关键词: 图论 频繁子图 数据挖掘

Abstract:

Nowadays association rule mining techniques are applied to many non-traditional domains, existing approaches for frequent itemsets discovery are inapp licable. An alternate way to solve these problems is to represent the transactions of those domains by graph,and find the frequent subgraphs by using gr aph-based data mining techniques. We propose a new algorithm named SLAGM, which is based on Apriori＇s idea. It can mine frequent subgraphs from simple  graphs efficiently. After being evaluated by experiments with synthetic datasets, this algorithm show better performance than another subgraph mining a lgorithm AGM.

Key words: graph theory, frequent subgraph, data mining