📰 Fake News Detector (BERT + Streamlit)

A real-time fake news detection web application built using a fine-tuned BERT model with an elegant Streamlit interface. The system supports manual input and URL-based article analysis with multilingual support and clean visual feedback.

The Fake News Detection Web Application is a locally deployed tool that uses traditional machine learning algorithms like Logistic Regression and Random Forest to classify news articles as real or fake. It offers a simple web interface for real-time predictions, aiming to combat misinformation and promote media awareness.

---

🧠 Model Architecture: BERT-Powered Fake News Detection System

🧪 Screenshots

🔍 Features

• 🧠 BERT-based classification of news articles as Fake or Real
• 🌐 URL scraping using newspaper3k and auto language detection
• ✨ Real-time feedback with model-wise comparison: Logistic Regression, Random Forest, SVM, Naive Bayes, and BERT
• 📊 Dynamic confidence bars (green = Real, red = Fake)
• 🌍 Auto-translation for non-English news using Google Translate
• 📱 Streamlit-powered UI with dark mode and modern UX

---

🛠️ Tech Stack

• Frontend: Streamlit
• ML Models:
  • Logistic Regression
  • Random Forest
  • Naive Bayes
  • SVM
  • BERT (transformers by HuggingFace)
• Vectorizers: TfidfVectorizer, CountVectorizer
• Tools: joblib, torch, langdetect, deep-translator, newspaper3k

---

🚀 How to Run

1. Clone the Repository

git clone https://github.com/PankajMahanto/FakeNewsDetection.git
cd FakeNewsDetection

2. Create a Virtual Environment (Optional but Recommended)

python -m venv venv
source venv/bin/activate  # On Windows: venv\Scripts\activate

3. Install Requirements

bash
pip install -r requirements.txt

4. Run the App

bash
streamlit run app.py

📁 Project Structure

bash

|—— .gitattributes
|—— app.ipynb
|—— app.py
|—— app_rf.ipynb
|—— bert_fakenews
|    |—— config.json
|    |—— model.safetensors
|    |—— special_tokens_map.json
|    |—— tokenizer.json
|    |—— tokenizer_config.json
|    |—— vocab.txt
|—— bert_fakenews_vs.ipynb
|—— copy-app.py
|—— Datasets
|    |—— Fake.csv
|    |—— True.csv
|—— Datasets.zip
|—— lr_model.jb
|—— model.jb
|—— model_load.ipynb
|—— nb_model.jb
|—— r.md
|—— requirements.txt
|—— rf_model.jb
|—— rf_model.zip
|—— svm_model.jb
|—— tfidf_vectorizer.jb
|—— train_bert.py
|—— train_bert_news.ipynb
|—— Train_nb.ipynb
|—— train_svm.ipynb
|—— vectorizer.jb
|—— vectorizer_nb.jb
|—— vectorizer_rf.jb
|—— visualization_of_project.ipynb

Code Details

Tested Platform

• software

  OS: Debian unstable (May 2021), Ubuntu LTS
  Python: 3.8.5 (anaconda)
  PyTorch: 1.7.1, 1.8.1
  

• hardware

  CPU: Intel Xeon 6226R / Normal CPU work
  GPU: No Needed
  RAM: At least 8GB
  

👤 Author

Made with ❤️ by Pankaj Mahanta