Comparison of Selected System Approaches for Solving the Optimization Problem of ETL Processes in a Data Warehouse

This thesis analyzes and compares three systems approaches, Checkland's soft systems methodology, system dynamics and operational optimization, in solving problems with the efficiency of ETL processes in a data warehouse. The aim is not only to assess the applicability of these methods in terms of time requirements, human resource costs, measurability of results, stakeholder satisfaction and functionality of the solution, but also to compare the results of their practical application in solving a specific problem in the field of ETL processes. The motivation for this comparison is the frequent ambiguity in choosing the appropriate approach and the effort to identify in which situations individual methods can bring the greatest benefit. Checkland's soft systems methodology reveals key problems and relationships in the system, which contributes to a better understanding of the processes and stakeholder involvement, but its results remain only at a qualitative level. System dynamics provides deeper insight into the dynamic relationships between factors such as automation and manual interventions, while identifying key links affecting the dynamics of the system. Finally, the selected linear programming optimization method offers concrete and measurable results, focused on minimizing delays and costs, but less takes into account broader system relationships. The comparison showed that each method has its specific advantages and limitations, while their combined or integrated use can bring the most comprehensive approach to optimizing ETL processes. The conclusion of the thesis summarizes the results and offers recommendations for their application in other areas of data processing.