The Research Of Large-Scale Social Network Visualization Method

Visualization analysis is an important method for social network analysis. Through social networking visualization display, people can obtain clear intuitive understanding of not only the structure of the network, as well as the link between nodes. And through visual processing and the user's interaction, users can concentrate on the local network or the focus on the linkages of nodes and get into targeted analysis, research, and thus go on a more in-depth and effective exploration and discovery for social networks. However, with the development of society, social networks have been expanding, how to display large-scale network data clearly and effectively became the focus of visualization analysis of social network.In this paper, we first study the layout visualization algorithms of large-scale network in a more detail, and upon original classical model based on force-directed layout algorithm practise the idea of approximate calculation introduced in Nbody problem. The improved layout algorithm, through experimental comparison, was proved efficient and better for displaying of large-scale network. However, as the limit of the screen size and capabilities of human visual recognition, the size of the original network showed directly is limited. By the research on existing multilevel technique and experimental comparison, constructing a tree-based multilevel display model, this paper presents hierarchical model-based visualization prototype system. By multilevel processing and user interaction, the system can not only display the large-scale networks effectively, and based on levels of network analysis, users can interact directly, to obtain the network information of interest.