A Research On Logical Omniscience Based On Epistemological Analysis

Logical omniscience is an important problem in cognitive logic which needs to be solved,and is the manifestation of the difference between the modeling of epistemic logic and the actual cognition of rational agents.The knowledge of rational agents is locality,weak inconsistency and dynamic development.Locality shows that the agents does not know all the consequence of his belief set,which also shows the counterintuition of logic omniscience problem.Weak inconsistency shows that the agents may not be aware of the implied inconsistency,he may hold beliefs p and q,where p can lead to ～q,but he is not aware of such implied inconsistency.The agent constantly learns,receives new information,revises and improves his knowledge,so the agent’s knowledge develops dynamically.This paper holds that modeling these important properties and making agents active may help solve logic omniscience.Then,this paper studies the mainstream solutions to logic omniscience,including Konolige’s deductive model,the awareness logic of Levesque and Fagin,and the scheme of impossible worlds.This paper divides these solutions into four categories:starting from modeling locality,increasing weak inconsistency of agent’s knowledge,increasing both locality and weak inconsistency of agent’s knowledge,and modeling the dynamic development characteristics of agent’s knowledge.Finally,this paper introduces dynamic awareness logic,which can not solve the logic omniscience problem completely,but can be regarded as a direction to solve the logic omniscience problem.