FAPERJ LNCC / INRIA DNET: Complex Dynamic Networks project

PIs: Artur Ziviani, Researcher, LNCC & Eric Fleury, Professor at ENS de Lyon

The main goal of the CoDyN project is to lay solid foundations to the characterization of dynam- ically evolving networks, and to the field of dynamical processes occurring on large scale dynamic interaction networks. In order to develop tools of practical relevance in real-world settings, we will ground our methodological studies on real data sets. One set was collected within the iBird project partially funded by the MOSAR European program. Other sets are based on social blog traces or twitter traces1. In the very active field of complex networks, the development of a large body of research in the last ten years has largely been stimulated by the availability of empirical data, and the increase in computer power needed to analyze these data. This has allowed to point out the similarities in the structures of networks arising in very different fields, and to develop a body of knowledge, tools and methods to characterize and model these networks, and to understand the impact of their structure on the dynamical processes occurring on these networks.

While many interesting questions remain open even for what concerns static networks, the next challenging issues, which the community is starting to address, are given by the generalization of known tools and the development of new methods able to deal with dynamically evolving networks. While some theoretical studies on simplified models have started to appear, important and relevant development has to be based on real empirical data. Such characteristics of the dataset will allow the project teams to tackle important issues such as how to analyze dynamical networks in which the nodes have several internal properties, possibly sampled on different scales, how to determine and assess correlations between these properties. Many complex dynamic networks are often only components of larger systems, where dynamic graphs somehow interact and depend on each other. These larger systems, which are composed by complex networks that interact and depend on each other, are defined as layered complex networks and impose specific challenges to their character- ization, modeling, and analysis. There is therefore a high need for the development of tools and methods for the analysis of the dynamics of graphs and the possible interaction of them in layered complex networks.