基于粗糙近似的Web事务聚类改进算法

基于粗糙近似的Web事务聚类改进算法 Clustering Web Transactions Based on Rough Approximation 李石君 LI Shijun1 1 2 3 4 first-author 张瑞 ZHANG Rui2 1 2 3 4 卢哲延 RO Cholyon1 1 2 3 4 金索哲 KIM Sokchol3 1 2 3 4 武汉大学计算机学院 武汉大学计算机学院 School of Computer, Wuhan University School of Computer, Wuhan University 新汶矿业集团职工大学信息系 新汶矿业集团职工大学信息系 Department of Information, Xinwen Mining Industry Group Employees’ University Department of Information, Xinwen Mining Industry Group Employees’ University 金哲柱师范大学数学与计算机科学系,湖北,武汉,430072 金哲柱师范大学数学与计算机科学系,湖北,武汉,430072 Department of Mathematics and Computer Science, Kim Chol Ju Normal University, Pyongyang, DPR of Korea Department of Mathematics and Computer Science, Kim Chol Ju Normal University, Pyongyang, DPR of Korea ,平壤,朝鲜 ,平壤,朝鲜 针对现有Web使用记录挖掘方法发现的知识和规则存在不精确或不完全的问题,提出了一种改进的基于粗糙近似的Web事务聚类算法.该算法首先对原有算法得到的每个类,用关系矩阵找出在这个类中不满足传递性的对象,然后在该对象的集合（类）上用关系矩阵找出在这个类中不满足传递性的对象,再对每个类进行分割,使其满足关系R的传递性,克服了现有算法不满足关系的传递性和在某些情况下得到的结果不准确的缺点. Web usage mining can mine useful information from Web access log, which can help designers design the sites and service. In general, the discovered knowledge or unexpected rules are likely to be imprecise or incomplete. Rough set theory is the mathematic tool dealing with imprecise and incomplete problems. In this paper, we present a novel extend algorithm for rough approximation-based clustering to cluster web transactions from web access logs. Web使用记录挖掘 粗糙集 聚类 Web usage mining rough sets clustering TP311.13;TP393.09 国家自然科学基金（600773007）;湖北省自然科学基金（2005ABA238）资助项目 2008-01-01 2021-04-01 1