面向转移转化的专利技术组合方法对比研究  

A Comparative Study of Patent Portfolio Methods for Technology Transfer

作　　者：李姝影[1] 张丽华[2] 刘春江[1] 许轶[1] Li Shuying;Zhang Lihua;Liu Chunjiang;Xu Yi(Chengdu Library and Information Center, Chinese Academy of Sciences, Chengdu 610041;Shanxi University of Finance and Economics,Taiyuan 030006)

机构地区：[1]中国科学院成都文献情报中心,成都610041 [2]山西财经大学,太原030006

出　　处：《情报杂志》2022年第4期63-69,53,共8页Journal of Intelligence

基　　金：中国科学院文献情报中心青年人才领域前沿项目“面向转移转化的专利技术组合判别方法研究”(编号:Y8C0641002)研究成果之一;2022年中国科学院青年创新促进会资助项目。

摘　　要：[研究目的]基于供需双方转移转化场景,尝试从专利组合相似原理、实际可行性、聚类结果的客观性和稳定性考虑,筛选并对比分析有助于转移转化的专利组合方法。[研究方法]面向转移转化需求,选取三种关联组合:产业层面上的行业应用方向相似、技术层面上的技术引用相似以及研发层面的要素相似,分别基于专利共类、引文耦合以及主题词共现分析来识别。通过对比研究三种方法聚类生成的专利组合包在两年后专利实际转化情况,分析其优势和劣势。[研究结论]专利共类方法识别的产业应用组合包更适合面向研究机构/高校的回顾性专利打包;基于专利引文耦合的技术承接组合包在发生转移转化的预测上非常精准,更适合企业精准性开展技术对接;基于专利主题词共现的研发要素组合包细粒度较高,在发生转移转化的预测上较为精准,也能覆盖大部分的专利,更适合面向企业的前瞻性技术对接和技术问题的解决。三种方法都可以有效地用于识别潜在的合作对象,从各自的视角为技术转移转化提供更多的技术机会选择。[Research purpose]Based on the scenarios of technology transfer between supply and demand sides,this paper attempts to compare and analyze patent portfolio methods for technology transfer on the principle of similarity,practical feasibility,objectivity and stability of clustering results.[Research method]Three kinds of patent portfolios are selected:similar industry application directions at the industry level,similar technology references at the technology level and similar elements at the R&D level,which are identified based on patent co-classification,citation coupling and subject word co-occurrence,respectively.The advantages and disadvantages of the patent portfolios generated by the three methods of clustering are analyzed through a comparative study of the actual transfer of patents after two years.[Research conclusion]The industrial application portfolio identified by the patent co-classification is more suitable for retrospective patent packaging for research institutions/universities.The technology portfolio based on patent citation coupling is relatively precise in predicting the occurrence of technology transfer and more suitable for company to carry out matching precisely.The R&D portfolio based on keyword co-occurrence has higher granularity and is more accurate in predicting the occurrence of technology transfer.It can also cover most of the patents,which is more suitable for forward-looking technology matching and technology problem solving for company.All three methods can effectively be used for identifying potential collaboration targets,and will provide more options for technology transfer.

关 键 词：专利技术组合 技术转化 专利组合方法 专利共类 引文耦合 主题词共现 组合方法识别 

分 类 号：G250[文化科学—图书馆学]