Research On Approximate Information Mining Of The Upper Approximation And Its Application

Granular computing is a research topic in information science. Its basic idea from a whole toparts and from the parts to the whole has been recognized by researchers who are focusing their attentionon this topic.Different researchers study on granular computing from different point of view.Some researchachievements focus on computer applications and other aspects, other results focused discussion on thetheoretical aspects of.The mathematical methods used in the study are not the same.Such as logic method,algebraic method and topological method or probability method, etc. There are three major theoreticalmodels when researching granular computing.They are fuzzy set, rough set and quotient space. Models aredifferent, but their original intentions to solve the issues and methods are same.They have their owncharacteristics. That is structured processing of the problem space.Small the complex issues and establishthe appropriate hierarchy.Solving the issue according to the hierarchical structure when the universal setwas seperated into different levels.This article analyses the development process of the granular computing theory and constructsthe space structure by the existing theories and the hierarchical structure. Integrating with the upperapproximation theory to investigate the approximate information contained in it. Obtain the followingresults:(1) Based on an information system, n-order-granule partitions are introduced in, which make theinformation system achieve different granule’s degrees, so that granular structures are obtained.(2) Furthermore, by arranging elements of the universe with a special way, and after integratingn-order-granule partitions with upper approximation, a data hiding and recognizing method is established,which plays an important role in this area. Moreover, by adjusting the order of the elements of the universe,the data hiding and recognizing are more flexible. Therefore, data hide and recognition can be realized invariable ways.(3) Define the T-approximate space by extending the approximate space in rough set theory. Andthen study the natures and conclusions to expand the research space of the approximation space.(4) Integrating upper approximation with T-approximate space, imprecise reasoning is introduced. This indicates that the imprecise information contained in upper approximation can be developed by areasoning approach. Furthermore, as a mathematical structure, a T-approximate space with imprecisereasoning can be used to model practical problems.