专利技术组合机会与合作潜力研究  

Exploring Potential Patent Portfolios: an Integrated Approach Based on Topic Identification and Correlation Analysis

作　　者：张娴 许海云[1] 方曙[1] 胡正银[1,2] 

机构地区：[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,北京100190

出　　处：《情报杂志》2015年第7期39-46,共8页Journal of Intelligence

基　　金：中国科学院知识产权专项工作"中国科学院知识产权信息服务"(编号:KFJ-EW-STS-032);中国科学院西部之光-西部博士项目"基于内容分析的领域核心专利挖掘及专利权人竞合态势分析方法研究"(编号:Y4C0011001)

摘　　要：提出了一种专利技术组合与合作潜力分析方法,针对特定机构或领域的专利产出,利用数据挖掘、主题识别、关联分析、可视化技术等方法与手段,挖掘专利技术特征,判断相似程度,从中寻找有望进一步构建专利组合的潜在技术、有望建立合作关系的专利主体。针对国内某科研机构开展的实验分析结果显示,该方法能够有效地映射出专利技术的知识单元,能够为机构挖掘自身专利技术组合可能、机构间寻找战略合作潜在方向等提供有效的决策参考,在大科学时代的协同创新形势下具有广阔的应用前景。This paper proposes a framework for the identification of the potential patent portfolios, so as to identify potential patent pools or further collaborators for assignees, the proposed framework employs the topic identification and association analysis of patent documents to identify the correlation technologies, including the following steps：Extracting the technology topics of the patent solutions in two dimen- sions - classification codes and topic terms ;Constructing the association networks, clustering in topics,computing the similarities and visu- alizing with NetDraw;Interpreting the maps, to investigate the potential portfolios and partnerships. A case study is conducted, experimen- ting with 27401 chemistry patents filed by a national research institute, the empirical results show that the framework proposed can perform well

关 键 词：专利技术组合 专利合作 主题识别 关联分析 

分 类 号：G306[文化科学]