🎬 Sentiments Analysis

📌 Overview

A Natural Language Processing (NLP) project that uses a Recurrent Neural Network (RNN) built with PyTorch to classify IMDB movie reviews as Positive or Negative.

📂 Dataset

IMDB Movie Reviews Dataset

• Binary sentiment classification
• 0 → Negative
• 1 → Positive

🧠 Model Architecture

Movie Review
     ↓
Text Preprocessing
     ↓
Tokenization & Padding
     ↓
Embedding
     ↓
RNN
     ↓
Fully Connected Layer
     ↓
Positive / Negative

🛠️ Technologies

• PyTorch
• Scikit-learn
• Python
• NumPy
• Pandas
• Matplotlib
• Seaborn

📊 Results

Metric	Score
Accuracy	88.99%
Precision	89.02%
Recall	88.99%
F1 Score	88.99%

📁 Folder Structure

Sentiment-Analysis-RNN/
│
├── Notebook/
│   └── RNN_Sentiments_Analysis.ipynb
│
├── models/
│   └── best_RNN_model.pt
│
├── Output/
│   ├── confusion_matrix.png
│   └── loss_curve.png
│
├── README.md

🎯 Key Learning

• Text prep-rocessing and tokenization
• Sequence padding
• Word embeddings
• RNN-based text classification
• Model training and validation
• Classification metrics and visualization

🔮 Future Improvements

• Compare RNN, LSTM, and GRU
• Experiment with pretrained embeddings
• Explore Transformer-based sentiment analysis