Twitter_Sentimental_Analyzer is a Python-based project that performs sentiment analysis on tweets — classifying them as positive, negative, or neutral.
It includes preprocessing steps, a trained ML model, and a simple web interface to test sentiment predictions.
- Pretrained sentiment analysis model (
trained_model.sav) - Vectorizer for text transformation (
vectorizer.pkl) - Complete training workflow in Jupyter Notebook
- Web-app interface (Streamlit / Flask)
- Easy to extend, retrain, and integrate
This project uses a publicly available sentiment dataset from Kaggle:
🔗 Kaggle Dataset Link:
https://www.kaggle.com/datasets/kazanova/sentiment140
You can download the dataset and place it in your project folder before running the notebook.
| File / Folder | Description |
|---|---|
twitter_sentiment_analysis.ipynb |
End-to-end model training + preprocessing. |
vectorizer.pkl |
Saved text vectorizer used during training. |
trained_model.sav |
Final trained model for prediction. |
streamlit_app.py / app.py |
Web interface for sentiment prediction. |
kaggle.json |
API key for Kaggle dataset download (optional). |
| Other files | Supporting scripts/config. |
pip install -r requirements.txt