Research On Community Model And Structure Of Complex Networks

In the real world,many complex systems can be expressed by the network, such as the Intemet networks,collaboration networks,metabolism networks,aeromagnetic networks and so on.In the network,a node is the individual and the edges between the nodes can expresse the relationship between individuals.The focus in this dissertation is community stucture which is an important characteristic in real complex networks,specifically in the complex social networks. In the community network model of exploration and study, this article will be interested in network nodes to give identity and expression using vector space model, the new node with interest as the similarity of characteristics of the network nodes with edges already in the choice of a good assessment.In the community network model of exploration and study, this article will be interested in network nodes to give identity and expression using vector space model, the new node with interest as the similarity of characteristics of the network nodes with edges already in the choice of a good assessment.Numerical analysis shows that scale-free network model has a characteristic, BA scale-free network models are only a special case of this model; also interested in features similar to the close connections between the nodes, interests and characteristics of the larger difference between the sparse connections between the nodes, which has obvious the structure of societies. Compared with existing network models, a society based on interest features of preferred network model rules are simple, is a natural evolution of societies in the network model.On the other hand, find and analyze organizations contribute to a better understanding of network structure, efficient and accurate algorithm to detect communities in complex networks is an important research challenge. In this paper, the starting node potential of association proposed a new detection algorithm, analog physical midfield and potential of the concept, the algorithm will be in cyberspace as a node in the level of potential value of the tightness of the reaction, was connected surrounded by a low potential zone various high-potential areas as a reasonable division of the network society. Compared to traditional detection algorithm for solving the entire network without the need of the IL, and therefore the algorithm has better performance and more natural division of the result.In addition, the community network survivability research in the military field has its special status, this study shows that the network of community detection analysis to give priority to communities in the network attack "bridge" has a better damage effects, which the network Protection provides a new way of thinking.