Dynamic Scientific Co-Authorship Networks

Abstract

AbstractScientific collaboration networks have been studied systematically since 1960 by scholars belonging to various disciplinary backgrounds. As a result, the complex phenomenon of scientific collaboration networks has been investigated within different approaches. Although the term “scientific collaboration network” has different connotations in the literature, we use the term more narrowly to focus on scientific collaboration resulting in co-authored public documents. We broaden this beyond journal articles to include many types of scientific productions in addition to journal articles and books. We insist that these productions are public items available in each field. In this chapter, we focus on the main quantitative approaches dealing with the structure and dynamics of scientific collaboration networks through co-authorized publications. We provide a brief history of social network analysis that serves as a foundation. We further review earlier conceptual classifications of co-authorship networks and distinguish cross-disciplinarily, cross-sectoral and cross-national levels. We couple the newer ideas of “small world” models and “preferential attachment” to older sociological conceptions of scientific collaboration. This is followed by descriptions of deterministic and stochastic models that have been used to study dynamic scientific collaboration networks. We stress the importance of delineating the topology of collaboration networks, understanding micro-level processes and then coupling them. We conclude by outlining the strengths and limitations of various modeling strategies.KeywordsRandom GraphSocial Network AnalysisPreferential AttachmentCitation NetworkScientific CollaborationThese keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.