The Evolution of Big Data Architectures: From Traditional Data Management to Data Lakehouses

The rapid growth in data volume, variety, and generation speed has led to a significant transformation in data management architectures. This study examines the evolution of Big Data architectures from traditional file-based systems and relational databases to Data Warehouses, Data Lakes, and Data Lakehouses. The main characteristics, data processing approaches, advantages, and limitations of these architectures are discussed comparatively. The study also examines cloud-based Big Data platforms, distributed data processing technologies such as Apache Spark and Apache Kafka, and the integration of Big Data with Business Intelligence and Artificial Intelligence applications. In addition, current challenges such as data quality, metadata management, data governance, and the risk of Data Lakes becoming Data Swamps are discussed. Finally, the study highlights the management of data structures required by Machine Learning applications and high-dimensional data as important research topics for the future development of Data Lake and Lakehouse architectures.