An Approach to Converting Relational Database to Graph Database: from MySQL to Neo4j

Abstract

At present, there are few research methods that can convert any relational database into a graph database, and most of them are based on a specific field data set to build a relational database, and then perform simple conversion through the characteristics of the data set.Aiming at this problem, a universal conversion method is proposed. Firstly, converted the most basic component tables name, records, and fields in the relational database into labels, nodes, and corresponding attributes of the nodes under the graph database; secondly, used the intermediate connection table method to convert the foreign keys in the relational database into the relationship of a graph database between the nodes; then some constraint issues in relational databases, such as multiple primary key issues, indexes, and no default values, were optimized to form a final graph database model that met expectations; finally, Realized the effective migration of large quantities of data in the relational database to the constructed graph database model. In the experiment, the above method was used to successfully convert a relational database to a graph database, and the database construction, data import, SQL query and Cypher language query were performed for the database before and after the conversion, and through the analysis and comparison of data integrity, time cost, result validity,which shows that the integrity and operability of the database before and after conversion are consistent, and the data processing efficiency of the database is much higher than that of the relational database, which verifies that the method in this paper is feasible.