AI Based Phishing and Scam Message Detector: A Machine Learning and NLP Based System to Detect Fraudulent SMS, Email and Chat Messages in Real Time

Abstract: Phishing and scam messages sent through SMS, email, and messaging platforms have become a major cybersecurity threat. These messages are used to steal sensitive information, spread malware, and commit financial fraud. This project proposes an AI-Based Phishing and Scam Message Detector that uses Natural Language Processing (NLP) and Machine Learning techniques to automatically classify incoming messages as legitimate, phishing, or scam. The system extracts linguistic, structural, and URL-based features from message content and trains supervised learning models using labelled datasets. The models are evaluated using accuracy, precision, recall, and F1-score. The proposed solution aims to provide a lightweight, accurate, and easily deployable detection mechanism that can be integrated into messaging applications or used as a standalone tool to alert users before they become victims of online fraud. Keywords: Phishing Detection, Scam Message Detection, Machine Learning, Natural Language Processing, Artificial Intelligence, Cybersecurity, SMS Classification, Email Security, URL Analysis, Fraud Detection Preprint Notice: This manuscript is an open-access scientific preprint published in the Global Youth Science Journal (GYSJ) Preprints repository. It has not undergone formal journal peer review. Permanent Repository URL: https://globalyouthsciencejournal.app/preprint/202611.0016