CallShield - Real Time Fraud Call & OTP Scam Detection App

Abstract

Spam calls and OTP scams have become a major concern in recent years, affecting millions of mobile users around the world. Fraudsters often pretend to be representatives of banks or trusted organizations and trick people into revealing sensitive details such as one-time passwords (OTPs), PINs, or account numbers. Many existing spam detection methods depend on number blocking or community reports, which can only identify scams that have already been detected. These traditional methods fail to protect users from new or unknown fraud numbers.

This review paper studies different approaches used for spam call and message detection, including keyword-based filtering, machine learning, and natural language processing (NLP) techniques. It also reviews the growing use of artificial intelligence (AI) for real-time fraud detection. The paper focuses on the CallShield concept, a mobile-based system that uses speech- to-text conversion, keyword spotting, and a spam number database to detect and alert users during live calls. When suspicious words like “OTP,” “PIN,” or “bank account” are detected, the system immediately warns the user to stay cautious. The review highlights the advantages of AI-powered detection, identifies gaps in existing research, and emphasizes the importance of developing secure, real-time systems to reduce financial loss and increase user awareness against fraud calls and OTP scams.