Research on opinion polarization by big data analytics capabilities in online social networks

Opinion polarization in online social networks causes a lot of concerns on its social, economic, and political impacts, and is becoming an important topic for academic research. Based on the system theory, a theoretical framework on analyzing opinion polarization combining big data analytics capabilities (BDAC) is proposed. A web crawler is used to collect data from the Sina Weibo platform on the topic of “Tangping”. Concerning the characteristics of the big data environment, social network analysis (SNA), machine learning, text clustering and content analysis are used to mine opinion polarization of “Tangping” on Weibo. Results show that social network users holding the same opinion indicate the phenomenon of aggregation. Although no influential users support the opinion of “Tangping” on Weibo, a high percentage of people advocate the idea. The supporting group has the most clusters while the opposing group has the highest density of keywords. The research contributes to the existing literature on applying BDAC to analyze online polarization from the perspective of the system from user behavior and interaction to topic clustering and keywords identification. The conceptual system framework shows superiority in the integration of information coordination of microsystem and exosystem. Guidance strategies are put forward to supplement the formation theory of opinion polarization and provide suggestions to reasonably regulate network group polarization. • This paper develops a big data-driven conceptual model based on system theory to investigate the mechanism of opinion polarization on Weibo. • Big data analytics capabilities (BDAC) approaches are used for sentiment analysis, community detection and topics identification on opinion polarization. • This paper aims to understand, measure and quantify online opinion polarization on the controversial issue and put forward guidance for opinion management.