基于社会网络分析的指标群划分及群间关系研究  

Research on the Analysis of Index Group Division and Inter Group Relationship Based on Social Network

作　　者：程豪 易丹辉[2] 牟宗毅[3] Cheng Hao;Yi Danhui;Mou Zongyi(National Academy of Innovation Strategy,China Association for Science and Technology,Beijing 100012,China;School of Applied Statistics,Renmin University of China,Beijing 100872,China;Affiliated Hospital of Changchun University of Traditional Chinese Medicine,Changchun 130117,China)

机构地区：[1]中国科协创新战略研究院,北京100012 [2]中国人民大学应用统计科学研究中心,北京100012 [3]长春中医药大学附属医院,长春130117

出　　处：《世界科学技术-中医药现代化》2018年第4期505-513,共9页Modernization of Traditional Chinese Medicine and Materia Medica-World Science and Technology

基　　金：教育部人文社会科学重点研究基地重大项目(16JJD910002):基于大数据的精准医学生物统计分析方法及其应用研究;负责人:易丹辉

摘　　要：目的:从网络中指标节点间的关系信息出发,定量测度指标节点成群及群间结构关系。方法:Tabu搜索算法来实现指标子群、指标群的划分及指标群间关系的测度。结果:通过仿真实验和真实数据,Tabu搜索算法较准确的实现了不同类型指标网络的划分,同时定量测度了不同指标群划分后的结构关系。结论:Tabu搜索算法可以较为准确反映网络中不同指标的"凝聚"情况,同时又可以定量测度指标群间的关系,在一定程度上可以推广到规模较大或网络结构更复杂的指标间关系的测度问题。To measure the structure relationship among indexes and different indexes groups based on the networkrelationship information. Methods： Tabu search algorithm was used as the concrete algorithm to study the composition ofindex groups and the structure relationship between them. Result： The results of simulation and real data show that Tabusearch algorithm accurately solve the division of indexes and measure the structure relationship among different groups.Conclusion： Tabu search algorithm measure the structure relationship among indexes and different indexes groups, whichextend to larger network and more complex situations.

关 键 词：社会网络分析 指标群 群间关系 

分 类 号：R1[医药卫生—公共卫生与预防医学]