 徐源1, 许海云1, 文奕1,4   

  1. 1. 中国科学院成都文献情报中心 成都 610041;
    2. 中国科学院大学 北京 100049;
    3. 山西财经大学信息管理学院 太原 030006;
    4. 电子科技大学计算机科学与工程学院 成都 611731
  • 收稿日期:2016-09-09 修回日期:2016-12-21 出版日期:2017-01-05 发布日期:2017-01-05
  • 作者简介:胡正银(ORCID:0000-0002-5699-9891),副研究员,博士,E-mail:huzy@clas.ac.cn;刘春江(ORCID:0000-0001-8934-339X),助理研究员,硕士;隗玲(ORCID:0000-0002-0594-3134),讲师,博士研究生;杨宁(ORCID:0000-0001-9121-4458),副研究员,硕士;徐源(ORCID:0000-0003-4246-5730),助理研究员;许海云(ORCID:0000-0002-7453-3331),副研究员,博士;文奕(ORCID:0000-0001-9871-1035),副研究员,博士研究生。
  • 基金资助:

    本文系中国科学院知识产权信息服务专项(项目编号:KFJ-EW-STS-032)和中国科学院西部之光项目“基于本体的专利文献技术挖掘系统研究与实践”研究成果之一。

Design and Practice of Domain Patent Tech Mining System Oriented to TRIZ

Hu Zhengyin1, Liu Chunjiang1, Wei Ling1,2,3, Yang Ning1, Xu Yuan1, Xu Haiyun1, Wen Yi1,4   

  1. 1. Chengdu Document and Information Centre, Chinese Academy of Sciences, Chengdu 610041;
    2. University of Chinese Academy of Sciences, Beijing 100049;
    3. School of Information and Management, Shanxi University of Finance and Economics, Taiyuan 030006;
    4. University of Electronic Science and Technology of China, Chengdu 611731
  • Received:2016-09-09 Revised:2016-12-21 Online:2017-01-05 Published:2017-01-05

摘要:

[目的/意义] 针对面向TRIZ的专利技术深度、精准挖掘的需求,设计并开发一套领域专利技术挖掘系统。[方法/过程] 首先,归纳面向TRIZ的专利技术挖掘的具体需求,分析现有工具的不足。其次,提出领域专利技术挖掘系统的体系结构,总结其关键技术及解决方案。最后,开发一套面向TRIZ的领域专利技术挖掘原型系统,并进行大口径光学元件(LAOE)领域专利技术挖掘实践。[结果/结论] 该系统基于SAO(subject-action-object)三元组与简单知识对象,集成文本挖掘技术构建细粒度、多维度的领域技术索引,实现领域知识棱镜、面向TRIZ的语义检索与专利可视化分析功能,可以支持深度、精准的专利技术挖掘应用。

关键词: 技术挖掘, SAO, 语义TRIZ, 知识棱镜, 专利分析

Abstract:

[Purpose/significance] According to the requirements of deep and precise patent tech mining, this paper designed and developed a domain patent tech mining system oriented to TRIZ.[Method/process] This paper induced the requirements of patent tech mining oriented to TRIZ, analysed the disadvantages of exists tools, proposed a framework of domain patent tech mining system, and summarized key technologies and corresponding solutions. Based on these, a prototype of a domain patent tech mining system was developed.[Result/conclusion] The system considers SAO triples and simple knowledge objects as basic semantic units and includes a fine-grained, multi-dimensional domain tech index, which offers three services, semantic search service oriented to TRIZ, domain knowledge lens, and visualization analysis service, to support deep and precise patent tech mining.

Key words: tech mining, Subject-Action-Object, semantic TRIZ, knowledge lens, patent analysis

中图分类号: