Operator fuzzy logic based on argumentation

The knowledge bases in use, especially those of large scale, distributed expert systems, are often inconsistent. The inconsistency may be caused by different sources of knowledge or the relative correctness of the inference rules of the expert systems. As a result, it is difficult to maintain the consistency of a large-scale knowledge base. Thus, a robust inference engine must be able to cope with such inconsistency and should produce rational reasoning in the presence of inconsistent knowledge. In this paper, in order to handle inconsistency, we establish a conflict resolution model based on operator fuzzy logic. Firstly, we propose a kind of lattice-valued operator fuzzy logic, called AOFL, based on argumentation considerations. Secondly, we present the supported model semantics of generalized fuzzy Horn clause sets in AOFL, and develop a mechanical algorithm, SM, to calculate the supported models of a cyclic, non-free generalized fuzzy Horn clause sets in AOFL. Finally, we discuss the application of AOFL in conflict resolution and nonmonotonic reasoning.