Agents in a network want to learn the true state of the world from their own signals and their neighbors' reports. Agents know only their local networks, consisting of their neighbors and the links among them. Every agent is Bayesian with the (possibly misspecified) prior belief that her local network is the entire network. We present a tractable learning rule to implement such locally Bayesian learning: each agent extracts new information using the full history of observed reports in her local network. Despite their limited network knowledge, agents learn correctly when the network is a social quilt, a tree-like union of cliques. But they fail to learn when a network contains interlinked circles (echo chambers), despite an arbitrarily large number of correct signals.

中文翻译：

---

网络中的局部贝叶斯学习

网络中的代理希望从自己的信号和邻居的报告中了解世界的真实状态。代理只知道他们的本地网络，包括他们的邻居和他们之间的链接。每个代理都是贝叶斯（可能是错误指定的）先验信念，即她的本地网络是整个网络。我们提出了一个易于处理的学习规则来实现这种本地贝叶斯学习：每个代理使用她本地网络中观察到的报告的完整历史来提取新信息。尽管他们的网络知识有限，但当网络是一个社会被子，一个像树一样的集团联盟时，代理可以正确学习。但是，尽管有任意大量的正确信号，但他们无法了解网络何时包含相互关联的圆圈（回声室）。