Service-Based Credit Card Fraud Detection Using Oracle SOA Suite

Abstract

Credit card fraud detection techniques help to capture fraudulent transactions carried out by illegitimate users and thus prevent any misuse of the credit card. Due to the technological advancement, credit card usage has been on the rise of financial transactions keeping aside the risk of increase in number of fraudulent transactions. Thus, some sort of improved strategies are desired to be implemented to curb and avoid such fraudulent transactions. This study intends to propose a fraud detection technique by implementing various machine learning techniques on cloud platform which itself is based on service-oriented architecture (SOA). SOA helps to create applications by making use of services available over the network. Furthermore, this credit card fraud detection technique, focuses on orchestration of various services using Oracle SOA suite mingled with different machine learning models, such as support vector machine (SVM), isolation forest, random forest regressor, local outlier factor (LOF), and different neural networks, such as multilayer perceptron (MLP), autoencoder, and convolutional neural network (CNN). The outputs of all the machine learning models are integrated with Oracle SOA suite to provide proper agility and efficiency. In addition, this Oracle SOA suite model has been deployed on Google cloud platform (GCP) for providing reliable solution in an online mode. A comparative analysis on performance of different machine learning algorithms has been presented for their critical assessment.

A systematic review of literature on credit card cyber fraud detection using machine and deep learning

“…In addition to fraud location, an algorithm to determine the timing of the fraud is required ( Alghofaili, Albattah & Rassam, 2020 ; Chen & Lai, 2021 ). In addition, an algorithm can be developed to predict fraudulent transactions in a real-time and deploying the service on various cloud platforms to make it easily accessible and reliable ( Ingole et al, 2021 ).…”