Abstract: In order to mine association rules efficiently，and find the transaction items including Frequent Item Sets（FIS） when mining FIS，a new algorithm based on graph with character weights is given.Firstly，this paper proposes the graph with character weights，finds and approves some properties of such a kind of graph.Then，a novel method that mines FIS and the transaction items in which FIS appears is designed.This algorithm is validated by analyzing its time and space complexity.

摘要： 为了提高关联规则挖掘效率，在挖掘频繁项目集的同时，挖掘出包含频繁项目集的事务集，提出了基于字符权图的关联规则挖掘算法。首先，提出了字符权图的概念，发现和证明了它的一些性质。基于此，提出了挖掘频繁项目集及包含频繁项目集的事务集的算法。时间和空间复杂性的分析表明，该算法是合理和高效的。