摘要
现有的数据挖掘算法和模型主要是基于大型数据库或数据仓库的环境,大多采用集中式处理.而目前绝大部分的大型数据库都是以分布式的形式存在的,因此.提出新的分布式关联规则挖掘算法是非常必要的.针对FDM 算法中可能造成频繁项集丢失的缺点,提出了一种改进的分布式关联规则挖掘算法 DARM,该算法同时也减少了各分站点问的通讯量.从而提高了整个挖掘算法的效率.
At present,the data mining algorithms and models that are mainly based on large databases or data warehouse environment are centralized processing.However,as many databases are distributed,it is necessary to present a new algorithm of distributed association rules mining.In the paper,an improved algorithm DRAM is proposed to overcome the shortcoming of FDM algorithm that may lose frequent item- sets,reduce the communication consuming among the sites,and improve the efficiency of the entire mining algorithm.
出处
《湛江师范学院学报》
2007年第6期74-77,共4页
Journal of Zhanjiang Normal College
关键词
分布式数据挖掘
频繁项集
关联规则
distributed data mining
frequent itemset
association rules
apriori algorithm
