vikasgautam2003/Text_Classification_Using_FastText

📰 Fake or Real News Classifier Using FastText

⚡ A text classification project that classifies news as Fake or Real using FastText.

---

🚀 Features

✅ Train and test a FastText classifier
✅ Support for CSV datasets
✅ Preprocessing and simple text tokenization
✅ Save/load trained FastText model
✅ Ready for deployment or further experimentation

---

🛠️ Tech Stack

• Model → FastText
• Language → Python 3.10+
• Libraries → pandas, fasttext
• Dataset → Custom CSV dataset (ecommerce_dataset.csv)

---

📂 Project Structure

├── .ipynb_checkpoints/           # Jupyter notebook checkpoints
├── TextClassification.ipynb       # Main notebook
├── TextClassificationPart2.ipynb  # Extended analysis notebook
├── requirements.txt               # Python dependencies
├── .gitattributes                 # Git LFS configuration
├── ecommerce.train                # Training dataset (LFS)
├── ecommerce.test                 # Test dataset (LFS)
├── ecommerce_dataset.csv          # Full dataset (LFS)
└── README.md                      # Project documentation


⚙️ Setup Instructions
1️⃣ Install Dependencies
pip install -r requirements.txt

2️⃣ Run Training / Evaluation

Open the Jupyter notebooks:

jupyter notebook TextClassification.ipynb
jupyter notebook TextClassificationPart2.ipynb


Make sure to have datasets available locally (ecommerce.train, ecommerce.test, ecommerce_dataset.csv) or tracked via Git LFS.

🎯 Use Cases

Detect fake news in real-time applications

Build NLP portfolio projects

Experiment with FastText embeddings and text classification

💡 Future Improvements

Add real-time API for classification

Integrate with Hugging Face Datasets

Hyperparameter tuning for better accuracy

Add visualization dashboards for analysis

👨‍💻 Author

Vikas Gautam