Statistically Validated Networks in Bipartite Complex Systems

Abstract:     
Many complex systems present an intrinsic bipartite nature and are described and modeled in terms of networks. Bipartite networks are often very heterogeneous in the number of relationships that the elements of one set establish with the elements of the other set and the heterogeneity makes it very difficult to discriminate preferential links between the elements from randomly occurring links reflecting system heterogeneity. Here we introduce an unsupervised method to statistically validate each link of a projected network against a null hypothesis, which takes into account system heterogeneity. We apply our method to a biological, an economic and a social complex system. Our method is able to detect network structures which are informative about the organization and specialization of the investigated systems. Specifically, our method (i) identifies the preferential relationships between the elements, (ii) highlights the clustered structure of systems, and (iii) defines and classifies links according to the type of statistically validated relationships between the connected nodes.

Number of Pages in PDF File: 34

Keywords: Networks, Complex Systems

JEL Classification: C10, C80

working papers series