改进的最大频繁子图挖掘算法  

Improved Algorithm for Mining Maximal Frequent Subgraph

作　　者：柴然[1] 郭彦颖[1] Chai Ran;Guo Yanying(Computer Department,Cangzhou Medical College,Cangzhou Hebei 061001,China)

机构地区：[1]沧州医学高等专科学校计算机教研室,河北沧州061001

出　　处：《信息与电脑》2017年第18期42-43,47,共3页Information & Computer

摘　　要：在保证挖掘信息完整的前提下,为了避免挖掘所有频繁子图、降低挖掘难度,笔者进行自顶向下的最大频繁子图挖掘。挖掘过程中,应用迭代分区得到规范邻接矩阵,利用决策树计算支持度。最后,通过对比实验结果来论证算法的高效性以及可行性。Under the premise of ensuring the integrity of mining information and in order to avoid mining all frequent subgraph and reduce the difficulty of mining,the author carries on top-down algorithm for mining maximal frequent subgraph.In the process of mining,the canonical adjacency matrix is obtained by iterative partition,canonical adjacency matrix is used to calculation support.Finally,the efficiency and feasibility of the algorithm are proved by comparing the experimental results.

关 键 词：最大频繁子图 自顶向下 规范编码 决策树 

分 类 号：TP311[自动化与计算机技术—计算机软件与理论]