The study on data migration from relational database to graph database

Abstract

Under the background of big data, using relational databases to manage massive data may have some problems just like storage capacity and query efficiency. So, there is a new kind of databases called NoSQL to store data. However, the data models of NoSQL databases are different from relation databases. In order to finish migrating historical data from relational databases to NoSQL databases, in terms of graph database in the NoSQL databases, this paper takes ER diagram as the original model, graph model as target, and makes some transformational rules by using the relationships of entities. And this paper proposes an algorithm which can finish data migration by traversing ER diagram and using the transformational rules. This method can reduce the impact of model differences between relational databases and graph databases, ensure the integrity constraint of data, and automatically complete data migration. The experimental results show the validity and correctness of the data migration.