中间存储的联机分析关联规则挖掘方法  

Method of Association Rule of Intermediate Storage Based On-line Analytical Mining

作　　者：陈旭辉[1,2] 刘东坡[2] 武玉杰[2] 

机构地区：[1]厦门理工学院福建省高校物联网应用技术重点实验室,福建厦门361024 [2]兰州理工大学计算机与通信学院,甘肃兰州730050

出　　处：《厦门理工学院学报》2012年第2期57-61,共5页Journal of Xiamen University of Technology

基　　金：国家自然科学基金项目(61065007);厦门理工学院高层次人才科研项目(YKJ11013R)

摘　　要：针对"挖掘不同支持度下的关联规则需要反复扫描多维数据库所带来大量的系统I/O开销"这一缺限,提出一种基于中间存储的联机分析关联规则挖掘OLAM方法.在自行开发的联机分析处理OLAP的基础上,对待分析的多维数据集利用BCTree存储谓词集频度索引数,利用改进的Apriori关联规则挖掘算法分析用户感兴趣的维度,减少挖掘维度的个数,提高了反复挖掘的系统效率和系统针对性.生产数据证明了该方案的正确性和有效性.With the development of business intelligence technology, the research of effective techniques in multi-dimensional database for mining association rules is becoming more and more important. In order to solve the defects of multi-dimensional database system caused by I /O overhead repeatedly under different support of mining association. A new association rule mining measure based on the intermediate storage of OLAM（ On- Line Analytical Mining） is put forward. In the self-developed on the basis of OLAP （On-line Analytical Processing） to treat analysis, using BC-Tree storage cube to set the frequency of the index number of the predicate, then using the improved Apriori association rule mining algorithm to analysis the dimension of interest to users, improving the＇efficiency of repeated mining system. This can reduce the number of mining dimensions, increasing the system＇s target, and take the produce data for example to prove the correctness and effectiveness of the program.

关 键 词：多维数据库 关联规则 联机分析挖掘 联机分析处理 维度 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]