Online news diffusion dynamics and public opinion formation: A case study of the controversy over judges’ personal opinion expression on SNS in Korea

Abstract

On the Internet, the line between news and opinion becomes blurred in terms of content and form. In this sense, the processes of news diffusion on the Internet and online public opinion formation are integrated. This study selects one Korean example about the private tweet of a judge about government social network service (SNS) regulation and analyzes citation networks from one news/opinion to another news/opinion. By relying on social network analysis (SNA) methods, this study examines how news/opinions diffuse and how they interact and influence each other's ideological frame within agencies sharing the same ideology and between agencies whose ideologies contrast. The results show that public discourse on the Internet is clearly polarized and fragmented along political ideological lines. Additionally, the findings show that the winner is the conservative bloc because it succeeds in achieving within-bloc solidarity and ensures the resonance of news frames with the cultural values of Korean society. Implications for these findings in modern democracy are also discussed.

Keywords:

• News diffusion
• Public opinion process
• Homophily
• Fragmentation
• Polarization
• ERGM

Notes

1 In this study, social network analysis refers to a statistical approach that will be introduced with ordinary terms in Section 3. Social network analysis in this study does not exclusively refer to analysis of specific social networks formed by social actors or organizations in social settings, as attempted in prior studies, such as Cnaan, Kang, Kim, and Lee (2014).

2 Some readers may be curious why Pearson's correlation was used instead of popular inter-coder reliability like Krippendorff's alpha when evaluating the consistency between machine coding and human coding. This study relied on Pearson's correlation for the sake of simplicity.

3 Except in a few cases, it was impossible to track the detailed profiles of bloggers, which should be noted as a potential limitation of this study.