An Optimization of Closed Frequent Subgraph Mining Algorithm

Graph mining isamajor area of interest within the field of data mining in recent years. Akey aspect of graph mining is frequent subgraph mining. Central to the entire discipline of frequent subgraph mining is the concept of subgraph isomorphism. One major issue in early subgraph isomorphism research concerns computational complexity. Normally, the subgraph isomorphism problem is NP-complete. Previous studies of frequent subgraph mining have not solved NP-complete problem in the subgraph isomorphism. In this paper, we proposeanew algorithm which can deal with this problem. The proposed algorithm can solve the subgraph isomorphism in polynomial time in some settings. Moreover, the new algorithm is proved theoretically more effective than previous studies in closed frequent subgraph mining.