<![CDATA[Market service levies are one of the primary sources of Local Own-Source Revenue (PAD) used to develop and revitalize markets, as well as improve comfort, security, and local economic growth. Although the Electronic Transaction System for Local Government (ETPD) has been implemented in South Tangerang City to facilitate payments through various digital channels, the payment compliance rate remains low. In 2023, South Tangerang's market service levy revenue only reached 29.94% of the set target. Of this amount, only 16.86% of levy obligors paid regularly and on time, while 83.14% failed to make timely payments. This condition may hinder optimal regional revenue generation. This study aims to develop a classification model that can predict compliance in market service levy payments in South Tangerang City. The research utilizes the Random Forest algorithm and follows the Cross-Industry Standard Process for Data Mining (CRISP-DM) framework to classify market service levy data from November 2021 to November 2024 obtained through the ETPD system. The developed model classifies payment compliance into two categories: timely and delinquent. Analysis results using K-Fold Cross Validation show that the Random Forest model achieves strong performance metrics with 94.94% accuracy, 99.49% precision, 90.34% recall, 94.69% F1-score, and 95% AUC, indicating the model's excellent classification capability. These findings can serve as recommendations for the South Tangerang City government to formulate strategies for improving levy payment compliance. KEYWORDS]]>
Copyrights © 2026
