Optimizing The Static Non-revision Approach To Clausal Belief Set

In the process of people knowing the world, they will accept new knowledge ceaselessly, and people’s cognitive states have changed with it as well. Otherwise, some of the new knowledge will strike the original one, at which time the original knowledge base needs to be updated. Through revising the original belief states to obtain a new consistent belief is just the meaning of belief revision. Although belief revision is the main method to solve the problem of the inconsistent belief currently, there are still some problems, such as the lost of expected information, the generation of unwanted conclusions, and the difficulties of choosing conclusions, etc.The method of belief non-revision is proposed to handle the problems of belief revision, which proceeds with inferences with the existence of inconsistent information. It will not revise the possible inconsistent belief, and add the new knowledge to the knowledge base directly, which avoids the lost of useful information in the process of belief revision. And then, it promises the proceeding of inference without the maintenance of the knowledge base in the process of knowledge increasing. Furthermore, it makes the conclusion without reverse information, which could be inferred with the original belief, as final conclusion in the system. This makes the system has the unique conclusion set, and solves the difficulty of choosing the conclusion when there are inconsistent beliefs in the system.In the existing belief non-revision methods, general resolution principle is applied. In the process of the application of general resolution principle, many useless redundancies will emerge. To avoid this problem, in this paper, the author puts forward methods to optimize the static non-revision methods of belief set in the propositional logic, and utilizes some methods to limit the resolution process, such as deletion strategy and lock resolution which could improve the efficiency of resolution. This paper proves that, the hypothesis expansions got after improvement do not change essentially, and the same good properties still exist.Besides, this paper makes the algorithm of the static non-revision methods basing on the deletion strategy and lock resolution, and confirms the feasibility of optimization.