摘要： 粒度计算理论为知识发现的研究提供了新的理论工具，通过引入全粒度空间的概念，对信息系统的论域空间进行等价划分，给出了论域子空间可精确粒度表示的充分条件，建立了目标概念的外延和内涵的粒度计算关系，揭示了目标概念学习中的粒度计算过程，提出了一种新的数据分类模型。

关键词: 粒度计算, 全粒度空间, 数据分类, 目标概念

Abstract: The problem of dada classification is a fundamental issue in KDD. There are still many open theoretical problems such as modeling for data classification. Granular computation theory offers a proper new theoretical basis for KDD. In fact, the process of constructing the connotation of a concept from its extension can be considered as a granularity computation process. This paper proposes a novel granularity representation method to study the problem of data classification in KDD and a novel data classification model.

Key words: granularity computation, complete granularity space, data classification, object concept

中图分类号: 

• TP 311