Fast Stylometry Python library: Natural Language Processing tool

You can run the walkthrough notebook in Google Colab with a single click:

☄ Fast Stylometry - Burrows Delta NLP technique ☄

Developed by Fast Data Science. Fast Data Science develops products, offers consulting services, and training courses in natural language processing (NLP). Subscribe to our blog for regular news from the NLP universe.

Source code at https://github.com/fastdatascience/faststylometry

Tutorial at https://fastdatascience.com/fast-stylometry-python-library/

Fast Stylometry is a Python library for calculating the Burrows' Delta. Burrows' Delta is an algorithm for comparing the similarity of the writing styles of documents, known as forensic stylometry.

A useful explanation of the maths and thinking behind Burrows' Delta and how it works

💻 Installing the Fast Stylometry Python package

You can install from PyPI.

pip install faststylometry

🌟 Using Fast Stylometry NLP library for the first time 🌟

⚠️ We recommend you follow the walk through notebook titled Burrows Delta Walkthrough.ipynb in order to understand how the library works. If you don't have the correct environment set up on your machine, then you can run the walkthrough notebook easily using this link to create a notebook in Google Colab.

💡 Usage examples

Demonstration of Burrows' Delta on a small corpus downloaded from Project Gutenberg.

We will test the Burrows' Delta code on two "unknown" texts: Sense and Sensibility by Jane Austen, and Villette by Charlotte Bronte. Both authors are in our training corpus.

You can get the training corpus by cloning https://github.com/fastdatascience/faststylometry, the data is in faststylometry/data. Or you can call download_examples() from Python after importing Fast Stylometry:

from faststylometry import download_examples
download_examples()

📖 Create a corpus

The Burrows Delta Walkthrough.ipynb Jupyter notebook is the best place to start, but here are the basic commands to use the library:

To create a corpus and add books, the pattern is as follows:

from faststylometry import Corpus
corpus = Corpus()
corpus.add_book("Jane Austen", "Pride and Prejudice", [whole book text])

Here is the pattern for creating a corpus and adding books from a directory on your system. You can also use the method util.load_corpus_from_folder(folder, pattern).

import os
import re

from faststylometry.corpus import Corpus

corpus = Corpus()
for root, _, files in os.walk(folder):
    for filename in files:
        if filename.endswith(".txt") and "_" in filename:
            with open(os.path.join(root, filename), "r", encoding="utf-8") as f:
                text = f.read()
            author, book = re.split("_-_", re.sub(r'\.txt', '', filename))

            corpus.add_book(author, book, text)

💡 Example 1

Download some example data (Project Gutenberg texts) from the Fast Stylometry repository:

from faststylometry import download_examples
download_examples()

Load a corpus and calculate Burrows' Delta

from faststylometry.util import load_corpus_from_folder
from faststylometry.en import tokenise_remove_pronouns_en
from faststylometry.burrows_delta import calculate_burrows_delta

train_corpus = load_corpus_from_folder("faststylometry/data/train")

train_corpus.tokenise(tokenise_remove_pronouns_en)

test_corpus_sense_and_sensibility = load_corpus_from_folder("faststylometry/data/test", pattern="sense")

test_corpus_sense_and_sensibility.tokenise(tokenise_remove_pronouns_en)

calculate_burrows_delta(train_corpus, test_corpus_sense_and_sensibility)

returns a Pandas dataframe of Burrows' Delta scores

💡 Example 2

Using the probability calibration functionality, you can calculate the probability of two books being by the same author.

from faststylometry.probability import predict_proba, calibrate
calibrate(train_corpus)
predict_proba(train_corpus, test_corpus_sense_and_sensibility)

outputs a Pandas dataframe of probabilities.

✉️ Who to contact

Thomas Wood at Fast Data Science

🤝 Contributing to the project

If you'd like to contribute to this project, you can contact us at https://fastdatascience.com/ or make a pull request on our Github repository. You can also raise an issue.

Developing the library

Automated tests

Test code is in tests/ folder using unittest.

The testing tool tox is used in the automation with GitHub Actions CI/CD.

Use tox locally

Install tox and run it:

pip install tox
tox

In our configuration, tox runs a check of source distribution using check-manifest (which requires your repo to be git-initialized (git init) and added (git add .) at least), setuptools's check, and unit tests using pytest. You don't need to install check-manifest and pytest though, tox will install them in a separate environment.

The automated tests are run against several Python versions, but on your machine, you might be using only one version of Python, if that is Python 3.9, then run:

tox -e py39

Thanks to GitHub Actions' automated process, you don't need to generate distribution files locally. But if you insist, click to read the "Generate distribution files" section.

🤖 Continuous integration/deployment to PyPI

This package is based on the template https://pypi.org/project/example-pypi-package/

This package

uses GitHub Actions for both testing and publishing
is tested when pushing master or main branch, and is published when create a release
includes test files in the source distribution
uses setup.cfg for version single-sourcing (setuptools 46.4.0+)

🧍 Re-releasing the package manually

The code to re-release Fast Stylometry on PyPI is as follows:

source activate py311
pip install twine
rm -rf dist
python setup.py sdist
twine upload dist/*

😊 Who worked on the Fast Stylometry NLP library?

The tool was developed by:

Thomas Wood, Natural Language Processing consultant and data scientist at Fast Data Science.

📜 License of Fast Stylometry library

✍️ Citing the Fast Stylometry library

If you are undertaking research in AI, NLP, or other areas, and are publishing your findings, I would be grateful if you could please cite the project.

Wood, T.A., Fast Stylometry [Computer software], Version 1.0.4, accessed at https://fastdatascience.com/fast-stylometry-python-library, Fast Data Science Ltd (2023)

@unpublished{faststylometry,
    AUTHOR = {Wood, T.A.},
    TITLE  = {Fast Stylometry (Computer software), Version 1.0.4},
    YEAR   = {2023},
    Note   = {To appear},
}

jeffkaufman/faststylometry