Understanding Scholar Social Networks: Taking SCHOLAT as an Example

Min Gao, Yang Chen, Qingyuan Gong, Xin Wang, Pan Hui

Abstract: Scholar social networks are composed of scholars and social connections among them. Studying such social networks can help promote academic exchanges and cooperation, and predict future trends in research. In this paper, we analyze SCHOLAT, a representative scholar social network in China, from three perspectives. First, we explore SCH-OLAT’s social graph, and we find this graph has a smaller average shortest-path length and a higher clustering coefficient than other social networks, for example, the collaboration network of Google Scholar and the Flickr social network. Moreover, we leverage the structural hole theory to identify important users on SCHOLAT. By comparing the top-500 structural hole spanners with 500 randomly selected users, we have found that the former have the higher values of several graph-based metrics, and they also connect more communities. Finally, we also undertake user group-based analysis, and we discover that the users belonging to Guangdong province, and the users from the top universities in China are well-connected and occupy important positions in the network.