A data reduction approach using hypergraphs to visualize communities and brokers in social networks

Abstract

The comprehension of social network phenomena is closely related to data visualization. However, even with only hundreds of nodes, the visualization of dense networks is usually difficult. The strategy adopted in this work is data reduction using communities. Community detection in social network analysis is a very important issue and in particular detection of community overlapping. In this approach, the information extracted from social networks transcends cohesive groups, enabling the discovery of brokers that interact among communities. In order to find admissible solutions in hard problems, relaxed approaches are used. Quasi-cliques are generated, and partition is found using a partial set covering heuristic. The proposed method allows the identification of communities and actors that link two or more groups. In the visualization process, the user can choose different dimension reduction approaches for the condensed graph. For each condensed structure a hypergraph can be drawn, identifying communities and brokers.