Density-Based Clustering of Social Networks

Abstract The idea of the modal formulation density-based clustering is to associate groups with regions around modes probability density function underlying data. correspondence between clusters and dense in sample space here exploited discuss an extension this approach analysis social networks. Conceptually, notion high-density cluster fits well one community a network, regarded as collection individuals local ties its neighbourhood. lack probabilistic networks turned into strength proposed method, where node-wise measures that quantify role actors are used derive different configurations. allows for identification hierarchical structure clusters, which may catch degrees resolution structure. This feature nature networks, disentangling involvements aggregations.