How to Mitigate Misinformation Diffusion on Blockchain‐Based Decentralized Social Network Platforms? Insights From an Agent‐Based Simulation Model
通过对比去中心化平台Steemit和中心化平台Pokec的智能体仿真实验，发现无缓解机制时去中心化平台错误信息影响更多用户、传播更快、距离更短；而用户标记、文章评分和声誉三种机制能有效缓解传播。
ABSTRACT Although the spread of misinformation on centralized social media platforms has received significant attention, few studies compare centralized and decentralized platforms, and how to mitigate misinformation diffusion in newly emerging blockchain‐based decentralized social network platforms. We study misinformation diffusion between the decentralized and the centralized platforms and identify three decentralized governance mechanisms to mitigate the spread of misinformation in decentralized networks: user flagging, user article ratings, and user reputation. Our empirical experiments using agent‐based simulations leveraging real‐world data from two platforms reveal two findings. First, comparing misinformation diffusion between the decentralized Steemit and the centralized Pokec platforms suggests that in the absence of any mitigation mechanisms, misinformation in decentralized platforms affects more users, at faster rates, and reaches shorter distance from the misinformation initiating user. Second, within the decentralized platforms, misinformation affects fewer users, at slower rates, and reaches shorter distance from the misinformation initiating user in the presence of mitigating mechanisms than in their absence. We discuss the implications of these results both for the understanding of misinformation diffusion and for the governance of decentralized social network platforms.