Modelling transition phenomena of scientific coauthorship networks

Abstract

In a range of scientific coauthorship networks, transitions emerge in degree distribution, in the correlation between degree and local clustering coefficient, etc. The existence of those transitions could be regarded because of the diversity in collaboration behaviors of scientific fields. A growing geometric hypergraph built on a cluster of concentric circles is proposed to model two specific collaboration behaviors, namely the behaviors of research team leaders and those of the other team members. The model successfully predicts the transitions, as well as many common features of coauthorship networks. Particularly, it realizes a process of deriving the complex “scale‐free” property from the simple “yes/no” decisions. Moreover, it provides a reasonable explanation for the emergence of transitions with the difference of collaboration behaviors between leaders and other members. The difference emerges in the evolution of research teams, which synthetically addresses several specific factors of generating collaborations, namely the communications between research teams, academic impacts and homophily of authors.

Modeling the coevolution between citations and coauthorship of scientific papers

“…1). The combination makes the new provided model not only reproduce a range of statistical features of empirical coauthorship and citation networks (which have captured by our previous models [30,32]), but also predict some features generated by the interactions between coauthorship and citations, e.g. the distribution of citations per author, self-citation, etc.…”

“…The first part of the model is a generalization of the coauthorship model in Reference [30]. The parameter p in Step 1.b is newly introduced to tune the average number of papers per author.…”

“…4). With some modifications, the analysis and calculations in References [25,[30][31][32][33] can be employed to show how the model works, which are described in Appendix. Here an intuitionistic explanation is given as follows.…”

“…However, the respectable model-data fit after data binning is still impressive to us because the model involves no true free parameters to tune P analysis. In addition, a similar analysis has been applied to P AA (k) [30].…”

“…For each kind of indexes, split authors into two non-overlapped parts by the algorithm in Reference [30] (Table 6 in and then publish more papers to receive more citations. Therefore, our model successfully reproduces the power-law tails in the distributions of those indexes (Fig.…”