demographicx: A Python package for estimating gender and ethnicity using deep learning transformers

Summary

This package was used in the following publication to estimate gender and ethnicity of scientists

Acuna, Daniel E., Liang, Lizhen (2021), Are AI ethics conferences different and more diverse compared to traditional computer science conferences? In Proceedings of the AAAI/ACM Conference on AI, Ethics, and Society https://doi.org/10.1145/3461702.3462616

Installation

$ pip install git+https://github.com/sciosci/demographicx

Example

from demographicx import GenderEstimator

gender_estimator = GenderEstimator()
gender_estimator.predict('Daniel')

{'male': 0.9886190672823015,
 'unknown': 0.011367974526753396,
 'female': 1.2958190945360288e-05}

from demographicx import EthnicityEstimator
ethnicity_estimator = EthnicityEstimator()
ethnicity_estimator.predict('lizhen liang')

{'black': 7.76920258508755e-05,
 'hispanic': 0.00034410537213250747,
 'white': 0.0008992292872395202,
 'asian': 0.9986789733147773}

ethnicity_estimator.predict('daniel acuña')

{'black': 0.06771294974368015,
 'hispanic': 0.49868134219755395,
 'white': 0.1521847780511786,
 'asian': 0.2814209300075871}

Dependencies

• torch
• transformers
• numpy
• scipy

License

See LICENSE

Citation

Liang, L., Acuna, DE., demographicx: A Python package for estimating gender and ethnicity using deep learning transformers. (2021).