BEHAVIOR-DRIVEN DEVELOPMENT FRAMEWORK FOR BANKING FRAUD DETECTION: AN ARCHITECTURAL DESIGN WITH SERENITY AND SELENIUM

Abstract

In this study, a Behavior-Driven Development (BDD) framework to detect banking fraud, supporting Serenity BDD and Selenium was developed and proved. The aim was to improve automation features of detecting frauds in digital banking processes through incorporation of structured BDD scenarios, automated UI interactions, and reportable systems. A hypothetical banking setting and various fraud conditions- internal fraud, external attacks and behavior anomalies- were run to test the performance of the framework. The findings showed high success rate of execution, high accuracy of fraud detection and significant time saving of testing as opposed to manual validation. Serenity reporting enhanced better visibility and traceability of defects, whilst Selenium supported realistic simulation of transactions. Despite a few false-negative and false-positive cases that indicated that additional refinements can be done on the anomaly rules, the general results indicated that BDD-based automation is a scalable, efficient, and reliable methodology to enhance the financial institutions fraud detection systems.