Economic Complexity in Mono-Partite Networks

TL;DR

This paper extends the concept of economic complexity from bipartite graphs to general graphs, introduces new centrality measures, and uncovers the underlying cost functions, broadening its applicability across various network types.

Contribution

It presents a novel framework that generalizes economic complexity to any graph structure and introduces fitness and orthofitness centrality measures.

Findings

Extended economic complexity to all graph types.

Introduced fitness and orthofitness centrality measures.

Identified cost functions underlying complexity algorithms.

Abstract

Initially designed to predict and explain the economic trajectories of countries, cities, and regions, economic complexity has been found applicable in diverse contexts such as ecology and chess openings. The success of economic complexity stems from its capacity to assess hidden capabilities within a system indirectly. The existing algorithms for economic complexity operate only when the underlying interaction topology conforms to a bipartite graph. A single link disrupting the bipartite structure renders these algorithms inapplicable, even if the weight of that link is tiny compared to others. This paper presents a novel extension of economic complexity to encompass any graph, overcoming the constraints of bipartite structures. Additionally, it introduces fitness centrality and orthofitness centrality as new centrality measures in graphs. Fitness Centrality emerges as a promising metric for assessing node vulnerability, akin to node betweenness centrality. Furthermore, we unveil the cost functions that drive the minimization procedures underlying the economic complexity index and fitness centrality algorithms. This extension broadens the scope of economic complexity analysis, enabling its application in diverse network structures beyond bipartite graphs.