Collaborative authorship patterns in computer science publications

Abstract

Based on the analysis of data we observe that the share of single-authored papers was significantly high in theoretical
computer science, while collaborative efforts dominate computer science system research like PL, AI, ML, etc.
Collaborative authorship is higher in journals over conferences. Further, values of collaborative indicators are also high for
journals except for the Machine Learning (ML) subfield. In addition, the author distribution patterns are different for
conferences and journals. The findings also exhibited diversity in authorship trends across sub-fields of CS research. Our
results show collaboration trends in conferences and journals of major CS subfields. Such collaborative patterns benefit the
funding agency, policymakers, scientific community, and researchers to plan and execute their research.