Grc-Based Research On Information System Knowledge Discovery

Information system is an important form of mass data,and the process of searching information from information system through algorithm is the main content of knowledge discovery.Truth table is a special form of information system that occupies an important position in the application of digital logic.Granular computing(GrC)is a method of research information system developing in recent years.It is a new way to solve complex problems and deal with intelligent information.Rough set theory(RST)is an important theoretical tool in granular computing.Through analysis and reasoning the data,RST discover the implied knowledge and reveal the potential laws.Rule extraction is one of the important research contents of RST knowledge discovery.It is a theoretical method to obtain the implicit knowledge of information system.Based on the study of GrC and RST,this paper studies the knowledge discovery of information system,mainly discusses the existing rule extraction algorithm and the existing defects,and proposes a new rule extraction algorithm for information system based on GrC.And a new parallel reduction algorithm is proposed for the truth table.Specific work is as follows:First of all,aiming at the main form of the information system--the decision table,use the thought of granulation in GrC,we define distinguish vector from the multi-granularity point of view.By analyzing the decision table in the coarse to fine granularity space,we can extract the rules in the information system based on the element values of the distinguish vectors.Aiming at the inconsistent decision table,before rule extraction we need convert the inconsistent decision table into a consistent decision table.In this paper,the validity of the new algorithm is illustrated by the theorem and the case analysis.By comparing with the existing rule extraction algorithms in the UCI dataset,the test results show the effectiveness and fastness of the new algorithm.Then aiming at the special form of information systems--truth table,firstly,the shortcomings of the traditional reduction algorithm are analyzed,and we define distinguish matrix based on GrC knowledge.In the multi-granularity space,according to the result element values of the distinguish matrix we can extract the simplest rules of each output for achieving the reduction of the truth table and through the parallel computing to speed up the efficiency of the algorithm.In this paper,we let the truth table of light-emitting diodes as an example,through the specific process showed the new algorithm calculation.By comparing the formula method,Karnaugh map method,Q-M algorithm and other traditional truth table reduction algorithms,tests results of dataset show that the new algorithm has accuracy and fastness.Finally,a simple information system knowledge discovery system is designed on the basis of this paper.The system integrates some existing decision table rule extraction algorithms,and we also design a subsystem to reduce the truth table for users easy to operate.The three kinds of information system discovery algorithms proposed in this paper overcome some drawbacks of the existing algorithms.The decision rules obtained by the algorithm are improved in terms of accuracy and simplicity,and the fast rule extraction process of data is realized.