计算机工程与应用 ›› 2007, Vol. 43 ›› Issue (20): 169-172.

• 数据库与信息处理 • 上一篇    下一篇

基于深度优先序列模式挖掘的预取模型

卫 琳1,石 磊2   

  1. 1.郑州大学 升达经贸管理学院资讯系,郑州 451191
    2.郑州大学 信息工程学院计算机系,郑州 450052
  • 收稿日期:1900-01-01 修回日期:1900-01-01 出版日期:2007-07-11 发布日期:2007-07-11
  • 通讯作者: 卫 琳

Prefetching model based on depth-first sequential pattern mining

WEI Lin1,SHI Lei2   

  1. 1.Shengda Economics Trade and Management College,Zhengzhou University,Zhengzhou 451191,China
    2.School of Information Engineering,Zhengzhou University,Zhengzhou 450052,China
  • Received:1900-01-01 Revised:1900-01-01 Online:2007-07-11 Published:2007-07-11
  • Contact: WEI Lin

摘要: 序列模式挖掘能够发现隐含在Web日志中的用户的访问规律,可以被用来在Web预取模型中预测即将访问的Web对象。目前大多数序列模式挖掘是基于Apriori的宽度优先算法。提出了基于位图深度优先挖掘算法,采用基于字典树数据结构的深度优先策略,同时采用位图保存和计算各序列的支持度,能够较迅速地挖掘出频繁序列。将该序列模式挖掘算法应用于Web预取模型中,在预取缓存一体化的条件下实验表明具有较好的性能。

关键词: 序列模式, 深度优先, Web缓存, Web预取, Web挖掘

Abstract: Sequential pattern mining can find the user access patterns hidden in the Web log,which can be used to predict the upcoming Web access objects in the Web prefetching model.Most sequential pattern mining algorithms are based on Apriori breadth-first search strategy.A new sequential pattern algorithm based on depth-first search strategy is introduced and the mining mechanism based on lexicographic tree is presented in this paper.The data structure of bitmap is used in order to save and calculate the support of sequences fast.By the use of which,a prefetching model is proposed in integrated Web prefetching and caching environment.The experimental results show that the prefetching model based on depth-first sequential pattern mining can have a good performance.

Key words: sequential pattern, depth first, Web cache, Web prefetching, Web mining