earticle

영어

With the development and popularization of the Internet, the different microblog topics are disscussed everyday, so the microblog can product a large number of various topics,which can reflect the influence of different users in a given topic. In the microblog topics,the key users of microblog topics are found by discovering the influential sensitive topics and calculating the influence value of the user,which are the focus of attention in the fields of microblog public opinion supervision and safety management. In order to accurately measure the influencers in the given topic and to calculate the user influence value, the thesis proposes the method of constructing the propagation network which is based on attention and forward relationship between users, and then proposes TDN-If algorithm by using PageRank algorithm. When we calculate the transition probability in propagation network by using the TDN-If algorithm, the information propagation is considered to measure influencers. This method can resolve the defects of discovering the user's influence by only using followers of this single indicator in the current microblog topics. The experimental results show that the TDN-If algorithm has important theoretical and practical value, which is better than TwitterRank algorithm and other influential individuals found algorithm. Thus, the method proposed in this paper can not only effectively solve the problem about discovering and persuading the key users in the sensitive topics who have influences and have the unique insights on the significant events, for example, which can provide the strong guarantee for the governments in the fight against terrorism, but also provide the important theory and method for the complex network community discovery, microblog public opinion supervision , microblog safety management and so on.