A trust-based sentiment delivering calculation method in microblog

Abstract

Public sentiment analysis in microblog has become a flourishing frontier in social network big data mining. A trust-based sentiment delivering calculation method is proposed from the view of sentiment delivering and interaction to solve the problem of discovering sentiment propagation mechanism in microblog. Firstly, a trust model of microblog user, which evaluates users' trusts-based on aspects of user past interaction data and composition of integration path, is presented. Secondly, models of microblog sentiment delivering and sentiment delivering interaction were proposed for describing the process of sentiment delivering formally. Further, calculation methods of sentiment delivering effect and sentiment delivering interaction in above two models are given respectively. Finally, experimental results verified the effectiveness and feasibility of our proposed method.