Multi-value Information System And Knowledge Reduction In Decision Table

Data mining aims to find out the way of abstracting automatically and intelligently valuable information or knowledge from a huge amount of data. At present this is an active research field. Rough set theory is a relatively new mathematical tool for use in computer application in circumstances which are characterized by vagueness and uncertainty, which has been widely used in many fields, such as machine studying, data mining, intelligent data analysis, algorithm of control. This article carries on in-depth research to extension of classical rough set theory and knowledge reduction in decision table, gains a good effect of theory and experiment. So works in this filed are theoretically important and practically valuable.This thesis mainly deals with theory and application of combination of rough set and relational database, besides knowledge reduction in decision table.With respect to extension of rough set theory under incomplete information systems, rough set modal based on class similarity relation is proposed in this thesis, this modal can reclassify to objects through equivalence partition of domain values.With regard to rough data query, the focus is firstly placed on analysis of advantages and disadvantages different methods bear in dealing with this problem, and a new method of rough query based on SQL is put forward in this chapter. This method can improve efficiency of query, decrease storage space and posses a good application values.Concerning knowledge reduction, a algorithm based on condition information entropy for knowledge reduction is bought forward in information system. This algorithm can measure the importance of attribute. By running practical system, we can find the minimal reduction for decision table.Multi-valued information system and knowledge reduction in decision table are two important problems in research of rough set theory. The work of this paper on them is valuable.Combination of rough set theory and relational database is in a booming stage and there are many problems to study on the relation of them. Our work is just a beginning, and related work needs to be further developed.