Putting Fairness Principles into Practice: Challenges, Metrics, and Improvements

By Alex Beutel, Jilin Chen, Tulsee Doshi, Hai Qian, Allison Woodruff, Christine Luu, Pierre Kreitmann, Jonathan Bischof, Ed H. Chi. 2019

176 citations recorded by the source corpus.

As more researchers have become aware of and passionate about algorithmic fairness, there has been an explosion in papers laying out new metrics, suggesting algorithms to address issues, and calling attention to issues in existing applications of machine learning. This research has greatly expanded our understanding of the concerns and challenges in deploying machine learning, but there has been much less work in seeing how the rubber meets the road. In this paper we provide a case-study on the application of fairness in machine learning research to a production classification system, and offer new insights in how to measure and address algorithmic fairness issues. We discuss open questions in implementing equality of opportunity and describe our fairness metric, conditional equality, that takes into account distributional differences. Further, we provide a new approach to improve on the fairness metric during model training and demonstrate its efficacy in improving performance for a real-world product.

Subfields: Safety Research, AI Ethics, Fairness & Society

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Safety Research

1. Nithum Thain
2. Anthony Rios
3. Dana Pessach
4. Erez Shmueli
5. Alex Beutel
6. Ed H. Chi
7. Jilin Chen
8. Flavien Prost

In AI Ethics, Fairness & Society

1. Dana Pessach
2. Erez Shmueli
3. Anthony Rios

Citations in the corpus

• A Step Toward More Inclusive People Annotations for Fairness (2021, background): “Finally, many bias mitigation strategies require attribute labels on at least a portion of the model’s training set [3, 19].”
• Algorithmic Fairness (2020, background): “[16] have raised some concerns about fair adversarial learning that should be further investigated.”
• Toward a better trade-off between performance and fairness with kernel-based distribution matching (2019, methodology): “Numerous regularization approaches have been proposed to encourage models during training to optimize for various fairness definitions [20, 4, 18, 1, 21, 11].”
• FuzzE: Fuzzy Fairness Evaluation of Offensive Language Classifiers on African-American English (2020, background): “ly/2GFZmEC Many metrics and strategies have been proposed to evaluate fairness in recent years (Zliobaite, 2015; Hardt et al., 2016; Dixon et al., 2018; Borkan et al., 2019; Beutel et al., 2019; Mitchell et al., 2019).”
• Fairness without Demographics through Adversarially Reweighted Learning (2020, methodology): “To ensure fair comparison we instantiate Min-Diff with similar neural architecture and model capacity as ARL.”
• Debiasing Embeddings for Reduced Gender Bias in Text Classification (2019, background): “There are a number of metrics and criteria that offer different interpretations of model fairness (Dixon et al., 2018; Narayanan, 2018; Friedler et al., 2019; Beutel et al., 2019).”
• A Review on Fairness in Machine Learning (2022, background): “[22] have raised some concerns about fair adversarial learning that should be further investigated.”