Telecommunication Information Network Construction Based On Individual Attributes And Community Detection

Social network is a kind of basic paradigm for describing the relationships among individuals in society.With the increase of social network data and further research,the analysis of social network becomes more valuable.The global topology,community structure and node topology depict the social network from three perspectives of macro,meso and micro.The community discovery of the network not only helps to understand the overall structure of the network,but also closely related to the analysis of node topology properties.However,the current validation to the results of community discovery does not consider the role of nodes.What is more crucial is that current mainstream community discovery algorithms do not concern the form of the internal structure of the community.As a theme of social network,telecommunication information network has the characteristics of huge data and obvious degree feature.Since the current call information network data is only composed by the caller's call behavior frequently,it can't explain the call particularity of the single node.Therefore,this paper designs a telecommunication information network with individual attribute and a communication detection algorithm which aims to detect the inner structure of communication,in order to solve these two problems in the community discovery and the limitations of the network analysis of current call information.First of all,the paper draw on the analysis results of the call information network in the current social network,correlating a variety of individual attributes which are in the social society dataset and constructing a high-dimensional call information network.From the perspective of node attribute analysis,the paper analyzes the relationship between the individual attributes and the node topology attributes of the call information network.Then according to sociological theory,a core-edge-based community discovery algorithm and node role identification method are proposed.Finally,the paper analyzes this algorithm by combining with a number of real social networks and call information network of this paper.The experimental result shows that the divergence of individual attributes in the call information network can be reflected in the node topology of the social network.For example,there are significant different on degree between gender.In the experiments of ten real network and the telecommunication information network,the community discovery algorithm can discover the community of networks and its internal core-edge structure in most instances effectively,it can also identify the core nodes and bridge nodes in the network accurately.