集群系统中基于MPI的关联规则快速挖掘算法  

Reseach on Fast Algorithm Based on MPI for Mining Association Rules in Cluster

作　　者：安立奎[1] 钱伟懿[1] 韩丽艳[2] 

机构地区：[1]渤海大学数学系,辽宁锦州121003 [2]渤海大学公共计算机教研部,辽宁锦州121003

出　　处：《三峡大学学报（自然科学版）》2010年第1期95-97,共3页Journal of China Three Gorges University：Natural Sciences

基　　金：国家自然科学基金资助项目(10871033);辽宁省高等学校科研项目资助计划(2008004);辽宁省教育厅A类项目(2009A034)

摘　　要：集群系统是一种分布存储系统,它主要利用消息传递方式来实现各结点之间的通信.在集群系统中通过利用MPI中的消息传递调用接口,提出了一种基于二进制形式的数据存储和计算的高效并行挖掘算法,通过对数据对象进行一些"或"、"与"、"异或"等逻辑运算操作,降低了算法的实现难度,减小了网络通信量,提高了挖掘效率,具有较好的可行性和可扩展性.Cluster is a kind of distributed storage system, which mostly adopts the method of message passing to realize the communication between every node. Through the interface for message passing of MPI, puts forword a fast parallel algorithm for mining association rules based on binary data storage and computating, which needs only some logic operations such as ‘ and＇, ‘or’ and ‘xor＇, decreases the realizing difficulty, reduces the network traffic and increases the mining efficiency. It is proved theoretically that the algorithm has the advantage of improving the effectiveness and scalability of mining association rules.

关 键 词：集群系统 并行计算 MPI 关联规则 

分 类 号：TP393[自动化与计算机技术—计算机应用技术]