A comprehensive BPMN approach for data-driven optimization of Italian civil proceedings

The integration of Business Process Model and Notation (BPMN) and process mining techniques offers a promising path for analysing and improving judicial proceedings. This study investigates into the utilization of BPMN for modelling judicial processes and employs process mining to analyse process logs obtained from the Employment section of the Turin Court. Data extraction and identification of the attributes were followed by a Statistical analysis, to reveal general patterns of the logs. Preprocessing included removal of outliers, cleaning of inconsistencies and redundancies, data aggregation. The result allowed our interdisciplinary team to focus only on critical elements. Thanks to a process mining tool (Apromore) we could visualize and analyse the process map, allowing comparison with the BPMN model from previous work. The “Hearing waiting time and hearing” activity significantly impacts the duration of the proceeding. This activity may occur repeatedly, and its duration is especially noticeable for the initial hearing. The analysis of process variations concentrated on differences in the closing phase of the proceeding and its influence on the overall duration. Variations ending with premature and anomalous termination show the smallest average time, followed by proceedings involving settlements or Fornero procedures. Understanding the factors influencing these outcomes contributes to a deeper understanding of process dynamics. Overall, this research underscores the value of process-oriented analysis in judicial settings, offering useful vision to improve efficiency and performance. By focusing on specific activities and phases, our study provides a foundation for further investigations into court proceedings, with broader implications for judicial practice and policy.