Scientific collaboration networks are a prime example of complex system. If we create a graph in which nodes represent researchers and edges indicate a collaboration among them, we obtain a complex network exhibitting very specific properties emerging from the particular interaction dynamics in the field under scrutiny.
We have conducted such a study on the community of researchers in computational intelligence and games. Using data grabbed from the DBLP, we have constructed and analyzed the resulting network from a dynamic perspective, studying how it evolves in time when considering either its cumulative state or a moving time-frame. This analysis has been conducted at different scales, from the macroscopic to the microscopic, and has paid special attention to issues such as the growth dynamics of the network (driven by preferential attachment), the collaboration patterns and the community structure. The network is shown to be at an early stage of development, with percolation starting to give rise to a large giant component. Comparisons have been also drawn to other related collaboration networks. Our findings have been published in Physica A.