Forgetting What Agents Know: Removing (Positive) Information in Epistemic Logic

Timo Eckhardt (Ludwig Maximilians Universität, München)

Timo Eckhardt (Melbourne) will present "Forgetting What Agents Know: Removing (Positive) Information in Epistemic Logic", at 11 on 30 November in Old Quad G10.

Abstract: In the standard Kripke model representation of epistemic logic it is not possible to model loss of information for agents of the systems. Once something is known it cannot be unlearned. In "Forgetting complex propositions" by Fernandez-Duque et al., a method was introduced that allows us to model forgetting of propositional formulas in a single agent system. The goal of this talk is to show a generalisation of this approach to a) multiple agents and b) positive knowledge formulas, i.e. formulas that do not include negated knowledge statements, i.e. statements that an agent does not know something.