概念格递增修正关联规则挖掘方法

研究了一种知识发现与数据挖掘中关联规则的发现方法 .关联规则是数据挖掘的重要方法之一 ,其核心是各大项目集的获取 .针对货篮关联规则挖掘方法 ,提出了一种改进的概念格递增修正方法 .该方法适应于数据库的动态数据递增或递减更新 ,通过记录项目集 (即概念格中的结点 )在数据库中出现的频率值 ,不需要构造完整的格即可求得项目集的支持度值和可信度值 ,以获取大项目集 ,进而求得关联规则 .同时 ,该方法运用 Hasse图解进行可视化操作 ,降低了算法的时间复杂度

Concept Lattice Based Approach for Incremental Association Rules Mining

A novel association rules mining algorithm in knowledge discovery and data mining was present- ed.Association rule is an important database discovery method,whose kernel is the acquisition of large itemsets.According to K.Hu's basketassociation rule mining algorithm,a modified concept lattice based approach for incrementally acquiring large itemsets was introduced.The approach is efficient when the database is dynamically updated( whether insertion or deletion) ,the frequency value of each itemset( each node in the lattice) is recorded;so the corresponding supportand confidence value can be obtained without constructing the complete lattice,which generates the association rules.Hasse diagram is used to visualize the process,and the algorithm's time complexity can be reduced.