科研合作网络形成机理——基于随机指数图模型的分析  

An Empirical Study of Mechanism for Scientific Collaboration Network: An Analysis Based on Exponential Random Graph Model

作　　者：刘璇[1] 汪林威 李嘉[1] 张朋柱[2] LIU Xuan;WANG Linwei;LI Jia;ZHANG Pengzhu(Business School,East China University of Scienceand Technology,Shanghai 200237,China;Antai College of Economicsand Management,Shanghai Jiao Tong University,Shanghai 200030,China)

机构地区：[1]华东理工大学商学院,上海200237 [2]上海交通大学安泰经济与管理学院,上海200030

出　　处：《系统管理学报》2019年第3期520-527,共8页Journal of Systems & Management

基　　金：国家自然科学基金资助项目(71471064,71371005);上海市浦江人才计划资助项目(15PJC019);教育部人文社会科学研究资助项目(18YJC630068);中央高校基本科研业务费专项基金资助项目(50321051924003)

摘　　要：知识时代的到来,使得学者个人的奋斗难以完成复杂的科研工作要求,学者之间的合作交流显得越来越重要。科研合作有助于产生创新思想,提高科研工作者的工作效率,促进多领域、多学科的交叉与融合,及缩短科研产出的周期等。因此,研究科研合作网络的形成机理具有重要意义。以维普为来源数据库,选取"知识管理"相关研究领域的文献,构建所有学者之间的合作网络,并运用指数随机图模型探究了网络结构和节点属性对合作网络形成的影响机理。结果显示:合作网络具有很好的传递性,网络分布较为均匀,未出现明显的核心节点;相同地区的学者,热门研究领域学者之间的合作更为普遍;此外,拥有高结构洞特征的学者和团队凝聚力高的学者与其他学者的合作更为普遍。With the advent of knowledge era, it is more difficult for scholars to accomplish complicated scientific research depending on individual efforts. The collaborations and communications among scholars are becoming more and more important. Scientific collaboration helps produce innovative thinking, increase efficiency, promote the interactions and integration of various domains and disciplines, and shorten the cycle of scientific products. It is, therefore, worthwhile to study the mechanism of collaboration network formation. In this paper, by collecting relevant papers in the "knowledge management" domain from Vip database, a collaboration network was established among all scholars, and the influencing mechanism of network structure and node attributes on the formation of cooperation network was explored by using the exponential random graph model(ERGM). The results indicate that the collaboration network has a good transitivity, and the network distribution is uniform, which means there are no obvious core nodes. Meanwhile, the homophily of region and research field among scholars increases the likelihood for their collaboration, and scholars with high structure holes and high group cohesiveness have more collaborations.

关 键 词：知识管理 科研合作网络 随机指数图模型 网络结构 节点属性 

分 类 号：G302[文化科学] G203