基于变精度粗糙集的Web用户聚类方法

Web User Clustering Method Based on Variable Precision Rough Set

针对Web使用挖掘中的用户聚类问题,提出一种基于变精度粗糙集理论的粗糙聚类方法,该方法放宽经典粗糙集中不可区分关系的传递性将其扩展为相容关系,使用变精度粗糙集的相对错误分类率β来形成新的相似β上近似,从而将一个用户划分到多个聚类,该方法不需要区分用户会话,降低了数据预处理的难度,通过理论推导和实例证明了其有效性。

Focus on solving the user clestering issues, a rough clustering method based on the variable precision rough set theory is proposed. The indiscernibility relation in classical rough set is extended to a tolerance relation with the transitivity property being relaxed. The new proposed similarity β upper approximations are formed using the relative degree of misclassification β, so a user can be assigned to more than one cluster, and this approach does not need to identify the users＇ sessions, therefore, the complexity of data preprocessing decreases. Experimental example shows the effectiveness of the proposed algorithm.