Research And Implementation Of Dynamic Mobile Community Discovery Method Based On The Mobile User Behaviors

It becomes a core issue in network analysis that community discovery helps to solve the problem of personalized mobile user needs. In recent years, the research about community, which has been widely used in biology, sociology, physics, computer and mobile fields. Since the mobile social networking topologies are dynamic and the communities are overlap among the friends, colleagues and family. Considering the dynamic and overlapping of mobile networks, there are three aspects of the study of dynamic mobile network community discovery:(1) An Overlapping Community Discovery Algorithm in Mobile Communications Network. Taking the characteristic of overlapping and the features of mobile networks into account, this study found an overlapping community discovery algorithm combine with idea of traditional methods. Because many traditional community discovery algorithms can’t find overlapping communities, our research allows overlapping in mobile network. There are differences between the traditional network and mobile network. So the main point of this research is the study of feasibility and effectiveness of overlapping community discovery algorithm.(2) A Dynamic Community Discovery Algorithm Based on Incremental Mobile Network. Based on the mobile users’behaviors, a new dynamic community discovery algorithm is proposed to solve the community discovery issues in mobile communication network, which is combined with the idea of traditional dynamic community discovery methods. Dynamic mobile social network model is constructed by this algorithm, and it gets the communities kernel nodes in current time according to the history information of network and four basic changes showed in current time; and then partial satisfied communities kernel nodes are merged to get the current local community, as partial satisfied nodes are added to the local community, community structure in current time is discovered; finally, this algorithm is proved to be feasible and effective though experiments of the public and simulation data sets.(3) A Dynamic community discovery algorithm Based on Timeliness In Mobile Network. Based on first and second points on research, a new dynamic timeliness community discovery algorithm is proposed to solve the community discovery issues in mobile communication network, which is combined with the idea of traditional dynamic community discovery methods. Most studies have focused on static network, which ignores the dynamic in mobile network. Taking the historical information of the time slice network snapshot into account, we use a forgetting factor, the far from the current time, it gets the lower the weight; the closer to the current time, it gets the higher weight. So considering the evolution of community and the stable historical information of community, the algorithm is proved to be feasible and effective though experiments of the public and simulation data sets.