摘要： 归纳推理检索是基于事例推理(CBR)中常用的检索方法之一，是基于ID3算法的检索方法。文章在基于事例推理方法的背景下论证了事例的属性的重要性和冗余之间的关系，并在此基础上从属性相对于其属性的重要性角度来构造启发函数。和ID3算法相比较，该算法不但降低了计算复杂性，而且在一定程度上可以消除样本中的噪声，使归纳推理检索方法的检索效率有所提高。

关键词: 基于事例推理, 归纳推理检索, 决策树, 粗集, 重要度

Abstract: Inductive retrieval is one of the retrieval methods commonly used in case-based reasoning systems, which is based on ID3. In this article, the relationship between weightiness and redundancy of attributes is demonstrated. A decision tree algorithm is given based on rough set in the point of view of the instances attributes weightiness. It has advantages in computational complexity and in gaining more effective cases retrieval. Key words Case-based reasoning (CBR);Inductive retrieval;Decision tree;Rough set;Weightiness

中图分类号: 

• TP301