The Proposed System Architecture Of PhishGuard-AI Follows A Structured Pipeline That Connects Backend

The proposed system architecture of PhishGuard-AI follows a structured pipeline that connects backend intelligence with an interactive frontend for real-time phishing detection. It starts with the Input Layer, where the user uploads or enters an email through the web interface. The email is sent to the backend via an API for analysis. In the Preprocessing and Feature Extraction Layer, the text is cleaned by removing noise like special symbols and stopwords, then converted into numerical vectors using TF-IDF, allowing the system to process it mathematically. The Machine Learning Layer uses a trained Random Forest Classifier to determine whether the email is phishing or safe. It also provides a confidence score indicating the certainty of the prediction. Finally, the Result Generation Layer sends the outcome back to the Frontend Interface, where the result and confidence percentage are displayed clearly. This end-to-end architecture effectively combines Natural Language Processing (NLP) and Machine Learning to deliver accurate, explainable, and user-friendly phishing detection through an interactive dashboard.

04.11.2025 18:35