Fairness in representation: quantifying stereotyping as a representational harm

By Mohsen Abbasi, Sorelle A. Friedler, Carlos Eduardo Scheidegger, Suresh Venkatasubramanian. 2019

58 citations recorded by the source corpus.

While harms of allocation have been increasingly studied as part of the subfield of algorithmic fairness, harms of representation have received considerably less attention. In this paper, we formalize two notions of stereotyping and show how they manifest in later allocative harms within the machine learning pipeline. We also propose mitigation strategies and demonstrate their effectiveness on synthetic datasets.

Subfields: Artificial Intelligence, Safety Research, Natural Language Processing, Safety, Robustness & Interpretability

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 Safety Research

In Natural Language Processing

In Safety, Robustness & Interpretability

Citations in the corpus

On The Impact of Machine Learning Randomness on Group Fairness (2023, background): “For all (t, s) ∈ [1, 50], we perform s unique training runs (while changing both forms of randomness, i.”
The Values Encoded in Machine Learning Research (2021, methodology): “, LWPR [8], standard GPR [1], sparse online Gaussian process regression (OGP) [5] and υ-support vector regression (υ-SVR) [11], respectively.”
UNQOVERing Stereotypical Biases via Underspecified Questions (2020, background): “Hence, if unchecked, such representational harms in model predictions would percolate into allocational harms (cf. Crawford, 2017; Abbasi et al., 2019; Blodgett et al., 2020).”
Fairness for Unobserved Characteristics: Insights from Technological Impacts on Queer Communities (2021, background): “is possible to root the analysis in proxy groups [63], there is a risk of incorporating harmful stereotypes in proxy group definitions, potentially resulting in harms of representation [1].”
Neural Machine Translation Doesn’t Translate Gender Coreference Right Unless You Make It (2020, background): “System users also experience representational harms via the reinforcement of stereotypes associating occupations with a particular gender, as well as lower quality of service in receiving grammatically incorrect translations (Abbasi et al., 2019).”
OSCaR: Orthogonal Subspace Correction and Rectification of Biases in Word Embeddings (2020, background): “In an effort to prevent such transition of representational harms into allocational harms (cf. Crawford, 2017; Abbasi et al., 2019; Blodgett et al., 2020), we look at mitigating stereotyping biases at the source, i.”
Your fairness may vary: Pretrained language model fairness in toxic text classification (2021, background): “Abbasi et al. (2019) discuss how harms from stereotypical representations manifest as allocative harms later in the ML pipeline.”
Equity Beyond Bias in Language Technologies for Education (2019, background): “This may take the form of search results returning stereotypical images of minorities (Noble, 2018) or other algorithmic stereotyping (Abbasi et al., 2019); much of the work in word embeddings falls into this category (Caliskan et al., 2017), though research on downstream tasks and outcomes often…”
Overcoming Failures of Imagination in AI Infused System Development and Deployment (2020, background): “…a wide spectrum of negative overtones or impacts for human stakeholders, ranging from denigration and emotional distress to physical harm and loss of opportunities (Blodgett et al., 2020; Barocas et al., 2017; Kay et al., 2015; Abbasi et al., 2019; Baker and Potts, 2013; Otterbacher et al., 2017).”