Paper Contents
Abstract
Fraudulent signups on digital platforms pose risks beyond financial losses, affecting data security, system integrity, and platform metrics. This research presents a multi-layered fraud detection model analyzing email, IP, and browser data to identify suspicious profiles at signup. Email verification employs SMTP-MX validation and domain analysis using NLP, trained on datasets of 170,000 disposable and 430,000 trusted domains. IP analysis involves reverse DNS lookup, DNSBL filtering, port scans, and latency checks to flag VPN or proxy use. Browser analysis detects anomalies in device configurations and anti-fingerprinting measures. Using logistic regression and random forest classifiers, the model generates fraud scores, enabling early detection and reducing risks like data breaches and metric inflation.
Copyright
Copyright © 2024 Archish Sharma. This is an open access article distributed under the Creative Commons Attribution License.
