首页|动态专利技术网络视域下技术集群识别

动态专利技术网络视域下技术集群识别

扫码查看
原文链接
  • 万方数据
  • 维普
[研究目的]国与国之间的技术资源竞争已经从单一技术竞争转向技术集群竞争.从动态专利技术网络的角度识别技术集群,有助于企业和研发人员了解技术结构、挖掘技术机会和预测技术未来,从而在技术研发和竞争中获得优势.[研究方法]采用自注意力机制学习出专利文档特定词向量,融合属性和功能技术因子等先验信息生成专利技术向量,计算专利技术相似度后构建动态专利技术网络,采用带节点增量的Louvain算法识别技术集群.[研究结论]在以美国电商数据技术为例的实证研究中发现了数据系统、数据设备、数据安全等 12 个技术集群及 1个技术集群融合事件,实证分析进一步验证了该方法的有效性.
Research on Technology Cluster Identification from the Perspective of Dynamic Patent Technology Network
[Research purpose]Technology resource competition among countries has shifted from single technology competition to tech-nology cluster competition.Identifying technology cluster from the perspective of dynamic patent technology network can help enterprises and researchers to understand technology structure,explore technology opportunities and predict the future of technology,so that they can gain advantages in technology research and development and competition.[Research method]The research uses self-attention mecha-nism to learn the specific word vector of patent document,and integrates the prior information such as attribute technology factor and func-tional technology factor to generate the patent technology vector.After calculating the patent technology similarity,a dynamic patent tech-nology network is built,and the Louvain algorithm with node increment is further used to identify the technology cluster.[Research con-clusion]Based on this method,12 technology clusters such as data system,data equipment and data security,and 1 technology cluster fu-sion event are identified in an empirical study,which takes the e-commerce data technology in the United States as an example and further verifies the effectiveness of the method.

patent technologytechnology clustertechnology cluster identificationdynamic patent technology networktechnical simi-larityLouvain algorithm

刘玉林、菅利荣

展开 >

南京航空航天大学经济与管理学院 南京 211106

安徽商贸职业技术学院电子商务学院 芜湖 241002

专利技术 技术集群 技术集群识别 动态专利技术网络 技术相似度 Louvain算法

安徽省高等学校自然科学研究重大项目安徽省高等学校中青年教师培养行动学科(专业)带头人培育项目芜湖市科技应用基础研究项目

2023AH040319DTR20230922022jc36

2024

10.3969/j.issn.1002-1965.2024.04.016

情报杂志
陕西省科学技术信息研究所

情报杂志

CSTPCDCSSCICHSSCD北大核心
影响因子：1.502
ISSN：1002-1965
年,卷(期)：2024.43(4)
  • 27