Sentiment Analysis of Movie Reviews

This project performs sentiment analysis on IMDB movie reviews using both traditional machine learning (Logistic Regression) and deep learning (LSTM) approaches.

Project Structure

• data_preprocessing.py: Preprocesses the raw IMDB dataset
• data_analysis1.py: Implements Logistic Regression model
• data_analysis2.py: Implements LSTM model
• data_visualization.py: Creates visualizations of the results
• app.py: Flask application for real-time sentiment prediction
• model1.pickle: Saved Logistic Regression model
• model2.pth: Saved LSTM model
• vectorizer.pickle: Saved TF-IDF vectorizer
• vocab.npy: Vocabulary for the LSTM model

Setup and Installation

1. Clone this repository
2. Install required packages: pip install -r requirements.txt
3. Download the IMDB dataset from Kaggle and rename it to "imdb_dataset.csv"
4. Place the dataset in the project root directory

Usage

1. Run data preprocessing:python data_preprocessing.py
2. Train and evaluate models: python data_analysis1.py python data_analysis2.py
3. Visualize results: python data_visualization.py
4. Run the Flask app: python app.py

Contributors

• Shilong Luo, Yuxuan Liu

License

This project is licensed under the MIT License - see the LICENSE.md file for details.