通过噪声缓解社交网络中的虚假信息
Mitigating Disinformation in Social Networks through Noise
AI总结：
本研究通过随机机器人和随机推荐两种噪声注入策略，发现噪声能提升社交网络智能体对基本事实的信念，且在虚假信息强大时效果与辟谣相当，但可能引发认知失调。
摘要：
大量文献表明，向复杂社会经济系统中注入噪声可以提高其韧性。本研究旨在探究这一结论是否同样适用于社交网络中的信息传播情境。具体而言，我们旨在理解，在一个由智能体组成的社交网络中，当智能体试图在一组相互竞争的假设中发现一个基本事实时，向该网络注入噪声是否能够增强其对虚假信息的抵御能力。我们实施了两种不同的风格化策略来向社交网络注入噪声，即通过随机机器人和随机推荐，并发现这两种策略都能提高群体对基本事实的整体信念。值得注意的是，我们发现当虚假信息特别强大时，噪声的效果与辟谣相当。另一方面，这种有益效果可能导致智能体私下持有的信念与其公开表达的信念之间出现不一致，这一现象让人联想到认知失调。
英文摘要：
An abundance of literature has shown that the injection of noise into complex socio-economic systems can improve their resilience. This study aims to understand whether the same applies in the context of information diffusion in social networks. Specifically, we aim to understand whether the injection of noise in a social network of agents seeking to uncover a ground truth among a set of competing hypotheses can build resilience against disinformation. We implement two different stylized policies to inject noise in a social network, i.e., via random bots and via randomized recommendations, and find both to improve the population's overall belief in the ground truth. Notably, we find noise to be as effective as debunking when disinformation is particularly strong. On the other hand, such beneficial effects may lead to a misalignment between the agents' privately held and publicly stated beliefs, a phenomenon which is reminiscent of cognitive dissonance.
机构、总结、中文摘要均由AI分析而来，可能存在错误，请以论文正文为准。
