The study examines how far-right and fringe online groups share news media articles to advance their narratives. It distinguishes between far-right and moderate publishers based on content, writing style, and sharing patterns by extreme online users. The findings suggest that misinformation consumption differs from production, with users preferring certain styles over content.
The research delves into the landscape of Australian news media, comparing ideological leanings in terms of topic coverage, trust index, and linguistic patterns. It identifies significant differences between left-leaning and right-leaning articles but minimal practical impact due to small effect sizes. The study also explores the classification of extreme online groups based on linguistic styles using LIWC, GRIEVANCE dictionary, and STYLOMETRIX.
Furthermore, the analysis extends to identifying different writing styles within extreme groups like far-right and antivax communities on Facebook. The results show that stylistic features can outperform content-based classifiers in distinguishing between these groups. Additionally, the study investigates distinguishing far-right production from consumption based on stylistic features, revealing distinct patterns between the two.
Overall, the research sheds light on how styled messages play a crucial role in spreading misinformation online and emphasizes the importance of understanding user preferences for effective intervention strategies against misinformation dissemination.
Naar een andere taal
vanuit de broninhoud
arxiv.org
Belangrijkste Inzichten Gedestilleerd Uit
by JooYoung Lee... om arxiv.org 03-14-2024
https://arxiv.org/pdf/2403.08391.pdfDiepere vragen