An Improved Solution to Detect Credit Card Fraud Using Apache Hadoop in Big Data Environment

Abstract

This paper presents an improved approach for identifying the pattern and detecting an online credit card fraud. Recent years have seen increasing amounts of data generated and stored in a geographically distributed manner for a large variety of application domains. Examples include social networking, Web and Internet service providers, and content delivery networks that serve the content for many of these services. This paper focus on designing an online credit card fraud detection framework with technologies, by which this can process large amount of data and to do detection in real time and to improve accuracy based on analyzing the factors such a processing speed, latency, fault tolerance, performance and scalability. On behalf of an evaluation about the techniques it was proposed that Apache Spark is performing better on Credit card fraud detection system when compared to other techniques or frameworks. Real time analysis is highly desirable to update models when new events are detected.