An AI-powered Email Security Platform that detects Spam, Phishing Attempts, Suspicious URLs, and Malicious Keywords using Machine Learning and Natural Language Processing (NLP).
- Machine Learning based email classification
- Spam/Ham prediction
- Spam probability score (%)
- URL extraction from email content
- Shortened URL detection
- IP-based URL detection
- Suspicious domain identification
- Suspicious keyword highlighting
- Phishing phrase detection
- Risk keyword scoring
- Security score calculation
- Threat level classification
- Spam vs Safe email statistics
- Interactive analytics dashboard
- Random Forest
- Support Vector Machine (SVM)
- Logistic Regression
- Naive Bayes
-
Accuracy comparison charts
-
Confusion Matrix visualization
-
Performance metrics
- Accuracy
- Precision
- Recall
- F1 Score
- Security analysis reports
- Dataset statistics
- Model performance summary
- HTML5
- CSS3
- JavaScript
- Chart.js
- Flask
- Scikit-Learn
- Pandas
- NumPy
- Joblib
- TF-IDF Vectorization
- Text Preprocessing
- Keyword Analysis
Clone the repository:
git clone https://github.com/harshu-belal/smart-email-security-system.git
cd smart-email-security-systemInstall dependencies:
pip install -r requirements.txtRun the application:
python app.pyOpen:
http://127.0.0.1:5000
Security Score =
100
− (Spam Score × 0.40)
− (Phishing Score × 0.35)
− (Keyword Score × 0.25)
Dataset used:
SMS Spam Collection Dataset
Contains:
- Spam messages
- Legitimate (Ham) messages
Used for training and evaluating machine learning models.
- PDF Report Generation
- Email File Upload (.txt)
- Real-Time Threat Monitoring
- Database Integration
- User Authentication
- Cloud Deployment
- Advanced NLP Models
Harshit Belal B.Tech CSE (AI & ML)
This project is developed for educational and research purposes.
© 2026 Harshit. All Rights Reserved.