A belief revision method based on argumentative dialogue model

摘要

Belief is Agents' cognition of surrounding. Due to the dynamics of the environment, the new received belief may be conflicted with the original belief set when an agent got new information. Belief revision is to make sure the new belief set is a conflict-free set in this moment. This paper proposes a combined with debate way of belief revision, which use certainty-factor based argumentative framework to represent the uncertainty of knowledge and uncertainty reasoning, to get a belief set with the corresponding certainty factor. Finally the result belief set will be received by a setting certainty-factor threshold value. Proved by the experiment, the new accepted belief may not be included in the revised credible belief set, and credible belief set is conflict-free. Belief revision algorithm, therefore, have broken the traditional way of belief revision, and effectiveness.