Extending ETL for NoSQL: A validated Framework for Document-Oriented Warehousing

Decision-making relies on the creation of a data warehouse, which in turn depends on the ETL process. The traditional ETL process is foundational for building Data Warehouses (DWs), particularly in relational environments where data is structured and adheres to strict schemas. With the rise of Big Data and NoSQL databases, new challenges have emerged due to schema flexibility and data heterogeneity. Given the critical need for timely and accurate decision-making in healthcare, it is essential to ensure efficient and scalable analytical processing of medical data. To address these challenges, this paper proposes an extended ETL approach—T-ETL-NoSQL—which introduces pre-aggregation during transformation phase and structure creation during loading phase to optimize the construction of NoSQL data warehouses. The focus of the work is on validating the applicability of this adapted ETL process for document-oriented NoSQL systems, specifically in the context of COVID-19 data.