Recognition of opinion leaders in blockchain-based social networks by structural information and content contribution

Abstract

Exploring the influence of social network users in the blockchain environment and identifying opinion leaders can help understand the information dissemination characteristics of blockchain social networks, direct the discovery of quality content, and avoid the spread of rumors. Members of blockchain-based social networks are given new responsibilities by token awards and consensus voting, which alters how users connect to the network and engage with one another. Based on blockchain theory and the relevant theories of opinion leaders in social networks, this article combines structural information and content contributions to identify opinion leaders. Firstly, user influence indicators are defined from the perspective of network structure and behavioral characteristics of user contributions. Then, ECWM is constructed, which combines the entropy weight method and the criteria importance through intercriteria correlation (CRITIC) weighting method to address the correlation and diversity among indicators. Furthermore, an improved Technique for Order Preference by Similarity to Ideal Solution (TOPSIS), called ECWM-TOPSIS, is proposed to identify opinion leaders in blockchain social networks. Moreover, to verify the effectiveness of the method, we conducted a comparative analysis of the proposed algorithm on the blockchain social platform Steemit by using two different methods (voting score and forwarding rate). The results show that ECWM-TOPSIS produces significantly higher performance than other methods for all selected top N opinion leaders.