<![CDATA[Background: Financial transactions in the mineral and coal mining sector will become increasingly complex, creating the potential for difficult-to-detect money laundering practices. Therefore, methods are needed to identify the reasonableness of the transaction relationship between the sender and recipient of funds. Objective: This research aims to design and implement a graph-based fuzzy Mamdani model to assess the risk level of transaction relationships that indicate money laundering in the mineral and coal mining sector. Methods: The research employed a quantitative modelling approach. The dataset consists of financial transaction reports obtained from the Indonesian Financial Transaction Reports and Analysis Center (INTRACT). The graph-based Mamdani Fuzzy Inference System (FIS) is applied to assess the risk level of transaction relationships. The model is evaluated with Mean Squared Error (MSE), Root Mean Squared Error (RMSE), and Coefficient of Determination (R²). Results: The results show FIS Mamdani model achieved superior performance with an MSE of 0.07, RMSE of 0.26, and R² of 0.93. Testing results with new data whose risk levels are known based on expert assessments and court decisions, show that all risk categories generated by the model are fully aligned with the classifications established based on professional evaluations. Conclusion: These findings confirm that the graph-based Mamdani fuzzy model is an adaptive and accurate tool for assessing transaction relationship risk, which can be used to detect money laundering in the mineral and coal mining sector. This research contributes to the advancement of graph-based financial transaction monitoring systems and expert systems, enhancing their ability to handle complex and dynamic transaction relationship risks. Further research is recommended to explore integration with advanced machine learning techniques such as deep learning. Keywords: Graph-Based Model, Mamdani Fuzzy, Money Laundering Risk, Mining Transactions, Artificial Intelligence ]]>
