算法推荐下社交媒体用户负向情绪扩散机理及规避研究

Diffusion Mechanisms and Mitigation Strategies of Negative Emotions among Social Media Users under Algorithmic Recommendation

摘要: 后信息时代，社交媒体平台借助算法推荐较为有效地实现了用户信息需求的精准预测及投送，用户也实现了信息接收、情绪表达和沟通反馈的传播闭环。本文以情绪传播为视角，探讨了推荐算法辅助下社交媒体用户分别在描述性、对抗性和煽动性话语中的情绪表达，展现出情绪模糊化、两极化和工具化的传播特征；同时，基于SIPS模型，借以情绪传染理论，阐述了社交媒体用户负向情绪的传播机理，包括共鸣、确认、参与和分享四个方面，展现了推荐算法下用户情绪的传播逻辑；并进一步从用户、技术、平台与社会四个维度提出规避措施建议。

Abstract: In the post-information era, social media platforms have effectively leveraged algorithmic recommendations to achieve precise prediction and delivery of user information needs, while users have formed a closed communication loop encompassing information reception, emotional expression, and interactive feedback. From the perspective of emotional communication, this paper explores how social media users express emotions under algorithmic recommendations in descriptive, confrontational, and inflammatory discourses, revealing emotional communication characteristics such as ambiguity, polarization, and instrumentalization. Furthermore, based on the SIPS model and emotional contagion theory, the study elucidates the dissemination mechanisms of users’ negative emotions on social media, including four aspects—resonance, validation, participation, and sharing—thereby clarifying the emotional propagation logic under algorithmic recommendations. Finally, recommendations for mitigation measures are proposed across four dimensions: user, technological, platform, and societal perspectives.