基于专利共现网络社区发现的军事智能技术演化分析

Evolution Analysis of Military Intelligence Technology Based on Community Detection of Patent Co-Occurrence Networks

[目的/意义]针对军事智能技术领域,采用基于社区发现的技术演化分析方法,实现对军事智能技术整体演化趋势和演化关系的分析,并验证基于社区发现的演化分析方法的有效性,为科技工作者更全面细致地掌握领域技术演化情况、探究技术起源和及早发现新技术萌生趋势提供依据和借鉴。[方法/过程]基于德温特专利数据库采集军事智能技术相关专利数据,利用专利分类代码在专利中的共现关系逐年构建共现网络,采用社区发现方法划分共现网络的社区结构,提出社区综合相似度计算方法计算社区相似度,形成各个技术子领域演化路径,分析军事智能技术的整体演化趋势和演化关系。[结果/结论]利用2012—2021年间16902条专利,结合本文的分析方法,从整体上识别出军事智能技术演化的趋势,并验证了继承、分化和融合关系分析方法的有效性。

[Purpose/significance]For the field of military intelligence technology,we adopt the community detection-based technology evolution analysis method to realize the analysis of the overall evolution trend and evolution relationship of military intelligence technology,and verify the effectiveness of the community detection-based evolution analysis method,so as to provide a basis and reference for science and technology workers to grasp the field technology evolution more comprehensively and carefully,explore the origin of technology and discover the trend of new technology emergence early.[Method/process]Based on the Derwent patent database to collect patent data related to military intelligence technologies,the co-occurrence network was constructed year by year using the co-occurrence relationship of patent classification codes in patents,the community structure of the co-occurrence network was divided by the community detection method,the community similarity was calculated by the proposed community comprehensive similarity calculation method,the evolutionary path of each technology subfield was formed,and the overall evolutionary trend and evolutionary relationship of military intelligence technologies were analyzed.relationships.[Result/conclusion]Using 16902 patents during 2012—2021,combined with the analysis method of this paper,the trend of military intelligence technology evolution is identified as a whole,and the effectiveness of the analysis method of inheritance,differentiation and convergence relationships is verified.