Dynamic Multi Level Approach for Community Detection

摘要

A dense connection within and a sparse connection between is what is assume right for a definition of community. It has been an existing research aim for most researcher recently in detecting community due to this definition. This paper proposes a new way of group the community is to give priority on the structure of the network for its community, rather than arbitrary addition of members with its only indicator is based on the value of modularity. Experiment and comparison of end result of found community shown a promising outcome. Hence, the new algorithm, MuLAN, is more robust in providing the detection where it forms the basic group of members as its first level of community and the it will check whether the remaining members are also connected with each other which form a strong structure for the community.