process quality

In today’s data-centric business landscape, characterized by the omnipresence of advanced Business Intelligence and Data Science technologies, the practice of Process Mining takes center stage in Business Process Management. This study addresses the critical challenge of ensuring the quality of event logs, which serve as the foundational data source for Process Mining. Event logs, derived from interactions among process participants and information systems, offer profound insights into the authentic behavior of business processes, reflecting the organizational rules, procedures, norms, and culture. However, the quality of these event logs is often compromised by interactions among various actors and systems. In response, our research introduces a systematic approach that leverages Python and the pm4py library for data analysis. We employ trace filtering techniques and utilize Petri nets for process model representation. This paper proposes a methodology demonstrating a significant improvement in the quality metrics of extracted subprocesses through trace filtering. Comparative analyses between the original logs and filtered logs show enhancements in fitness, precision, generalization, and simplicity, highlighting the practical importance of trace filtering in refining complex process models. These findings offer practical insights for practitioners and researchers involved in process mining and modeling, highlighting the significance of data quality in obtaining precise and dependable business process insights.