Topology of the Co-Authorship Graph in the Field of Physics in Russia
- PDF / 378,893 Bytes
- 5 Pages / 612 x 792 pts (letter) Page_size
- 49 Downloads / 143 Views
logy of the Co-Authorship Graph in the Field of Physics in Russia O. I. Ivanova, A. M. Kovalenkoa, A. V. Kolobova, V. V. Korolevaa, A. V. Leonidova,b,*, and E. E. Serebryannikovaa,b a Lebedev
b
Physical Institute, Russian Academy of Sciences, Moscow, 119991 Russia Moscow Institute of Physics and Technology, Dolgoprudnyi, Moscow Region, 141701 Russia *e-mail: [email protected] Received March 25, 2020; revised June 25, 2020; accepted June 26, 2020
Abstract—Topological properties of the co-authorship graph are considered as applied to the physical studies in Russia for 2012–2018. Objects of the study are two various weighted graphs; in the first one, edge weights correspond to the number of joint papers of two researchers; in the second one, weights are calculated by the technique accounting for the quality of journals in which joint papers are published. Based on an analysis of the PageRank centrality vector, it was shown that ranking of authors in the data of two graphs differs significantly. However, preferentially regularly published researchers annually become leaders in both graphs. Keywords: co-authorship graph, physical sciences, topology, PageRank DOI: 10.3103/S1068335620080060
Topological properties of the co-authorship graph are considered as applied to the physical studies in Russia for 2012–2018, constructed based on information on publication activity of Russian researchers from the WebOfScience database. The study of the co-authorship graph is one of important points of application of the complex network theory to a quantitative analysis of various phenomena of social and economic reality, see e.g. [1]. Extensive studies of the co-authorship graph involving large databases were initiated in [2], and since then are actively continued [3]. The co-authorship graph properties reflect features of one of the most important components of the scientific knowledge generation process, i.e., concentration of efforts of individual researchers for collaboration with the result of publication in a reviewed scientific edition. Let us define the co-authorship graph & t as a graph whose vertices are individual researchers, and edges between vertices arise in the case of joint publications of corresponding researcher pairs for a certain considered time interval, e.g., a year, with index t.1 Of particular interest is also an analysis of initial graph & modifications accounting for the intensity and/or quality of the collaboration of authors of joint publications satisfying the corresponding weighted graph defined below. Below we use the following notations: 1. t is the year for which the graph is considered (in data t = 2012,…,2018); 2. Rt is the number of graph & t vertices, i.e., the number of researchers being coauthors of at least one paper for year t; 3. 5 t = {id1t , … , idtRt } is the set of vertices of graph & t ; 4. & t is the unweighted co-authorship graph whose edge exhibits the existence of an only one joint publication of a researcher in year t; 5. & t = (5 t , %t ), where %t is the set of edges of grap
Data Loading...
