中文摘要：

传统的类关联规则挖掘方法在挖掘完整的规则数据集时往往需要消耗很长的时间。为了解决这个问题,提出一种高效的基于等价类规则树的类关联规则挖掘算法。首先,通过分析等价类规则树挖掘类关联规则算法存在的耗时问题,设计一个树结构存储数据集的频繁项集;接着,基于这棵树推导出一些修正树上节点和减少节点信息计算量的定理;最后,利用这些定理得到一个有效的适用于挖掘类关联规则的算法。实验结果表明,与其他较为先进的基于等价类规则树的关联规则挖掘算法相比,所提算法更加高效。

英文摘要：

Traditional class-association rules （CAR） mining methods usually need long time to mine a complete rule dataset. To address this issue, we propose an efficient CAR mining algorithm which is based on equivalence class-rules tree. First, by analysing the time consuming problem of equivalence class-rules tree in mining CAR algorithm, we design a frequent item sets for the storage datasets with tree structure. Then based on this tree we derive some theorems for pruning the nodes of the tree and decreasing node information computation load. At last, based on these theorems we obtain an effective algorithm suitable for mining the CAR. Experimental results indicate that the algorithm proposed is more efficient than other association rule mining algorithms based on equivalence class-rules tree.