AI - DRIVEN TELECOM THREAT DETECTION APP A Modular, Intelligent Framework for Mobile Security And Real -Time Fraud Prevention

I.               ABSTRACT:

With the increasing dependence on mobile communication, telecom users are increasingly exposed to security threats such as spam calls, smishing attacks, and caller ID spoofing. This project presents an AI-driven Telecom Threat Detection System designed to enhance user security through real-time monitoring and intelligent analysis.

The system consists of an Android application, a Flask-based backend API, and a machine learning engine. The Android app captures call logs and SMS data (with user permission) and forwards relevant information to the backend using REST APIs. The backend processes requests such as call analysis, SMS keyword scanning, and spoof detection.

The ML engine applies keyword-based filtering and scoring mechanisms to evaluate the likelihood of threats, generating a risk score between 0 and 100. The system also includes a chatbot module that provides clear, user-friendly explanations of detected threats, improving user awareness and decision-making.

By combining real-time data collection, machine learning analysis, and explainable AI, the proposed system offers a proactive and scalable solution to combat telecom-based fraud and cyber threats. This approach enhances both security and usability, making it suitable for modern mobile environments.

Keywords:

AI-driven threat detection, telecom security, spam call analysis, SMS phishing (smishing), caller ID spoofing detection, modular architecture, Android monitoring, Flask REST API, machine learning scoring, keyword-based filtering, real-time analysis, anomaly detection, risk scoring (0–100), explainable AI, chatbot interface, secure data transmission, Retrofit integration, JSON processing, user privacy, cybersecurity awareness.