A reasoning system about knowledge acquisition in bi-agent interaction

摘要

In artificial intelligence, knowledge acquisition and reasoning are the most basic key elements for intelligent agent to obtain the ability of the simulation thinking which capacitate intelligent agent to carry out a series of actions. At present, Dynamic Epistemic Logic (DEL) has been a primary technique for modelling knowledge acquisition and reasoning. This is mainly due to the dynamic epistemic logic has a stronger expression ability to deal with many problems about agent's cognitive reasoning in computer science. In this paper, we distinguish the major sorts of knowledge that the cognitive agent will hold in their interaction in the beginning, and then establish a Knowledge Acquisition System (KAS) which is added a new operator of ???knowledge acquisition??? to extend the basic epistemic language for bi-agent interaction. Other than that, we prove the soundness and completeness of this system, and analyze some properties about it. By this system, it is clear to explain how the agent acquires knowledge that will be transformed into common knowledge by observing the other agent's speech act. The acquired knowledge is further transformed into new private knowledge that provides the strategy options in agent's mind in interaction.