Abstract: The classical algorithm for mining association rules has low efficiency and high redundancy in the process of generating rules. To solve these problems, a new algorithm of finding non-redundant association rules based on frequent concept lattice is proposed. This concept lattice is generated on the basis of the FP-tree. During building lattice, all of nodes are formated according to the index of items which are found in the fequent-item head table. At the same time these nodes are selected by comparing supports. The Hasse graph is composed. In this graph, the intention of node is frequent itemset and the extension of node is support count of this itemset. And the non-redundant association rules are gained by scanning the leaf nodes of lattice. The simulation shows the feasibility of the algorithm proposed.

Key words: data mining, association rules, frequent itemsets, concept lattice

摘要： 为解决经典关联规则生成算法挖掘效率低及形成规则冗余性大的问题，提出在FP-tree基础上直接生成频繁概念格并提取无冗余关联规则的算法。其建格过程根据FP-tree频繁项目头表中各项的索引可分别独立进行，由支持度计数约束进行结点的筛选，形成频繁概念格的Hasse图，图中结点包含频繁项集及其支持度计数信息，通过对全部叶子结点的扫描可生成无冗余关联规则。通过实例验证该算法行之有效。

关键词: 数据挖掘, 关联规则, 频繁项集, 概念格