Topical influencers are experts on a specific topic, who always play an important role in the opinion dissemination. From the perspective of influence polarity, topical influencers can be classified into opinion leaders who gain a great deal of support, trolls who are widely condemned, and controversial figures who trigger debate. In this paper, discriminating topical influencers of social networks are addressed. First, the trained BiLSTM-CRF model is constructed to extract emotional elements, and, then, the proposed emotional matching and emotional transforming algorithms are leveraged to obtain the relative emotion of users. Second, the rank of influencers is calculated by the quantified user behavior characteristics and the multi-centrality algorithm. Finally, according to relative emotion of users, the topical influencers are divided into three categories-opinion leaders, trolls, and controversial figures. Experimental results show that the user relative sentiment analysis method proposed in this paper has higher accuracy, and the empirical analysis manifests that the three kinds of influencers affect the opinion distribution in various degrees.
scite is a Brooklyn-based organization that helps researchers better discover and understand research articles through Smart Citations–citations that display the context of the citation and describe whether the article provides supporting or contrasting evidence. scite is used by students and researchers from around the world and is funded in part by the National Science Foundation and the National Institute on Drug Abuse of the National Institutes of Health.