Intelligent Anti-Money Laundering Fraud Control Using Graph-Based Machine Learning Model for the Financial Domain

Abstract

Financial domains are suffering from organized fraudulent activities that are inflicting the world on a larger scale. Basel Anti-Money Laundering (AML) index enlists 146 countries, which are impacted by criminal acts like money laundering, and represents the country's risk level with a notable deteriorating trend over the last five years. Despite AML being a substantially focused area, only a fraction of such activities has been prevented. Because financial data related to this field is concealed, access is limited and protected by regulatory authorities. This paper aims to study a graph-based machine-learning model to identify fraudulent transactions using the financial domain's synthetic dataset (100K nodes, 5.3M edges). Graph-based machine learning with financial datasets resulted in promising 77-79% accuracy with a limited feature set. Even better results can be achieved by enriching the feature vector. This exploration further leads to pattern detection in the graph, which is a step toward AML detection.

Financial Anti-Fraud Based on Dual-Channel Graph Attention Network

“…As illustrated in Figures 7-9, it is evident that the accuracy, recall, and F1 scores of each algorithm exhibit an initial increase followed by stabilization with the variation in iteration cycles. In the comparative analysis involving model algorithms proposed by GCN, BiLSTM, CNN, and the algorithm presented by Usman et al (2023) [17], the accuracy of the model algorithm developed in this article reaches 93.82%. The classification recognition performance of each algorithm, ranked from highest to lowest, is as follows: model algorithm in this article > Usman et al (2023) [17] algorithm > GCN > BiLSTM > CNN.…”

“…The recognition performance of the model proposed in this article is compared with GCN, BiLSTM, CNN, and the model algorithm proposed by Usman et al (2023) [17] from related literature. The convergence situations are illustrated in Figure 6.…”

“…Their focus on safeguarding sensitive data during anti-fraud procedures maintained privacy while preserving an effective fraud detection mechanism. Exploring graph-based machine learning, Usman et al (2023) [17] proposed an intelligent anti-money laundering fraud control model. Leveraging graph-based machine learning technology, their model elevated anti-money laundering measures, consequently enhancing fraud control and compliance in financial transactions.…”

“…To assess the performance of the model proposed in this article, the model algorithm was compared with GCN, BiLSTM [35], CNN [36], and the model algorithm proposed by Usman et al [17]. (2023) from related literature.…”

Anti-money laundering method based on hierarchical risk control knowledge graph

“…Alarab et al 12 integrated the node embeddings of GCN with the original feature embeddings of transaction sequences. Usman et al 13 verified that GCN can better learn hidden structural features in transaction networks. Li et al 14 used a combination of GCN and Recurrent Neural Network (RNN) to identify money laundering transactions.…”