🔍 BiasLens

An AI-powered web application that detects bias in news articles using machine learning. Built with Random Forest models trained on 3.6M samples from the BEAD dataset.

🚀 Features

• Real-time Bias Detection: Instant analysis of news articles
• Dual-Model System:
  • Classification (Neutral vs Biased)
  • Regression (Bias Percentage 0-100%)
• High Accuracy: 71.1% accuracy on 3.6M sample dataset
• Confidence Scoring: Model certainty metrics
• Web Interface: User-friendly Flask web application
• Production Ready: 17+ hour trained models

📊 Model Performance

| Metric | Value | |--------|-------| | Overall Accuracy | 71.1% | | Biased Precision | 81% | | Neutral Precision | 66% | | Training Dataset | 3.6M samples (BEAD) | | Features | 15K TF-IDF features | | Training Time | 17+ hours |

🛠️ Technology Stack

• Backend: Python, Flask
• ML Models: Scikit-learn Random Forest
• Text Processing: TF-IDF Vectorizer (15K features, bigrams)
• Dataset: BEAD (Bias and Emotion Analysis Dataset)
• Frontend: HTML, CSS, JavaScript
• Data Processing: Pandas, NumPy

📁 Project Structure

BiasLens/
├── src/                    # Core application
│   ├── app.py             # Flask web server
│   ├── train_model.py     # Model training script
│   ├── load_bead_datasets.py  # Dataset loader
│   └── templates/         # Web UI templates
├── models/                # Trained ML models
│   ├── bias_classifier.pkl   # Classification model
│   ├── bias_regressor.pkl    # Regression model
│   └── vectorizer.pkl        # TF-IDF vectorizer
├── data/                  # Dataset files
├── analysis/              # Jupyter notebooks
├── docs/                  # Documentation
└── requirements.txt       # Dependencies

🚀 Quick Start

1. Clone Repository

git clone https://github.com/YOUR_USERNAME/BiasLens.git
cd BiasLens

2. Install Dependencies

pip install -r requirements.txt

3. Run Application

cd src
python3 app.py

4. Open Browser

Navigate to http://localhost:8080

📖 Usage

Web Interface

1. Open the web application
2. Paste your news article text
3. Click "Analyze Bias"
4. Get instant results:
   • Bias Level: Neutral or Biased
   • Bias Percentage: 0-100% intensity
   • Confidence: Model certainty
   • Probabilities: Detailed breakdown

API Usage

import requests

response = requests.post('http://localhost:8080/predict', 
                        json={'text': 'Your news article here'})
result = response.json()
print(result['bias'])

🎯 Model Details

Classification Model

• Algorithm: Random Forest (200 trees)
• Purpose: Neutral vs Biased classification
• Accuracy: 71.1%
• Features: 15K TF-IDF features with bigrams

Regression Model

• Algorithm: Random Forest Regressor (200 trees)
• Purpose: Bias intensity (0-100%)
• MAE: 21.84%
• R² Score: 0.2489

Training Configuration

# Classification Model
RandomForestClassifier(
    n_estimators=200,
    max_depth=25,
    min_samples_split=5,
    min_samples_leaf=2,
    random_state=42
)

# TF-IDF Vectorizer
TfidfVectorizer(
    max_features=15000,
    stop_words='english',
    ngram_range=(1, 2),
    min_df=3,
    max_df=0.95
)

📊 Dataset

BEAD (Bias and Emotion Analysis Dataset)

• Size: 3,674,925 samples
• Source: Hugging Face (shainar/BEAD)
• Labels: Neutral, Slightly Biased, Highly Biased
• Distribution: 51.3% Neutral, 48.7% Biased

🔧 Development

Retrain Models

cd src
python3 train_model.py

Load New Dataset

cd src
python3 load_bead_datasets.py

Run Analysis

jupyter notebook analysis/data_visualization.ipynb

📈 Results Examples

{
  "bias": {
    "prediction": "Biased",
    "bias_percentage": 47.8,
    "confidence": 0.642,
    "probabilities": {
      "Neutral": 0.358,
      "Biased": 0.642
    }
  },
  "summary": "Article summary..."
}

🤝 Contributing

1. Fork the repository
2. Create feature branch (git checkout -b feature/amazing-feature)
3. Commit changes (git commit -m 'Add amazing feature')
4. Push to branch (git push origin feature/amazing-feature)
5. Open Pull Request

📄 License

This project is licensed under the MIT License - see the LICENSE file for details.

🙏 Acknowledgments

• BEAD Dataset: Shainar et al. for the comprehensive bias dataset
• Hugging Face: For dataset hosting and tools
• Scikit-learn: For robust machine learning algorithms

📞 Contact

• GitHub: @YOUR_USERNAME
• Email: your.email@example.com

---

⭐ Star this repository if you found it helpful!