We propose a stochastic agglomerative algorithm to detect the local community of some given seed vertex/vertices in network. Instead giving deterministic binary output, our method assigns every vertex value that is probability this particular would be seed. The proposed procedure has several advantages over existing algorithms, including avoiding random tie-breaking, evaluating uncertainties, d...

Community detection in networks is the process by which unusually well-connected sub-networks are identified–a central component of many applied network analyses. The paradigm modularity quality function optimization stipulates a partition network’s vertexes that maximizes difference between fraction edges within communities and corresponding expected if were randomly allocated among all vertex...