一种新的用户事务算法  

New user transaction algorithm

作　　者：高集荣[1] 田艳[2] 邵海英[1] 

机构地区：[1]中山大学信息科学与技术学院,广州510275 [2]西安财经学院信息学院,西安710061

出　　处：《计算机应用》2009年第4期1099-1101,1105,共4页journal of Computer Applications

摘　　要：提出了双阈值用户事务算法。根据用户所访问的页面数来判断该用户是否为偶然用户,利用网络的拓扑结构和网页最低兴趣度来衡量一个网页是否为用户感兴趣的页面。改进了数据预处理过程,删除了偶然用户引起的访问记录,以及链接页面和用户不感兴趣的页面,生成一种有效的访问页面序列,即双阈值用户事务。通过事例对算法的有效性进行了论证。This study proposed a user business algorithm with double thresholds. This algorithm first acted according to the page number which the user visited to judge whether this user was the accidental user, and then the network topology and homepage lowest interest degree to judge whether the homepage appealed to the users. This method improved the data pretreatment process, and deleted the visit record which the accidental user caused, as well as the link pages and the pages that users were not interested in, produced one kind of effective visiting page sequence, namely double thresholds user business. This paper has proved the validity of the algorithm through an instance.

关 键 词：数据挖掘算法 数据预处理 用户事务 双阈值 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]