musfiqurtuhin/bengali-fake-news

🇧🇩 Bengali Fake News Detection Dataset

📄 Context

Official Novel dataset for the research paper:

"Bengali Fake News Detection: A Multi-Layered LSTM Ensemble Approach" Published in: 2025 International Conference on Quantum Photonics, Artificial Intelligence, and Networking (QPAIN)

This dataset focuses on mainstream Bangladeshi journalism, aggregating news from credible top TV channels and newspapers to provide a robust benchmark for authentic vs. fabricated news detection.

---

🔗 Associated Resources

Resource	Link
📄 Paper (IEEE Xplore)
🐙 GitHub Repository
🐍 Kaggle Dataset

---

💾 Dataset Specifications

Feature	Details
Total Documents	11,839
Authentic (Label 1)	6,000 documents
Fake (Label 0)	5,839 documents
Total Words	187,127 unique words
Language	Bengali (Bangla)

📝 Column Description

Column	Description
domain	The source publication or website.
category	The news category.
headline	The title of the news article.
content	The full text of the news article.
label	1 for Authentic, 0 for Fake.

---

🚀 Quick Start

from datasets import load_dataset

# Load the dataset
dataset = load_dataset("musfiqurtuhin/bengali-fake-news")

# View a sample
print(dataset['train'][0])

---

🤝 Citation

If you use this dataset, please cite our work:

@INPROCEEDINGS{11171927,
  author={Rinky, Susmita Roy and Tusher, Md. Mahbubur Rahman and Rahman, Md. Musfiqur and Porag, Prithweeraj Acharjee and Islam, Jahidul and Miah, Abu Saleh Musa},
  booktitle={2025 International Conference on Quantum Photonics, Artificial Intelligence, and Networking (QPAIN)}, 
  title={Bengali Fake News Detection: A Multi-Layered LSTM Ensemble Approach}, 
  year={2025},
  pages={1-6},
  doi={10.1109/QPAIN66474.2025.11171927}
}