Scientific Collaboration in a Multidisciplinary Organization Revealed by Network Science
- PDF / 1,845,944 Bytes
- 11 Pages / 595.276 x 790.866 pts Page_size
- 58 Downloads / 193 Views
ORIGINAL RESEARCH
Scientific Collaboration in a Multidisciplinary Organization Revealed by Network Science Ivan Bergier1 · Patrícia Menezes Santos2 · Andreia Hansen Oster3 Received: 16 October 2020 / Accepted: 5 November 2020 © Springer Nature Singapore Pte Ltd 2020
Abstract Multidisciplinary scientific organizations have sought to face the challenges of digital transformation through new governance models that optimize network collaboration and innovation. We studied the collaboration network from the long-term coauthoring system of a Brazilian multidisciplinary organization (Embrapa). The study shows that nodes degree distribution of the network is scale free and degree correlation analysis suggests a disassortative regime from competition and minimal but sufficient control that emerges as a hub-and-spoke pattern. The jobs of controller and researcher are twice as many occupied by males, except for the jobs of analyst, who act like network gatekeeper. With the largest number of individuals in product units, the southern region of the country is more likely to form clusters. Alternatively, hubs in thematic and ecoregional units in the Midwest have greater gravitational attraction, positioning themselves in the inner core of the giant component. The optimization of innovation by the organization should combine greater individual autonomy through improved human capital, with a universal labeling of units as, for instance, centers of innovation. Keywords Business intelligence · Innovation process · Hub-and-spoke · Organization control
Introduction Studies of scientific collaboration have fertile grounds on the principles of network science [1, 2]. Emerging from statistical physics and the science of complexity [1–4], network science is a transversal discipline providing the theoretical bases for studying and modeling real systems with empirical data [5]. Vertices (nodes) and links (edges) constitute a network, in which nodes with many edges have high-degree k. The distribution of values of k is important because the structure or the anatomy of a network reflects its internal Electronic supplementary material The online version of this article (https://doi.org/10.1007/s42979-020-00393-8) contains supplementary material, which is available to authorized users. * Ivan Bergier [email protected] 1
Embrapa Pantanal, Brazilian Agricultural Research Corporation, Corumbá, MS, Brazil
2
Embrapa Pecuária Sudeste, Brazilian Agricultural Research Corporation, São Carlos, SP, Brazil
3
Embrapa Uva e Vinho, Brazilian Agricultural Research Corporation, Bento Gonçalves, RS, Brazil
dynamics of evolution and affects important functions like the dissemination of information/disinformation and the resistance to failure [1, 3, 6]. Barabási and Albert [7] have shown that heavy-tailed degree distributions p (k) are emergent properties of stochastic growth models. New nodes continuously attach themselves to existing network nodes with probability proportional to k of the target node [8]. Observed in many empir
Data Loading...
