GitHub - alexiamuresan/AI-Master-Thesis-Fairness-in-AI-Based-Hiring

This repository contains:

6 file with the inal versions of code used in the bias evaluations. Note: this is the code only for scenario 2 (train of biased data, test on balanced data). The rest of the results can be reproduced by simply changing the datsets used in the notebooks.
3 synthetically generated datasets: 1 balanced, 1 gender biased, 1 biased in terms of ethnicity.
Code used to compute the number of trainable parameters in the instances of traditional models used in this thesis. The number of parameters for the LLMs is considered common
knowledge/found online, not directly computed.
Code used to generate the synthetic datasets.

Name		Name	Last commit message	Last commit date
Latest commit History 15 Commits
.DS_Store		.DS_Store
README.md		README.md
balanced_resumes_final.csv		balanced_resumes_final.csv
bias_eval_BERT_ethnicity.ipynb		bias_eval_BERT_ethnicity.ipynb
bias_eval_BERT_gender.ipynb		bias_eval_BERT_gender.ipynb
bias_eval_GPT_ethnicity.ipynb		bias_eval_GPT_ethnicity.ipynb
bias_eval_GPT_gender.ipynb		bias_eval_GPT_gender.ipynb
bias_eval_trad_models_ethnicity.ipynb		bias_eval_trad_models_ethnicity.ipynb
bias_eval_trad_models_gender.ipynb		bias_eval_trad_models_gender.ipynb
dataset_generation.ipynb		dataset_generation.ipynb
ethnicity_biased_final.csv		ethnicity_biased_final.csv
gender_biased_final.csv		gender_biased_final.csv
parameter_calculation.ipynb		parameter_calculation.ipynb

Provide feedback