Fraud Detection and Prevention in E-commerce using Decision Tree Algorithm

Abstract

Abstract: Fraud detection is an important part of e-commerce because it helps prevent fraud such as illegal transactions, identity theft, and money laundering. Recently, there has been a lot of literature on the application of machine learning algorithms to identify e-commerce fraud. These algorithms work by learning patterns in data that indicate fraud. Pattern checking deals with discovering differences in data, such as unusual products, locations, or behavior outside the norm for certain users, through machine learning. In this project, we propose a decision tree algorithm to detect fraud in e-commerce using newly generated data from various online products on e-commerce sites. In addition to fraud detection, we also provide advice on fraud prevention. We propose a new security model that will prove the user's identity. In this security model, users are required to register their profile with some questions. Our security systems will display relevant images in response to the registration question. The user has to click on the correct answer image within the time limit. We will ask the user 3 questions in graphic format. If the user selects the correct answer, the user will be considered a real user.