A Data-Driven Decision Support Approach for Recalibration Service Satisfaction Using Decision Tree Analysis (Case Study: Directorate of Metrology Bandung)

The increasing digitalization of public services has created new opportunities for government institutions to leverage data-driven approaches in improving service quality and decision-making processes. In the context of public metrology services, customer satisfaction plays a critical role in ensuring service effectiveness, institutional trust, and continuous improvement. This study proposes an interpretable machine learning–based approach to support service evaluation and managerial decision-making by applying a decision tree (C4.5) classification model to recalibration service satisfaction data at the Directorate of Metrology Bandung, Indonesia. Using historical customer satisfaction data collected through digital questionnaires, the study develops a classification model to identify satisfaction levels and uncover service patterns that influence user perceptions. Rather than focusing solely on predictive performance, this research emphasizes the role of explainable models in transforming operational data into actionable insights for public service governance. The results demonstrate that decision tree–based analysis can effectively support the identification of critical service attributes and provide transparent decision rules that are understandable to non-technical stakeholders. The findings highlight the potential of data-driven decision support systems to enhance service management, promote accountability, and support smarter public service delivery within a digital governance framework.