Agentic-News-Bot

A personalized press agent powered by AI, featuring news generation, fake news detection, and press conference simulation capabilities.

🎯 Key Features

• News Generation: Automated news creation
• Fake News Detection: ML-powered detection to identify unreliable news articles
• Press Conference Simulator: Interactive press conference simulation system

Note: At the moment This repository currently contains the complete fake news detection implementation. Other features (news generation and press conference simulator) are yet to be integrated .

📁 Project Structure

Agentic-News-Bot/
├── app.py                          # Main Flask application
├── requirements.txt                # Python dependencies
├── .env                            # Environment variables (not tracked)
├── .gitignore                      # Git ignore rules
│
├── architecture/                   # Project setup and documentation
│   └── project-structure-script.sh # Script to generate project structure
│
├── notebooks/                      # Jupyter notebooks for exploration
│   └── fake-news-detection.ipynb  # Fake news detection analysis
│
├── src/                            # Source code
│   ├── agents/                     # AI agents
│   │   └── news_prediction_agent.py
│   │
│   ├── data/                       # Datasets
│   │   └── News_dataset/
│   │       ├── Fake.csv            # Fake news samples
│   │       └── True.csv            # True news samples
│   │
│   ├── embeddings/                 # Text embedding models
│   │   └── embed_model.py
│   │
│   └── models/                     # Trained ML models
│       ├── best_model.pkl          # Best performing model
│       ├── logisticRegressor.pkl   # Logistic regression model
│       ├── minmax_scaler.pkl       # Feature scaler
│       ├── tfidf_vectorizer.pkl    # TF-IDF vectorizer
│       └── embedding_model/        # Pre-trained sentence transformer
│
├── templates/                      # HTML templates
│   └── index.html                  # Web interface
│
├── tests/                          # Test files
│   └── news_prediction.py          # Prediction tests
│
└── utils/                          # Utility functions
    ├── data_preprocessing.py       # Data cleaning and preprocessing
    ├── data_validation.py          # Input validation
    ├── simulation_helpers.py       # Simulation utilities
    └── train_and_save_model.py     # Model training pipeline

Quick Setup

To recreate the project structure from scratch, run:

bash architecture/project-structure-script.sh

🚀 Getting Started

Prerequisites

• Python 3.8+
• pip package manager

Installation

1. Clone the repository:

git clone https://github.com/IyedGuezmir/Agentic-News-Bot.git
cd Agentic-News-Bot

2. Install dependencies:

pip install -r requirements.txt

3. Set up environment variables:

cp .env.example .env
# Edit .env with your configuration

Running the Application

python app.py

The application will be available at http://localhost:5000

🧠 Fake News Detection

The fake news detection system uses a hybrid approach:

Agent Workflow

1. Text Embedding: News articles are converted to semantic embeddings using all-MiniLM-L6-v2 Sentence Transformer
2. ML Prediction: Pre-trained classifier predicts if the news is fake or true with confidence score
3. Web Verification: LLM (GPT-4) with web search tools verifies the news against credible online sources
4. Final Decision: If web verification finds credible sources, marks as True News; otherwise, defers to ML model prediction

Key Components

• Sentence Transformers (all-MiniLM-L6-v2): For semantic text embeddings
• Pre-trained ML Classifier: For initial prediction
• LangChain + OpenAI GPT-4: For intelligent web-based verification
• Hybrid Decision Logic: Combines ML predictions with real-time web verification

🤝 Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

📝 License

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