This interactive web app analyzes movie reviews and predicts their sentiment—Positive 😊 or Negative 😞—using a trained machine learning model and natural language processing. Built with Streamlit, it includes real-time predictions, data visualization, and optional voice playback.
- ✅ Clean NLP preprocessing with BeautifulSoup, regex, and stopword removal
- 🎓 Machine Learning using a scikit-learn classifier and TF-IDF vectorizer
- 📊 Summary stats with visual feedback (positive/negative proportions)
- 💬 Text-to-speech playback of predictions (browser-based voice)
- 💾 Save and display recent reviews and predictions
- 📁 Export all analyzed reviews to CSV
🖥️ Deploy your own version or use this repo with Streamlit Cloud.
Just make sure your main app script is correctly set in the app configuration (e.g., sentiment_webapp.py or app.py).
Install dependencies with: streamlit scikit-learn pandas beautifulsoup4 nltk pickle-mixin html5lib numpy requests streamlit==1.33.0 scikit-learn==1.4.1.post1 pandas==2.2.2 nltk==3.8.1 beautifulsoup4==4.12.3 dir
pip install -r requirements.txt
# imdb-sentiment-app