Structure, Vulnerability, and Dynamics in Socio-Economic and Ecological Networks

The applicability of state-of-the-art network algorithms is of- ten constrained by specific network architectures, limiting our understanding of complex systems. This thesis addresses two such challenges. First, we tackle the critical limitation of the influential Economic Complexity Index (ECI) and Economic Fitness and Complexity (EFC) algorithms, which are restricted to purely bipartite networks. The method involves introduc- ing a formal generalization of these algorithms, extending their framework to monopartite networks. The principal find- ing is a novel centrality measure, termed fitness centrality, which identifies “crucial” nodes that serve as essential hubs for a network’s dependent members. To demonstrate its util- ity, this framework is applied to ecological food webs, suc- cessfully characterizing species by both their systemic im- portance and their vulnerability to extinction. Second, we address the inadequacy of single-layer analyses for model- ing multifaceted socio-economic interactions. Our procedure is to construct and analyze a multilayer network of Euro- pean regional flows, integrating data on investment, migra- tion, and other interactions. The analysis reveals hidden in- terdependencies, identifies versatile regional hubs invisible to single-layer perspectives, and uncovers functionally inte- grated communities. This approach provides a more holis- tic and structurally accurate understanding of the European economy. Together, these contributions advance network sci- ence by developing more versatile analytical tools and ap- plying them to provide deeper insights into both ecological stability and regional economics.