序信息系统中知识粒库的研究及应用

The research of Knowledge-Granular in Ordered Information System and Its Application

借鉴已有的粒计算理论成果,从信息粒化的观点出发,给出了序信息系统中知识粒库的定义,分析了知识粒库的相关性质,构造了知识粒库贴近度,进一步指出利用知识粒库贴近度可以把知识之间的相似性度量转化为序信息系统中粒之间的运算,并基于知识粒库设计了对序信息系统进行无核属性约简的算法,说明了该方法是行之有效的.

Based on the theory of granular computing and from the viewpoint of information granularity , Knowledge-granular base in ordered information systems is defined in the paper .After studying its several proper-ties, this paper proposes the close-degree of knowledge-granular base, which is used to transform the problem of how to measure the similarity of knowledge into the problem of operations between granules in ordered information systems.Finally, an attribute reduction algorithm without core based on knowledge-granular base is presented , and this method is proved to be effective .