Logics for Personalized Announcements and Attention Dynamics

Access to information online usually undergoes double filtering: first, by the built-in algorithms of the platform in use, which select and propose information that is tailored to each user's preferences, and second, by the cognitive capacity and attention resources of the users themselves, which only allow the agent to receive a portion of the already personalized incoming information.Here, we introduce a framework based on dynamic epistemic logic, where announcements are personalized by semantically specified filtering conditions and are only received by the agents if their attention resources allow it.To achieve this, we first introduce opinion models to represent agents' opinions on topics and special edge-conditioned action models to more compactly represent the access of agents to personalized announcements.Then, we extend this framework to encompass attentive limitations as well.It is shown that the proposed logics are sound and complete with respect to the underlying classes of models.