Abstract:

High-impact users of microblog play an significant role in information recommendation, marketing, public opinion analysis and other areas. Aiming at the existing approaches to users’ influence only considering the network topology and taking the users of microblog as the entry point, a new model for calculating users’ influence is proposed-UIRank model. This model gives adequate consideration to the interaction between users and determines the distribution of the value of UIRank according to the weight difference of different user behaviors. It is proved by the results of the experiments that the method for calculating the users’ influence proposed is more accurate and efficient.

Key words: social networks; user interaction behavior; network topology; user influence; influence algorithm