Bayesian Emotions: Sentiment Analysis

Simple implementation of a Naive Bayes classifier for sentiment analysis on movie reviews built from scratch.

Installation

1. Clone the repository:

   git clone https://github.com/Pacatro/Bayesian-Emotions.git
   cd Bayesian-Emotions

2. This project uses uv for package management. You can run the following command to setup the entire project:

   uv sync

3. Run the project:

   uv run src/main.py --help

Usage

The application provides a CLI with two main commands: test and eval.

Usage: main.py [OPTIONS] COMMAND [ARGS]...

Options:
  -v, --verbose          Verbose mode
  -s, --show-data-stats  Show data stats
  -h, --help             Show this message and exit.

Commands:
  test  Tests the model with new reviews, either predefined or from user...
  eval  Performs a final evaluation using the optimal minimum word length.

You can use global options like --verbose (-v) to see more detailed output and --show-data-stats (-s) to display dataset statistics upon loading.

Test the model

The test command trains a model and predicts the sentiment of reviews.

Usage: main.py test [OPTIONS]

  Tests the model with new reviews, either predefined or from user input.

Options:
  -i, --interactive  Test model in interactive mode (Using user input)
  -h, --help         Show this message and exit.

• Test with predefined sample reviews:

  uv run src/main.py test

• Test with your own review in interactive mode:

  uv run src/main.py test --interactive

  You will be prompted to enter a review.

Evaluate the model

The eval command performs a more thorough evaluation. It first finds the optimal minimum word length via cross-validation on the training set, and then reports detailed metrics on the test set.

Usage: main.py eval [OPTIONS]

  Performs a final evaluation using the optimal minimum word length.

Options:
  -k, --k INTEGER               Number of folds  [default: 5]
  -m, --min-ocurrences INTEGER  Minimum word occurences  [default: 1]
  -s, --smoothing-factor FLOAT  Smoothing factor  [default: 1.0]
  -h, --help                    Show this message and exit.

• Run the evaluation with default parameters:

  uv run src/main.py eval

• Customize evaluation parameters: You can customize the number of folds (-k), minimum word occurrences (-m), and the smoothing factor (-s).

  uv run src/main.py eval -k 10 -m 2 -s 1.5

License

MIT - Created by Paco Algar.