A Novel Approach for Migrating Relational Databases into Graph Databases

Relational database (RDB) systems are widely used to store traditional data for many organizations despite their challenges when dealing with non-traditional data required by contemporary applications. These systems should be able to handle transactions and domains that require storing large amounts of data. Recently, data has been growing faster than ever before, as extracting valuable information from big data is a complicated task using traditional database systems. Therefore, most organizations are trending to use NoSQL systems to store and manipulate their data. Converting existing SQL-based data into NoSQL structures is a very complex task due to the need to ensure data equivalence and accuracy with better performance. This paper aims to present a new approach to migrating data from an RDB to a graph database based on a conceptual data model. Algorithms are developed to generate the target database according to a set of migration rules to translate all RDB constructs into a graph schema, based on which the RDB data is converted to the new environment. The proposed solution has proved effective and accurate after a prototype system has been implemented and the results empirically tested and evaluated.