一种基于FP-树的最大频繁模式增量更新挖掘算法  

AN FP-TREE BASED INCREMENTAL UPDATING ALGORITHM FOR MAXIMAL FREQUENT PATTERNS MINING

作　　者：李忠哗[1] 任春龙[2] 何丕廉[2] 

机构地区：[1]河北北方学院计算机系,河北张家口075000 [2]天津大学计算机学院,天津300072

出　　处：《计算机应用与软件》2007年第5期47-49,共3页Computer Applications and Software

基　　金：天津市科技发展计划资助项目(04310941R);天津市应用基础研究计划(05YFJMJC11700)

摘　　要：挖掘关联规则是数据挖掘领域的一个重要研究方向,人们已经提出了许多用于发现数据库中关联规则的算法,但对关联规则的增量维护问题的研究较少。深入分析了增量更新情况,使用了目前较高效的最大频繁模式挖掘算法FP-Max,并对其进行改进。基本思想:①基于FP-树;②考虑了数据集中,数据增加情况下FP-树的更新;③对FP-Max算法进行改进来更新、维护已经挖掘出来的最大频繁模式。Association rules is an important research topic in data mining area. Many approaches have been proposed to mine association rules. But little work was done on their maintenance. The case of adding new data into the database is analyzed in this paper. An algorithm which based FP-Max algorithm is presented. The main idea of this paper is that ,① It is a FP-tree based algorithm for mining maximal frequent patterns. ② Updating of FP-tree is considered. ③ FP-Max algorithm is improved and maximal frequent patterns which have been mined are updated.

关 键 词：数据挖掘 关联规则 频繁模式树 最大频繁模式 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]