Neo4j database based on graph model is different from other database on implementation.Firstly, this paper explores the architecture of the system and its internal mechanism.Secondly, Neo4j offers three ways to query, and evaluate its query performance from several dimensions: data size, query complexity, query number, etc. The results show that there are obvious difference in performance under different scenarios. This paper analyzes the experimental results and selection suggestions of query ways are recommended later. It provides a reference of query performance optimization for specific business applications.

No abstract