An Analysis of Social Network of Robotics Researchers in Japan

Abstract

In recent years, the social network has attracted attention as the system which causes the innovation, and various studies for researchers' networks have been carried out. This paper conducts social network analysis for a researchers' social network based on co-authorship in papers in the journal of RSJ. We investigate the 1,912 papers published from 1983 to 2010, which include 2,736 authors (nodes of the network) and 6,531 co-authorship (edges of the network). The network turns out to have general features as a complex network, such as scale-free structure, small-world structure, and cluster structure. We extract 710 core members from the network and analyze various centrality among them to show that some researchers demonstrate more importance in the network than their degrees. We also apply cluster analysis to the network to find 27 clusters in which researchers have close relationship, and give a coarse-grained representation of the clusters to clarify the adjacency among them.