Transformation of SQL Databases models to NoSQL

Abstract

The rapid growth of data-driven applications and the increasing need for flexibility and scalability in database systems have led to a surge in interest in NoSQL databases. As organizations transition from traditional relational databases to NoSQL databases, schema conversion becomes a critical challenge. This research presents a novel graph-based algorithm for converting relational database schemas to NoSQL formats, specifically MongoDB, while preserving data integrity and relationships. The proposed algorithm utilizes graphs to represent and analyze relationships between tables in a relational database schema. By calculating foreign key sequences, the algorithm guides the schema conversion process, ensuring that the relationships between tables are maintained during the transformation. The conversion process involves referencing and embedding techniques to construct a hierarchical data representation in the target NoSQL format. This research also explores multiple approaches to schema conversion and identifies the most suitable methods based on various factors. Although performance optimization was not the primary goal of this research, the proposed algorithm demonstrates promising results in terms of efficiency. The research highlights the algorithm's versatility and usefulness in various scenarios, aiding organizations in migrating from relational databases to modern NoSQL databases, such as MongoDB. Future work includes refining the algorithm to handle edge cases, supporting different NoSQL database paradigms, optimizing performance, and testing scalability for large-scale databases. By addressing these areas, the algorithm can be further developed and tailored to cater to a broader range of database structures and systems, paving the way for new advancements in database schema conversion.