Research On The Dissemination Subject Of Microblog Public Opinion Based On Social Network Analysis

With the drastic development of the Internet, more and more communication and emotional expression in people’s daily life can be reflected by the Internet. However, many of the consequent network security problems caused by communication and expression have already affected people’s life and social stability. There is urgent need to analyze and research the subjects of public opinion in order to find the core members, understand the state of public opinion development and dissemination, guide virtuous cycles of public opinion hotspots, and reduce the impact of harmful information to society.Social Network Analysis (SNA) is a structural research and analysis method, which studies individuals and the relations between individuals to effectively analyze and excavate the nodes residing in core spots by utilizing a series of centrality algorithms. However, there are plenty of "zombie fans" (fake followers) and "water army" (paid posters) in microblogs of real life, whose corresponding nodes in social network have an important significance on computing importance degrees of nodes although they seldom have real meanings. So it is necessary for us to deeply explore more effective approach to computing importance degrees of nodes.This paper presents a novel and practical algorithm for computing and raking importance degrees of network members based on Social Network Analysis by deeply exploiting the structure characteristics of social network consisting of microblog users. The specific contents of our research are as follows. First, the concept called "Node Attribute Value" (NAV) is presented as an attribute value of a node to measure whether a node is good or not by using calculating the relevant numbers of comments, forwards and likes to the node. Second, the concept called "Node Fans Influence Value" (NFIV) is proposed to calculate the influence value of fans of a node, which is determined by the NAVs of all fans of the node. Next, we improve the existing Point Centrality Algorithm in SNA. In this paper, the importance degree of a node is determined by the in degree of the node. Our approach based on influence values of fans of nodes aims to handle the inexact calculation of node important degree such that the analysis results can satisfy the true situation in microblog dissemination subject analysis. At last, several experiments are conducted to evaluate our approach. The experimental results show that our algorithm can improve the analysis precision of dissemination subjects within a relatively low time cost.