Social Network Analysis of China Computer Federation Co-author Network

摘要

We extract the data set from 2010 to 2014 of the China Computer Federation (CCF (http://www.ccf.org.cn)) by the distributed web crawler system and build the co-author network with R language. In this co-author network, authors represent nodes and a pair of authors is connected by an edge if they have co-authored at least one paper over the entire duration. We analyze this network with social-centric and ego-centric methods to study the situation of co-author network of CCF and visualize the analysis results. Social-centric measure reveals that the co-author network density, the usage of key words, the average number of authors per article, and confirms that most authors in computer science field publish a very small number of papers but have higher collaborators than those of other fields. Ego-centric analysis discovers betweenness centrality, closeness centrality, and degree centrality, indicates that only a small percentage of the authors locate in the center of the coauthor network. Finally, we pick out eight key persons and point out the group teams where the key persons are respectively. Based on these findings, we suggest that computer science field should promote wider collaboration, encourage more authors to publish their papers.