| Title | : | A Dynamic Modal Logic of Changing Agents and Ascribing Beliefs - Research Proposal Seminar |
| Speaker | : | Shikha Singh (IITM) |
| Details | : | Mon, 6 Feb, 2023 4:00 PM @ Meeting Room 1 (SSB- |
| Abstract: | : | Epistemic logic and its possible worlds semantics has been used by the logic-based AI community to model knowledge and beliefs of agent(s) in multi-agent scenarios. Information changing events and their effects in these systems are modeled with dynamic epistemic logic and its action frame semantics. The existing logic addresses world altering and belief changing actions such as communication and sensing activities limited to facts describing the world. In this seminar, I will discuss our approach to modelling actions that change a set of agents locally for some observing agents and/or globally for the entire model. The observers gain information on the existence of a new identity and need to ascribe them (higher-order) beliefs about the world as well as the existing set of agents. The problem becomes more interesting in the presence of asymmetric awareness about the action among the agents. It also opens up avenues of inventing and communicating fictional agent identities to others. We use action frames from dynamic epistemic logic to describe different views of a single action and extend them further to add or remove agents, we call these agent-update frames. This can be done selectively so that only some specified agents get information of the update, which can be used to model several interesting examples such as private update and deception. The product update of a belief model by an action frame is an abbreviated way of describing the transformed belief model which is the result of performing the action. This is substantially extended to a sum-product update of a belief model by an agent-update frame in the new setting. We show that dynamic epistemic logics, with update modalities - a more complex construct based on agent-update frames, continue to have sound and complete proof systems. Decision procedures for model checking and satisfiability have expected complexity. These ideas are applied to an AI problem of modelling a story in an epistemic planning framework. |
