基于模糊容差关系粗糙集的分解及其在信息检索中的应用

Decompositions of Rough Sets Based on Fuzzy Tolerance Relations

基于容差关系粗糙集的提出扩大了粗集的应用范围。本文在容差关系扩展为模糊容差关系的基础上引入支配关系对模糊容差关系的容差类进行进一步分解,分为模糊纯容差类、模糊正容差类、模糊负容差类,并分别定义上下近似集,进而提出了模糊容差关系粗糙集的分解。该方法能有效地处理连续属性,并用信息检索的例子证明了其应用价值。

The presentation of rough sets based on tolerance relations expands the application scope of rough set theory. By extending tolerance relations to fuzzy tolerance relations and introducing dominance relations, the present paper further decomposes the tolerance class of a fuzzy tolerance relation into pure fuzzy tolerance class positive fuzzy tolerance class and negative fuzzy tolerance class . Then the paper defines lower approximation and upper approximation under fuzzy tolerance relation respectively. Furthermore it suggests to divide rough sets of fuzzy tolerance relations. This method can deal with continuous attributes effectively. One example of information retrieval is used to demonstrate its application values.