Anomaly Detection in Graphs of Bank Transactions for Anti Money Laundering Applications

Dumitrescu, Bogdan; Bălțoiu, Andra; Budulan, Stefania

doi:10.1109/access.2022.3170467

Cited by 21 publications

(3 citation statements)

References 47 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Risk classification and outlier detection (considered to be typical machine learning tasks) employ machine learning techniques such as Support Vector Machines, Random Forest, Decision Trees, Bayesian approaches, Gradient Boosting, and regression (Alsuwailem & Saudagar, 2020;Tundis et al, 2021). Link and graph analytics involve the analysis of data in a graph format, using classic graph topology and structure analysis (Dumitrescu et al, 2022), and hybrid graph machine learning pipelines (Eddin et al, 2021).…”

Section: Aml Systemsmentioning

confidence: 99%

An Analysis of Novel Money Laundering Data Using Heterogeneous Graph Isomorphism Networks. FinCEN Files Case Study

Wójcik

2024

Econometrics

View full text Add to dashboard Cite

Aim: This study aimed to develop and apply the novel HexGIN (Heterogeneous extension for Graph Isomorphism Network) model to the FinCEN Files case data and compare its performance with existing solutions, such as the SAGE-based graph neural network and Multi-Layer Perceptron (MLP), to demonstrate its potential advantages in the field of anti-money laundering systems (AML). Methodology: The research employed the FinCEN Files case data to develop and apply the HexGIN model in a beneficiary prediction task for a suspicious transactions graph. The model's performance was compared with the existing solutions in a series of cross-validation experiments. Results: The experimental results on the cross-validation data and test dataset indicate the potential advantages of HexGIN over the existing solutions, such as MLP and Graph SAGE. The proposed model outperformed other algorithms in terms of F1 score, precision, and ROC AUC in both training and testing phases. Implications and recommendations: The findings demonstrate the potential of heterogeneous graph neural networks and their highly expressive architectures, such as GIN, in AML. Further research is needed, in particular to combine the proposed model with other existing algorithms and test the solution on various money-laundering datasets. Originality/value: Unlike many AML studies that rely on synthetic or undisclosed data sources, this research was based on a publicly available, real, heterogeneous transaction dataset, being part of a larger investigation. The results indicate a promising direction for the development of modern hybrid AML tools for analysing suspicious transactions, based on heterogeneous graph networks capable of handling various types of entities and their connections.

show abstract

Section: Aml Systemsmentioning

confidence: 99%

An Analysis of Novel Money Laundering Data Using Heterogeneous Graph Isomorphism Networks. FinCEN Files Case Study

Wójcik

2024

Econometrics

View full text Add to dashboard Cite

show abstract

“…Dumitrescu et al [22] introduced a new anomaly detection method for money laundering applications. The feature extraction method is used here that identifies the important set of features that are presented in the transaction process.…”

Section: Related Workmentioning

confidence: 99%

VTAIM: Volatile Transaction Authentication Insurance Method for Cyber Security Risk Insurance of Banking Services

Almatari

Wang

Zhang

et al. 2023

Preprint

View full text Add to dashboard Cite

Cyber threats in online banking services are increasing with the improvements in internet-aided communications and extended services. The threats gather personal and transactional information from the users from which illegitimate services are handled. For providing secure service-based risk mitigation, this article introduces a Volatile Transaction Authentication Insurance Method (VTAIM). This method generates volatile insurance authenticity for a transaction-initiated session. Depending on the transaction features, the insurance is constructed using a two-way volatile authentication key. This key extension/ validity is recommended through deep learning based on the user transaction interest. The type of security threat is first detected from the session interruptions and converging transaction period. Post the detection, the user and banking service-oriented authentication is used for ensuring end-to-end security. This recommendation is used for typical consecutive transactions and volatile security. The security features are updated periodically based on service availability and transaction support provided. Therefore, spoofed services are less available for distinct users across different time intervals, reducing the false rate and failures.

show abstract

“…(E) Reporting suspicion transactions, most commercial banks have tools to identify normal operation, abnormal operation, and suspicious transaction, based on laws, knowledge, and personal experience with illicit money activity. For instance, one of the alleged money laundering schemes involves shifting money from large to small or vice versa (Dumitrescu, Băltoiu, & Budulan, 2022).…”

Section: Introductionmentioning

confidence: 99%

The effect of the banking supervision on anti-money laundering

Benzerrouk,

Alnor,

Al-Matari

et al. 2023

View full text Add to dashboard Cite

This article's goal is to investigate the connection between banking supervision (BakSup) and anti-money laundering (AnMeyLg). The study's hypothesized model was examined using a survey questionnaire research methodology. The Central Bank of Sudan and the compliance departments of Sudanese banks provided the information. Just 247 valid surveys were received out of a total distribution of 450 questionnaires. The partial least squares method for structural equation modeling was applied. The statistical findings supported the impact of banking supervision on anti-money laundering. In-depth discussions of the study's specifics and consequences were included at various points. There are several real-world applications of the findings of this study. Management can use the findings to guide their decisions on whether or not to establish anti-money laundering procedures in their banking firms. Banking supervision may provide managers with effective anti-money laundering strategies for maximizing bank performance and maintaining market share. This is one of the few empirical studies of its kind that looks at how banking supervision affects efforts to combat money laundering.

show abstract

Anomaly Detection in Graphs of Bank Transactions for Anti Money Laundering Applications

Cited by 21 publications

References 47 publications

An Analysis of Novel Money Laundering Data Using Heterogeneous Graph Isomorphism Networks. FinCEN Files Case Study

An Analysis of Novel Money Laundering Data Using Heterogeneous Graph Isomorphism Networks. FinCEN Files Case Study

VTAIM: Volatile Transaction Authentication Insurance Method for Cyber Security Risk Insurance of Banking Services

The effect of the banking supervision on anti-money laundering

Contact Info

Product

Resources

About