Utilizing Data Mining Techniques to Analyze Payment Order Rejections in the State Treasury Office

Abstract: One of the challenges in budget implementation was the rejection of Payment Orders (SPM), which led to delays in bill payments and a decline in the quality of budget execution. To analyze SPM rejections, data mining techniques utilizing a decision tree model were applied to 15,030 data records obtained from State Treasury Office (KPPN) Palu, covering the 2019 fiscal year. The analysis revealed that SPM-Contractual, SPM-Others, Non-Salary, and PPNPN were the most frequently rejected Payment Orders. Furthermore, Payment Orders associated with a relatively large budget ceiling had a higher probability of rejection. The findings of this study can serve as a reference for KPPN Palu in developing technical guidance to minimize administrative errors based on the analyzed attributes. Keywords: Applied Machine Learning, Data Mining, State Budget, Payment Order ©2025 The Authors. This work is made available for academic use only.