Study of impact of seed selection based on centrality and information from overlapping communities on complex networks
complex networks
influence
seed selection
overlapping communities
Some scenarios where an infection, opinion or even ideas spread, share important characteristics; they need, for example, individuals involved in a social context to spread. Social networks have been widely used to model and simulate these processes. It is necessary to study its structure so that the choice of these individuals can guarantee the optimization of the diffusion process. This work focuses on comparing the scope of diffusion in two contexts. The first selects individuals based on measures of centrality, while the second chooses individuals using criteria based on overlapping communities in different ways. A broad comparison was made under the diffusion model: Independent Cascading Model and Threshold Model. By varying the parameters of the models, the results showed that in some scenarios the use of overlapping communities can bring improvements in the reach of diffusion.