Uncertainty reasoning based on filter of lattice implication algebra

Uncertainty reasoning is one of the important issues of many research fields such as artificial intelligence, knowledge-based system, data mining, etc. In the framework of lattice-valued propositional logic LP(X), the present work focus on how to deal with uncertainty reasoning with incomparable information by taking filter of lattice implication algebra as the measure of reliability of antecedents and conclusions. Firstly, we investigated the unsuitability of inference rules of classical logics in uncertainty reasoning and discussed the relation between inference rule and truth-values. Secondly, by taking filter as a measure of reliability, we studied the dynamic and hierarchical characteristics of uncertainty reasoning governed by the underlying lattice-valued propositional logic LP(X). Finally, we illustrate our uncertainty reasoning method through a simple example.