Beyond Parity: Fairness Objectives for Collaborative Filtering

By Sirui Yao, Bert Huang. 2017

414 citations recorded by the source corpus.

We study fairness in collaborative-filtering recommender systems, which are sensitive to discrimination that exists in historical data. Biased data can lead collaborative-filtering methods to make unfair predictions for users from minority groups. We identify the insufficiency of existing fairness metrics and propose four new metrics that address different forms of unfairness. These fairness metrics can be optimized by adding fairness terms to the learning objective. Experiments on synthetic and real data show that our new metrics can better measure fairness than the baseline, and that the fairness objectives effectively help reduce unfairness.

Subfields: Artificial Intelligence, Computer Science Applications, Safety Research, Safety, Robustness & Interpretability, Recommender Systems

How researchers see this paper

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

In Artificial Intelligence

In Computer Science Applications

In Safety Research

In Safety, Robustness & Interpretability

In Recommender Systems

Citations in the corpus

Does Gender Matter? Towards Fairness in Dialogue Systems (2019, background): “However, they have been proven to be unfair in many real-world applications such as computer vision [1], audio processing [2] and recommendations [3].”
Measuring Bias in Contextualized Word Representations (2019, background): “…2018) and language identiﬁcation (Jurgens et al., 2017), to downstream systems such as automated essay scoring (Amorim et al., 2018), exhibit inherent social biases which are attributed to the datasets used to train the embeddings (Barocas and Selbst, 2016; Zhao et al., 2017; Yao and Huang, 2017).”
Learning Gender-Neutral Word Embeddings (2018, background): “Researchers attributed such problem partly to the biases in the datasets (Zhao et al., 2017; Yao and Huang, 2017) and word embeddings (Garg et al.”
Empirical Risk Minimization under Fairness Constraints (2018, background): “In recent years there has been a lot of interest on algorithmic fairness in machine learning see, e.g., [1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13] and references therein.”
Mitigating Gender Bias for Neural Dialogue Generation with Adversarial Learning (2020, background): “A variety of AI applications have demonstrated common prejudices towards particular groups of people (Rodger and Pendharkar, 2004; Howard and Borenstein, 2018; Rose, 2010; Yao and Huang, 2017; Tolan et al., 2019).”
Balanced Datasets Are Not Enough: Estimating and Mitigating Gender Bias in Deep Image Representations (2018, background): “Concerns have been raised for applications such as recommender systems [35], credit score prediction [9], online news [24], and others [11] and in response various approaches have been proposed to mitigate bias [1, 10].”
Calibrated recommendations (2018, background): “Apart from that, [29] focused on rating prediction and RMSE, instead of the more relevant scenario of implicit feedback data and ranking metrics.”
Fairness in Machine Learning: A Survey (2020, background): “Subsequent publications expanded this view of recommender fairness by proposing additional metrics as well as algorithms [281, 294, 254, 29, 58].”
Fairness in Machine Learning (2020, methodology): “Some methods transform the constrained optimization problem via the method of Lagrange multipliers [3, 15, 34, 37, 38, 59, 97, 135, 183, 185] or add penalties to the objective [5, 14, 22, 44, 46, 56, 58, 62, 73, 74, 75, 79, 82, 87, 88, 89, 93, 94, 95, 98, 104, 115, 119, 123, 133, 134, 135, 138, 139,”
A Survey on the Fairness of Recommender Systems (2022, methodology): “Divided by Subject User [9, 19, 28, 39, 45, 53–55, 73, 76, 82, 86, 91, 95, 96] [11, 12, 22, 25, 43, 52, 56, 75, 77, 94, 99, 104, 114]”
Recommender systems and their ethical challenges (2020, background): “On the one hand, Yao and Huang (2017) consider several possible sources for unfairness in collaborative filtering, and introduce four new metrics to address them by measuring the distance between recommendations made by the system to different groups of users.”
Multistakeholder recommendation: Survey and research directions (2020, methodology): “The error independence criterion was approached through regularization in Yao and Huang (2017a).”
Quantifying Social Biases in Contextual Word Representations (2019, background): “, 2018), exhibit inherent social biases which are attributed to the datasets used to train the embeddings (Barocas and Selbst, 2016; Zhao et al., 2017; Yao and Huang, 2017).”
User-oriented Fairness in Recommendation (2021, background): “Yao and Huang [41] study fairness in collaborative filtering recommender systems, and propose four new metrics that address different forms of unfairness.”
Feedback Loop and Bias Amplification in Recommender Systems (2020, background): “Recommendationsgeneratedby CFgenerallysufferfrombiasagainst certain groups of users or items [13, 14].”
Bias and Debias in Recommender System: A Survey and Future Directions (2020, methodology): “[127], [128] proposed four fairness metrics in collaborative filtering, and used similar regularization-based optimization method to mitigate different forms of bias.”
Are My Deep Learning Systems Fair? An Empirical Study of Fixed-Seed Training (2021, background): “Bias metrics: Many metrics [6, 15, 19, 24, 26, 31, 35, 44, 46, 67, 82, 102, 103] have been proposed to measure biases of ML systems.”