Phishing Email Detection Tool Using Text Classification

Phishing remains the front door for most security breaches, and the people most at risk, including new employees, students, and older users, are the least equipped to spot a well-crafted lure. A tool that instantly evaluates a suspicious email provides genuine protection while teaching the vigilance that stops the next attempt. The intern builds a Flask web app in Python where users paste email text and a machine learning classifier predicts whether it is a phishing attempt. The intern assembles a labeled corpus of phishing and legitimate emails, engineers features from the text such as urgency language, suspicious links, and sender impersonation patterns, and trains a text classification model with standard data science libraries. The interface returns a verdict with a confidence score and highlights the phrases that influenced the decision, turning every check into a small lesson in phishing awareness. The project delivers a grounded introduction to both natural language processing and security: building text classifiers, evaluating them against adversarial examples that try to slip past, explaining model decisions to users, and packaging machine learning as a practical safety tool anyone can operate.