通过分析Eclat算法,对完全概念格按照支持度进行了裁减,得到了一个向下封闭的降序join半概念格,在构造半概念格的同时计算出每一个项集的支持度作为其权值,最后基于降序加权join半概念格对Eclat算法进行了改进,裁减了概念格中大量的冗余的连接,给出了一个快速的关联规则挖掘算法。经过分析,该算法与Eclat算法相比,效率更高。
Eclat algorithm is discussed and analyzed completely.Through pruning entireness lattice depend on support,a descend weighted join half concept lattice is obtained that is closed downward ,support for each node of concept lattice is calculated as constructing concept lattice,At last,a new fast mining association rules is presented,Its efficiency is more high than Eclat algorithm through analysis.