Technologies Used:

• Backend: PHP / Java / Node.js

• Frontend: HTML, CSS, Bootstrap, JavaScript

• ML Tools: Python, Scikit-learn, Pandas, NLTK / spaCy

• Database: MySQL / MongoDB

• Optional APIs: Gmail API / IMAP (to fetch emails)

Project Objective:

To develop a machine learning-based web application that classifies emails as fraudulent (phishing/spam) or genuine, helping users and organizations prevent cyber-attacks and data breaches.

Key Features:

1. Email Content Upload:
   Users can manually paste email content or connect their email account (optional) to scan received messages.

2. Fraud Detection Engine:
   An ML classifier identifies fraud indicators in the email’s subject, body, and metadata.

3. Highlight Suspicious Elements:
   The system highlights links, keywords, or patterns typically found in phishing/scam emails.

4. Real-time Classification:
   Instant output labeling an email as:

   •  Genuine

   •  Suspicious

   •  Fraudulent

5. Admin Panel:
   Admins can view flagged emails, model accuracy, and user-reported spam.

6. Feedback Loop:
   Users can mark wrongly classified emails to retrain and improve the model.

Dataset for Training:

Use publicly available datasets such as:

• Enron Email Dataset

• SpamAssassin Public Corpus

• Kaggle’s "Spam Email" datasets
  Features include:

• Subject line

• Email body text

• Sender domain

• Links and attachments

• Word frequency and pattern usage