🎬 Sentiment Analysis - RNN Text Classification System

A machine learning-powered Sentiment Analysis System that classifies movie reviews as positive or negative using Recurrent Neural Networks (RNN). This full-stack application features a trained SimpleRNN model, Streamlit web interface, and IMDB dataset integration for accurate sentiment classification with 85.8% accuracy.

Live Demo: https://sentiment-analysis-rnn-lccbnqgswwv7jtccdergwq.streamlit.app/

• ✅ Sentiment Classification - Binary classification of movie reviews (Positive/Negative)
• ✅ Pre-trained RNN Model - SimpleRNN with word embeddings achieving 85%+ accuracy
• ✅ IMDB Dataset - Trained on 50,000 real movie reviews
• ✅ Streamlit Web App - Interactive user interface for real-time predictions
• ✅ Fast Inference - Process predictions in <100ms per review
• ✅ Jupyter Notebooks - Complete training pipeline and analysis notebooks
• ✅ Word Embeddings - 32-dimensional embedding layer for semantic text representation
• ✅ Production Ready - Optimized model weights included (2.3 MB)

Input Text (Review)
    ↓
Embedding Layer (32 dims, 10K vocab)
    ↓
SimpleRNN Layer (128 units, ReLU)
    ↓
Dropout (0.5)
    ↓
Dense Layer (64 units, ReLU)
    ↓
Dropout (0.3)
    ↓
Output Layer (Sigmoid - Binary Classification)
    ↓
Sentiment Label + Confidence Score

# Clone repository
git clone https://github.com/Rahul6128/Sentiment-Analysis-RNN.git
cd sentiment-analysis-rnn

# Install dependencies
pip install -r requirements.txt

streamlit run main.py

Then open: http://localhost:8501

from tensorflow.keras.models import load_model
from tensorflow.keras.preprocessing.sequence import pad_sequences
import pickle

# Load model
model = load_model('simple_rnn_imdb.h5')
tokenizer = pickle.load(open('tokenizer.pkl', 'rb'))

# Predict
text = "This movie was absolutely fantastic!"
sequence = tokenizer.texts_to_sequences([text])
padded = pad_sequences(sequence, maxlen=500)
prediction = model.predict(padded)

sentiment = "Positive 😊" if prediction[0][0] > 0.5 else "Negative 😞"
confidence = prediction[0][0]

print(f"Sentiment: {sentiment}")
print(f"Confidence: {confidence:.2%}")

sentiment-analysis-rnn/
│
├── 📄 README.md                    # Project documentation
├── 📄 LICENSE                      # MIT License
├── 📄 requirements.txt             # Python dependencies
│
├── 🐍 main.py                      # Streamlit web application
├── 🤖 simple_rnn_imdb.h5          # Pre-trained model weights
│
├── 📓 simplernn.ipynb              # Model training pipeline
├── 📓 embedding.ipynb              # Word embedding analysis
├── 📓 prediction.ipynb             # Inference examples
│
└── 📁 images/                      # Project images (optional)

• Source: IMDB Movie Reviews
• Total Reviews: 50,000
• Training Set: 25,000 reviews
• Test Set: 25,000 reviews
• Sentiment Distribution: 50% Positive, 50% Negative
• Preprocessing: Tokenization, padding to 500 tokens, index encoding

Input: "Amazing movie! Best film I've seen all year!"
Output: Positive (98% confidence)

Input: "Worst movie ever made. Total waste of time."
Output: Negative (99% confidence)

Input: "It was okay, had some good parts but mostly boring."
Output: Negative (58% confidence)

tensorflow==2.12.0
keras==2.12.0
streamlit==1.28.0
numpy==1.24.3
pandas==2.0.3
scikit-learn==1.3.0
matplotlib==3.7.2

Install all: pip install -r requirements.txt

Model Configuration:

• Optimizer: Adam (lr=0.001)
• Loss: Binary Crossentropy
• Batch Size: 32
• Epochs: 10
• Validation Split: 20%
• Early Stopping: Yes (patience=2)

Performance Metrics:

• Training Accuracy: 88.5%
• Validation Accuracy: 86.2%
• Test Accuracy: 85.8%

Try the interactive application:

👉 Sentiment Analysis RNN - Live Demo

• Enter any movie review
• Get instant sentiment prediction
• See confidence score
• Real-time analysis

MIT License - See LICENSE file

This project is free to use, modify, and distribute for commercial and personal projects.

Copyright (c) 2024 Rahul Prajapati

Permission is hereby granted, free of charge, to any person obtaining a copy
of this software and associated documentation files (the "Software")...

Rahul Prajapati

• GitHub: @Rahul6128
• Repository: sentiment-analysis-rnn
• Location: Jaipur, India

Contributions welcome! Please:

1. Open an issue for bugs or suggestions
2. Fork the repository
3. Create a feature branch
4. Submit a pull request

• Report Bugs: GitHub Issues
• Questions: Open an issue with [question] tag
• Suggestions: Share ideas in GitHub Discussions

• TensorFlow & Keras teams
• IMDB dataset contributors
• Streamlit for web framework
• Open-source community