基于信息粒度的知识网的模糊分类与检索方法

摘要

To solve over-subjective retrieval method, repetition retrieval and operation of knowledge meshes in the self-reconfiguration of knowledgeable manufacturing system, the method of fuzzy classification and searching based on information granularity is proposed. The complexity of knowledge mesh solves the problem about the diversity of knowledge mesh caused by self-reconfiguration computing. The similarity degree which considers the differences in three aspects of quality, quantity and complexity has the characteristics of reflecting operation laws. The fuzzy clustering method of knowledge mesh does not need the classes number, meanwhile the permutation about target knowledge mesh is obtained.The search space is determined by centering at the clustering, which converts the problem from fine-grained space to coarse-grained space.%针对知识化制造系统自重构中知识网检索方法过于主观以及重复检索和运算等问题,提出基于信息粒度的知识网的模糊分类和检索方法.知识网复杂度解决了自重构运算导致的知识网存在多样性的问题.相似度考虑知识网在"质"、"量"和复杂性等方面的差异,具有反映知识网运算规律的特征.知识网模糊聚类方法不需要确定分类数,并且能够同时获得关于目标知识网的排序.以各聚类中心为中心确定的检索空间实现了问题由细粒度空间转化为粗粒度空间.