基于量化概念格的关联规则挖掘

在概念格的内涵中引入等价关系并将其外延量化 ,得到量化概念格。利用量化概念格可以清晰地表示知识 ,从而便于挖掘包括关联规则在内的多种规则 ,与经典的 A priori算法相比较 ,规则表示更简捷、直观 ,尤其重要的是用户可根据自己的兴趣交互地挖掘关联规则 ,不需要计算频繁项目集 ,因而提高了挖掘规则的效率 ,适用于大型数据库中关联规则的挖掘

Algorithm of mining association rules based on Quantitative Concept Lattice

The Quantitative Concept Lattice (QCL) evolves from concept lattice by introducing equivalent relation to their intensions and substituting the number of instance in extensions for their extensions. Knowledge on the QCL can be clearly discovered and some kinds of rules such as association rules easily mined. In comparison with Apriori algorithm, rules are presented succinctly and visually. Above all, users can mine association rules according to their subjective interests,and by using the algorithm of mining association rules on the QCL,frequent item sets need not be calculated, then the efficiency of mining association rule is improved. Therefore,the presented algorithm is suitable to the mining of association rules in large databases.