A Framework Analysis of Measuring the Index Node Considering Local and Global Networks.

This paper proposes a solution using multi-index method to evaluate node importance locally and globally with adjusted parameters in the network by considering several evaluation indexes. Especially, it adheres to the principles that the node importance of the network is relevant to the value of the node itself together with the centrality of neighbor nodes from a local point of view. Moreover, this paper considers both spreading information and speed of node in dynamic network. The chief influence on adjusted parameters in different network has been estimated. We can rank the influence of users and analyze their tags of interest from the result. In addition, we can make valuable classification of users and improve the service for more personalized services, which is very important for data mining.