Research Of Logical Foundation And Reasoning Algorithm In Fuzzy Information Processing

In the field of information science, knowledge division, knowledge representation and knowledge processing compose the base of information theory. While the cognition and processing of negative information also plays a great important role. In chapter 1, the paper mainly introduced the development of study about negative information.In chapter 2, it proposed that negation in fuzzy knowledge oughts to distinguish contradictory negation from opposite negation. A common character between opposite concepts was discovered. It considered that negation in fuzzy knowledge includes three different negations: contradictory negation, opposite negation and medium negation. Found on the cognition, a new Fuzzy Sets for differentiating Contradictory negation, Opposite negation and Medium negation (FScom) was defined, properties of which were also discussed.In chapter 3, based Medium Logic theory, it applied fuzzy negation and opposite negation to knowledge representation, made further differentiation in fuzzy negation as needed, gave new reasonable definitions to extend the traditional knowledge representation. Then it took a financial decision-making program for example. According to thoughts of infinite-valuedλ-interpretation of Medium Predicate Logic and [λ,μ]-cut segment, it discussed about measure of truth value and presented a kind of method to ascertain value ofλinλ-interpretation including its meaning. Finally fuzzy reasoning computing was realized. In chapter 4, proceeding from a principle that all things are related to all other things, it defined the mode of high-order fuzzy reasoning with many dimensions and discussed about the algorithm of the second order one in different circumstances based frame of CRI. It took the selection of"three good student"in a school for instance to study the application of the new model and testified that the multi-level fuzzy synthetic evaluation model can be seen as a special instance of high-order fuzzy reasoning with many dimensions.In chapter 5, the paper represented fuzzy knowledge based Medium Predicate Logic. It modified and-or graph with reason, regarded every predicate expression as state node to convert logic formulary into state space. The heuristic function of traditional searching method in and-or graph was also modified in order to treat fuzzy reasoning problem as problem of searching for the most reliable goal node. Besides, a new kind of method to process negative information was given, which is convenient for computer to realize.