Analysis of Co-Authorship Pattern in Mathematics-Related Fields

Introduction: Scientific collaboration is often a defining feature and critical component of modern academic research. The analysis of networks of collaboration between scientists reveals features of academic communities that help in understanding the specifics of collaborative scientific work and identifying the notable researchers.

Objectives: This paper aims to investigate the patterns in scientific collaboration in the field of mathematics and related areas via graph theory, social network analysis, and bibliometric analysis.

Methods: Our research materials consist of journals indexed by Math-Net.Ru, a Russian mathematical portal that provides different scientific resources in mathematics, computer science, physics, information technology and related fields of science. The study covers a 20-year period — from 2000 to 2020.

Results: Most of the papers (> 54%) are found to be single-author publications, while 39% are co-authored by 2 to 4 people. The remaining 2% are papers written by 5 to 44 authors. This co-authorship structure shows that our object of study, Math-Net.Ru, does not follow the widely reported trend of a surge in ghost and honorary/guest/gift authorship. The main features of the constructed co-authorship graph–graph diameter, average path length, and percentage of vertices belonging to the maximally connected component–stabilize when the authorship count reaches 4.

Conclusions: Contrary to existing recommendations to acknowledge and separately describe the presence, in any sample, of articles with more than 10 co-authors, our results suggest that such recommendations are more appropriate and precise for papers co-authored by over 5 persons.