<![CDATA[Recommendation algorithm architectures and content management policies on social media platforms play a crucial role in shaping public information consumption patterns, yet frequently trigger social division. This study aims to analyze the interaction between artificial intelligence-based recommendation systems, content moderation governance, and their impact on escalating social polarization. Employing a qualitative approach with a critical discourse analysis design, data were gathered through platform policy document reviews, transparency reports, and in-depth interviews with media sociology experts and digital regulators. The findings demonstrate that recommendation algorithms consistently prioritize high-emotion content to maximize user retention, directly isolating the public within homogeneous information filter bubbles. The absence of transparency in automated curation criteria and the limitations of content moderation accelerate the spread of misinformation and reinforce group bias. This study concludes that restructuring platform governance through algorithmic accountability and external regulation represents a crucial step in mitigating social division risks within virtual spaces.]]>
Copyrights © 2026
