Rough Data-deduction Based On The Upper Approximation And Its Applications

The study of information science involves various aspects of data processing.The related work facilitates the generation of research directions,and the achievement ad-vances the development of the subject.As research directions,data classification,data deduction,data warehouse,data reduction,data filtering,data mining,and so on not only demonstrate the wide and active research fields of data processing,but also reflects the idea of combining theory with practice.Although each of them has its research empha-sis,they often focus on some common aspects.In terms of data problems,the description method of the data connection that is unclear,nondeterministic,seemed or potential is re-lated to the directions,and also frequently occurs in practice.This gives rise to the notion of rough data-connection.The concentration on rough data-connection brings about the topic of rough data-deduction.Because this topic is rarely concerned,the research on it must be significant and advanced.Thus the topic serves as the focus in this paper.The completed work is as follows:A structure is constructed in this paper,called a rough deduction-space.It is an extension of an approximation space,and incorporates a deduction relation related to data connections.To describe rough data-connections,a notion of data deduction is introduced in the rough deduction-space,and is referred to as rough data-deduction.Rough data-deduction is based on the upper approximation of the information obtained by integrating an equivalence relation with the deduction relation.It accomplishes deductions among data,and leads to the main subject of this paper.The research conclusions show properties of rough data-deduction,including the characteristics,such as rough data-deduction keeps definite data-connections,rough data-deduction is closely linked with the approximate information in the upper approximation,rough data-deduction has the function of approximate description,rough data-deduction can be equivalently described by paths,and rough data-deduction corresponds to different equivalence relations,etc.As an example,a specific problem is modeled by a rough deduction-space which de-scribes different classifications of the enterprises in an automobile manufacturing industry chain,and records the supply channels between enterprises.In this rough deduction-space,rough data-deduction is used to characterize the potential supply channels from some enterprises to others.This may provide reference information for intelligent pro-cessing and automated management,and also shows the practical application of the the-oretical research.The discussion on rough data-deduction is developed in a tree deduction-space in which the deduction relation is taken as a tree.Based on the hierarchy information con-tained in the tree,a conclusion is proved and shows the fact that rough data-deduction relies on the hierarchies that the data are located in.Furthermore,by refining the tree deduction-space,the refined rough data-deduction demonstrates the characteristic of ap-proximating the precise information.At the same time,the refined rough data-deduction is used to analyze the product supply channels in the automobile manufacturing industry chain,which shows further applications of the theoretical method.In order to further study rough data-deduction,a notion of rough path is introduced in a rough deduction-space.It is proved that there exists the correspondence between a rough path and a deduction produced by rough data-deduction.Accordingly,the accurate degree of rough data-deduction can be described by use of a rough path.This makes it possible to distinguish between different meanings of the same form of rough data-deduction.A method is therefore set up and the accurate degree of rough data-deduction can be evaluated by it,which has a guiding role in practical application.By constructing a structure called a granulation tree,and by use of the hierarchy information of the tree,the notion of data association is introduced.This leads to rough data-deductions by way of association deduction.The association deduction takes an association data as a bridge to establish associations between the data in a data set and the data in another data set,which is combined with the data identity,closer identical.forms of data association,numerical representation of data association,maximal data association,and so on.At the same time,a special operation of upper approximation is used as a necessary and sufficient condition to determine whether data are associated.The association deduction embodies the characteristic related to the changes of hierarchy and granularity in granulation trees,and also shows the numerical representation of data association and accurate degree.In particular,the association deduction is bound up with practical problems as well.The data association method based on granular trees is used to describe the specific issues,thereby realizing the research expectation of applying theory to practice.The work of this paper takes rough data-deduction as the main subject,and takes the association deduction as an important part.The study progresses step by step,the discussion seeks clarity,the analysis advances gradually,and the research centers on the theme.The work shows the unique way of the research,and reflects the understanding and cognition on rough data-deduction and data association.The model description on practical problems and the characterization of actual data connections by use of rough data-deduction convey the research ideals of applying theory to practice and basing actual applications on theoretical research.