SCAM-VISION: HYBRID AI-BASED REAL-TIME DETECTION OF FRAUDULENT UPI IDS, SUSPICIOUS PHONE NUMBERS, AND QR CODE PAYMENT SCAMS

Read full text Summarise with VIRAT
Authors: Rajakumar  Varshitha, Pedamutti Rajesh, Veeranki Vasavi, Dr. Shaik Nazeer, Reddy Nandini
Cited by: 0
References: 13

Summary

No summary yet. VIRAT can read this paper and write a plain-language summary of its aim, methods, key findings and limitations.

Summarise with VIRAT Free to try with your Science Catalogs account.

Abstract

The rapid growth of digital payment systems has significantly increased the risk of financial fraud, particularly through fraudulent Unified Payments Interface (UPI) IDs and deceptive phone numbers.Traditional fraud detection methods often rely on static rule-based systems, which fail to identify newly evolving scam patterns.This paper presents SCAM-VISION, a real-time, AI-driven hybrid fraud detection framework designed to identify suspicious UPI identifiers and phone numbers with high accuracy.The proposed system integrates three complementary detection layers: rule-based keyword analysis, trusted database verification, and deep learning-based anomaly detection using a Variational Autoencoder (VAE).Character-level TF-IDF vectorization is employed to capture structural patterns in legitimate UPI handles, enabling the VAE to detect subtle deviations commonly used in scam identifiers.Additionally, phone number verification is enhanced through carrier and line-type analysis to detect VoIP-based and foreign scam numbers.SCAM-VISION supports QR code scanning for automatic UPI extraction and provides a real-time web dashboard that displays risk scores, detection reasons, and performance metrics.Experimental evaluation demonstrates that the hybrid approach significantly improves detection reliability compared to standalone rule-based methods, achieving higher precision and reduced false positives.The system is lightweight, scalable, and suitable for real-world deployment, offering an effective solution for enhancing digital payment security.

Fields of Study & Subjects

Spam and Phishing Detection Cybercrime and Law Enforcement Studies Imbalanced Data Classification Techniques

Publication details

Journal / source
International journal of engineering science and advanced technology.
Volume
26
Issue
3
Pages
380–386
Publication date
1 March 2026
Language
EN