Aristotle University Of Thessaloniki (AUTH)
Applications of statistical physics theories on research and innovation networks
Abstract
dc:descriptionIn the current dissertation, research and innovation networks are analyzed using theories derived from Statistical Physics and Network Theory. More specifically, their structure and evolution are examined, and an effort is made to explain their changes and transitions. The first network examined is a patent citation one, with patents registered from 1978 to 2016 in the European Patent Office and the Patent Cooperation Treaty. Its structure is analyzed by calculating the network’s degree distribution and the application of the k-shell decomposition method. What it does is to decompose the network into shells and identifying the most important nodes for the network’s structure. Next, a geographic and thematic analysis is conducted on the shells. The evolution of the network is also examined using percolation theory, where changes in the transition threshold are observed. In order to create networks with similar behavior and characteristics and to provide a plausible explanation for the observed changes a hybrid model is proposed. This hybrid model is a combination of the well-known models of Erdős–Rényi and Barabási–Albert. Next, a multiplex network is examined, which consists of two layers. The first one represents the inventor collaborations for the creation of patents registered in the European Patent Office, and the second layer represents the scientific collaborations in Framework Programmes 5-7 and Horizon 2020. The network nodes are the geographical regions of the inventors/scientists and a link between them exists, when two of them from different regions collaborate. For the structural analysis of the network, the notion of multilinks will be used, which is an indicator of node connectivity among the layers of a multilayer network. Their evolution is examined which exhibits similar behavior through the years, thus, we propose a kinetic model which consists of 3 differential equations with 6 parameters for its description. The equations are then fitted to the real data in order to acquire the parameter values. Finally, the same multiplex network is examined, using the notion of triangles, an indicator for the formation of three-way collaborations. The network is divided into various multiplex sub-networks, whose number of triangles is calculated for the two layers. In addition, the links of the real data are randomly shuffled, by preserving the network properties, and the same procedure is repeated to identify whether there is a mechanism that promotes non-random behavior. The same analysis is also conducted by calculating the clustering coefficient, which contains the notion of triangles.
Degree
thesis:*- Grantor dc:publisher
- Aristotle University Of Thessaloniki (AUTH)
- Year dc:date
- 2020
Author and committee
dc:creator, dc:contributor.*- Authors dc:creator
-
- Angelou, Konstantinos
- Αγγέλου, Κωνσταντίνος
Subjects
dc:subject × 10Rights
- Language dc:language
- eng
Identifiers
dc:identifier.*- Identifier
- 10.12681/eadd/48462