<![CDATA[This study aims to develop an Artificial Intelligence (AI)-based anomaly detection system to enhance financial transparency in government credit card transactions within ministries and government agencies. Although the government has implemented the integrated SAKTI financial management system, transaction verification processes still rely heavily on manual procedures, creating opportunities for fraud, irregularities, and weaknesses in financial oversight. To address this issue, the study proposes an anomaly detection system utilizing the Isolation Forest algorithm to identify suspicious transactions related to official travel expenditures and operational material purchases. The research adopts an unsupervised machine learning approach, enabling anomaly detection without requiring labeled data. The analysis is conducted using transaction datasets representing government credit card activities and focuses on identifying unusual transaction patterns that may indicate potential misuse or fraudulent behavior. The findings demonstrate that the proposed model effectively identifies anomalous transactions and supports the detection of potential financial irregularities. The study concludes that AI-based anomaly detection systems can strengthen internal control mechanisms, improve accountability, and support more transparent and efficient financial management. The proposed system also offers practical value as an early warning tool for government institutions seeking to enhance fraud prevention and financial governance.]]>
Copyrights © 2026
