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

This thesis analyzes and compares three systemic approaches—Checkland's Soft Systems Methodology, system dynamics, and optimization—in addressing the efficiency issues of ETL processes in a data warehouse. The aim is to evaluate their applicability in terms of time requirements, labor costs, result measurability, stakeholder satisfaction, and solution functionality. Checkland's Soft Systems Methodology identifies key problems and relationships within the system, contributing to a better understanding of processes and stakeholder involvement, though its results remain primarily qualitative. System dynamics provides a deeper insight into the dynamic relationships between factors such as automation and manual interventions, identifying key feedback loops that influence system dynamics. Finally, the selected optimization method, linear programming, offers concrete and measurable results focused on minimizing delays and costs, albeit with less emphasis on broader systemic relationships. The comparison highlights that each method has its specific strengths and limitations, and their combination provides the most comprehensive approach to optimizing ETL processes. The thesis concludes by summarizing the findings and offering recommendations for applying these methods in other areas of data processing.