Advancing Anomaly Detection in Banking Transactions
Open Access DepositedLeveraging Natural Language Processing and Artificial Neural Network Methods.
With the rise of technology and the convenience of making payments today acrossthe world, it’s easy to fall victim to a scam and make payments that may seem legitimate. As the fraudsters evolve using modern technologies to commit fraud, financial organizations are being pushed to figure out ways to identify such anomalous transactions and stop or review such payments. Governments across the world are addressing payment fraud issues through legislative efforts to hold financial institutions accountable for reimbursing money lost in fraudulent transactions to customers, unless the customer was clearly neglectful. The complexity for financial transactions along with sophisticated fraudulent techniques, the need for advanced approaches for anomaly detections is high. Advanced anomaly detection techniques should adopt modern methods to improve the detection of fraudulent activities. The research uses a modern technique leveraging Natural Language Processing (NLP) model —Bidirectional Encoder Representations from Transformers (BERT), combined with Artificial Neural Network (ANN) models for hybrid learning and refinement with fraud and legitimate customer profiles, to produce a robust and dynamic model for anomaly detection in cybersecurity within the banking domain. Additionally, a Generative Adversarial Network (GAN) will be used to address the imbalance of minority data during training of the model. The robust anomaly model will be evaluated on the performance metrics like precision, recall, and F1-score to assess model efficacy. These measures will determine the feasibility of the model for practical implementation in the finance domain. vi At the end, the research delivers a detailed research report, a trained BERT and ANN based anomaly detection model, and recommendations for practical implementation of the solution in banking cybersecurity systems.
- All rights reserved
Notice to Authors
If you are the author of this work and you have any questions about the information on this page, please use the Contact form to get in touch with us.
| Thumbnail | Title | Date Uploaded | Visibility | Actions |
|---|---|---|---|---|
|
|
Lakkireddy_gwu_0075A_17701.pdf | 2026-02-26 | Open Access |
|