Building and Analyzing a Global Co-Authorship Network Using Google Scholar Data

By publishing papers together, academic authors can form a co-authorship network, modeling the collaboration among them. This paper presents a data-driven study by crawling and analyzing the vast majority of author profiles of Google Scholar. We make the following major contributions: (1) We present a demographic analysis and get an informative overview of the authors from different aspects, such as the distribution of countries, scientific labels, and academic titles. (2) Based on the publication lists of crawled authors, we build a global co-authorship network with 402.39K authors to study the collaboration among authors. With the aid of social network analysis (SNA), we observe several unique features of this network. (3) We explore the relationship between the co-authorship network and citation metrics. We find a strong correlation between PageRank and h-index.