基于约简概念格的关联规则提取改进算法  

Mining associations based on simplified concept lattice by improved algorithm

作　　者：陈湘[1] 吴跃[1] 

机构地区：[1]电子科技大学计算机科学与工程学院,成都610054

出　　处：《计算机应用研究》2011年第4期1293-1295,共3页Application Research of Computers

基　　金：基于神经网络的大规模数值模拟数据分析技术与研究(10476006)

摘　　要：为了提高在概念格上生成频繁项集的效率,提出了一个基于约简概念格的生成频繁项集的新算法。该算法通过利用节点之间的父子关系能够直接生成全部频繁项集,省略了对节点进行排序的时间开销,大大减少了节点比较次数,从而提高了频繁项集的生成效率。实验结果证明了其可靠性和高效性。This paper proposed a new algorithm based on simplified concept lattice to improve the efficiency of generating large set form the concept lattice.The algorithm could generate all the large itemsets by using the parents-children correlation of the nodes.It improved the efficiency of generating large sets by cutting the time cost of sorting nodes and reducing the comparing times between the nodes.Through the experiments on performance study,it can be seen that the proposed algorithm has much superior performance in efficiency to the original algorithm.

关 键 词：知识发现 数据挖掘 关联规则 频繁项集 约简概念格 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]