Data Integration Approaches for AI-Ready Data Architecture: A Comparative Review of Relational Databases, Data Warehouses, and Knowledge Graphs

Abstract

The rapid development of Artificial Intelligence (AI)-based systems has increased demand for integrated data architectures that support operational, analytical, and semantic data management. Organizations currently utilize relational databases, Data Warehouses, and Knowledge Graphs to support data integration in modern AI environments. However, previous studies generally discuss these approaches separately and rarely provide an integrated perspective regarding their roles within AI-ready architectures. This article conducts a comparative literature review of Relational Databases, Data Warehouses, and Knowledge Graphs as data integration approaches for AI-based systems. The study applies a conceptual literature review approach by analyzing scientific publications related to data integration, data warehouses, the semantic web, and knowledge graphs. The findings indicate that Relational Databases mainly support operational layers, Data Warehouses support analytical layers, while Knowledge Graphs provide semantic representation and contextual reasoning capabilities for modern AI systems. This article proposes a three-layer data integration perspective consisting of operational, analytical, and semantic layers as a conceptual framework for AI-ready data architectures.