数据库中关联规则的并行/分布式采掘技术  

Techniques for Parallel and Distributed Mining of Association Rules in Database

作　　者：张惠民[1] 王晓卫[2] 肖庆[1] 

机构地区：[1]装甲兵工程学院信息工程系,北京,100072 [2]陆军航空兵工程学院基础部,北京,101114

出　　处：《装甲兵工程学院学报》2003年第2期38-41,共4页Journal of Academy of Armored Force Engineering

摘　　要：高效的并行/分布式采掘方法是数据库中关联规则发现技术的重要的研究方向,首先给出了关联规则的并行/分布式采掘的相关描述,针对关联规则并行/分布式采掘的基本思想,分析算法中可以改进效能的环节,提出生成更少候选集、局部剪枝、减少通讯量和优化扫描等提高性能和降低复杂度的方法.Techniques for parallel and distributed mining of association rules in database are an important research trend of data mining. We propose some additional descriptions on relevant problems, and describe the concise algorithm of parallel and distributed mining of association rules. Some approaches to improving the mining performances and reducing the complexity are summarized and provided via detailed analysis, such as less candidate set generation, branch pruning, communication reducing, and scan optimization.

关 键 词：关联规则 知识发现 并行采掘 分布式采掘 支持度 大项集 

分 类 号：TP311.13[自动化与计算机技术—计算机软件与理论]