=========================
Citation and bibliography
=========================

Citation
--------

If you use FairLangProc in your research, please cite our paper, **FairLangProc: A Python package for fairness in NLP**.

BibTeX
~~~~~~

.. code-block:: bibtex

   @misc{pérezperalta2025fairlangprocpythonpackagefairness,
         title={FairLangProc: A Python package for fairness in NLP},
         author={Arturo Pérez-Peralta and Sandra Benítez-Peña and Rosa E. Lillo},
         year={2025},
         eprint={2508.03677},
         archivePrefix={arXiv},
         primaryClass={cs.CL},
         url={https://arxiv.org/abs/2508.03677},
   }

APA Style
~~~~~~~~~

   Pérez-Peralta, A., Benítez-Peña, S., & Lillo, R. E. (2025). FairLangProc: A Python package for fairness in NLP. arXiv preprint arXiv:2508.03677.

IEEE Style
~~~~~~~~~~

   A. Pérez-Peralta, S. Benítez-Peña, and R. E. Lillo, "FairLangProc: A Python package for fairness in NLP," arXiv preprint arXiv:2508.03677, 2025.

MLA Style
~~~~~~~~~

   Pérez-Peralta, Arturo, et al. "FairLangProc: A Python package for fairness in NLP." arXiv preprint arXiv:2508.03677 (2025).

Related Publications
--------------------

If you use specific datasets, metrics, or algorithms from FairLangProc, please also cite the original papers:

Datasets
~~~~~~~~

- **BBQ** (Bias Benchmark for Question Answering): Parrish, A., et al. (2021). BBQ: A hand-built bias benchmark for question answering. arXiv:2110.08193.
- **BEC-Pro** (Bias Evaluation Corpus for Professions): Bartl, M., et al. (2020). Uncovering Hidden Bias in NLP. arXiv:2010.14534.
- **BOLD** (Bias in Open-Ended Language Generation): Dhamala, J., et al. (2021). BOLD: A large-scale bias benchmark for language model evaluation. ACM FAccT.
- **BUG** (Bias evaluation Universal Benchmark): Levy, S., et al. (2021). A Unified Approach to Benchmarking and Mitigating Bias. arXiv:2109.03858.
- **CrowS-Pairs**: Nangia, N., et al. (2020). CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Models. EMNLP.
- **GAP** (Gender Ambiguous Pronouns): Webster, K., et al. (2018). Mind the GAP: A systematic approach to measuring gender bias in coreference resolution. ACL.
- **HolisticBias**: Smith, E. M., et al. (2022). HolisticBias: Sensitivity analysis across sociodemographic identity. arXiv:2205.09209.
- **HONEST** (HOlingistic Neutrality Evalutation for gender-based Stereotypes): Nozza, D., et al. (2021). HONEST: Measuring Hurtful Language in Gendered Audio. NAACL.
- **StereoSet**: Nadeem, M., et al. (2020). StereoSet: Measuring stereotypical bias in language models. arXiv:2004.09456.
- **UnQover**: Li, T., et al. (2020). UnQovering Unintended Bias in Language Models through Compositional Reasoning. arXiv:2010.02428.
- **WinoBias+**: Vanmassenhove, E., et al. (2021). Toward Controllable Gender Bias Mitigation in Neural Machine Translation. arXiv:2109.06105.
- **WinoBias**: Zhao, J., et al. (2018). Gender Bias in Coreference Resolution. ACL.
- **WinoGender**: Rudinger, R., et al. (2018). Gender Bias in Neural Coreference Resolution. ACL.

Metrics
~~~~~~~

- **WEAT** (Word Embedding Association Test): Caliskan, A., Bryson, J. J., & Narayanan, A. (2017). Semantics derived automatically from language corpora contain human-like biases. Science.
- **LPBS** (Log Probability Bias Score): Kurita, K., et al. (2019). Measuring bias in contextualized word representations. ACL.
- **CBS** (Categorical Bias Score): Ahn, H., et al. (2021). Why Do Pretrained Language Models Cross-persona Conversational Tasks? EMNLP.
- **CPS** (CrowS-Pairs Score): Nangia, N., et al. (2020). CrowS-Pairs: A Challenge Dataset for Measuring Social Biases in Masked Language Models. EMNLP.
- **AUL** (All Unmasked Log-likelihood): Kaneko, M., et al. (2021). Gender and Racial Bias in Language Models. arXiv:2104.07496.
- **DR** (Demographic Representation): Liang, P. P., et al. (2022). Towards Fairness in Text Generation. arXiv:2211.09110.
- **SA** (Stereotypical Association): Liang, P. P., et al. (2022). Towards Fairness in Text Generation. arXiv:2211.09110.
- **HONEST** (HOlingistic Neutrality Evalutation for gender-based Stereotypes): Nozza, D., Bianchi, F., & Hovy, D. (2021). HONEST: Measuring Hurtful Language in Gendered Audio. NAACL.

Algorithms
~~~~~~~~~~

Pre-processing:

- **CDA** (Counterfactual Data Augmentation): Webster, K., et al. (2020). Measuring and Reducing the Correlation Between Word Embeddings and Gendered Constructs. ACL.
- **Projection-based Debiasing**: Bolukbasi, M., et al. (2016). Man is to Computer Programmer as Woman is to Homemaker? Debiasing Word Embeddings. NeurIPS.
- **BLIND** (Bias removaL wIth No Demographics): Orgad, H., et al. (2023). BLIND: Bias Removal via Debiasing at Inference Time. ACL.

In-processing:

- **ADELE** (Adapter-based DEbiasing of LanguagE models): Lauscher, A., et al. (2021). AdapterFusion: Non-invasive Transfer of Knowledge for Text Classification. arXiv:2109.03646.
- **EAR** (Entropy Attention Regularizer): Attanasio, G., et al. (2022). SENTeX: BERT-based Sentiment Analysis for Italian. arXiv:2203.09192.
- **Embedding-based Regularizer**: Liu, Y., et al. (2020). Fairness without Demographics through Adversarial Learning. arXiv:1910.10486.
- **Selective Unfreezing**: Gira, M., et al. (2022). Selective Knowledge Transfer for Democratizing NLP. ACL.

Intra-processing:

- **MoDDiffy** (Modular Debiasing with Diff Subnetworks): Hauzenberger, L., et al. (2023). Modular Debiasing for Language Models. ACL Findings.
- **EAT** (Entropy Attention Temperature scaling): Zayed, A., et al. (2023). DeBiasing: Learning about Bias Mitigation from Plain Models. arXiv:2305.13088.