Research Of User Influence Ranking Of Online Social Network

In recent years, with the broad application of Web2.0framework, online social network which allow people to connect each other closely, present yourself and resource-sharing have developed quickly, they generate significant impact on economic and social.While online social network have the advantages that large users population base and increasing web site traffic, they don’t leverage on their advantages to achieve sustainable revenues, the profitable model is still in the exploratory stage.In that context, this thesis focus on the user influence ranking which will support a theoretical basis for the s virtual marking and other business activities on online social network. This thesis devotes to the action and interpersonal relationships on online social network with the quantitative approaches, and it examines user influence in two ways: activity and the characteristics of user interpersonal relationship, the activity demonstrate the user’s real-time state and dynamic influence, but the interpersonal relationship demonstrate user’s implied and static influence, so they present user influence in different aspects. Therefore this thesis builds user influence ranking model based on activity graph and connection graph separately and ranks user influence based on PageRank algorithmic. To meet the needs of practical application of user influence, this thesis presents user integer influence ranking model which combines user activity and interpersonal relationship both two aspects.In the end, this thesis evaluates three ranking algorithms with the online social site Facebook’s dataset, and the experiment result showed that our framework is efficient and accurate.