With the development of science and technology service industry, make more and more the number of service resources, service resources will exist between a relationship, you can make multiple services together to complete the same thing, but in the service users demand for combination, due to the needs of users the same or similar, makes combination will be the same or similar, for the same combination of similar cases, After comparing MySQL and Neo4j databases, a combinatorial multiplexing method combining graph database and improved VF2 algorithm was proposed. First, the service combination was stored in Neo4j database, and then the improved VF2 algorithm was used for subgraph matching. Experimental results show that compared with MySQL database, this method can reduce the number of service combination.The improved VF2 algorithm can reduce the time of matching combination compared with VF2 algorithm.