Assessing the Reliability of Word Embedding Gender Bias Measures

Du, Yupei; Fang, Qixiang; Nguyen, Dong

doi:10.18653/v1/2021.emnlp-main.785

Cited by 12 publications

(15 citation statements)

References 32 publications

(48 reference statements)

Supporting

Mentioning

Contrasting

Order By: Relevance

“…This task contains 276 template sentences t ∈ T , where for each occupation o that sentence either starts with that occupation, "man", or "woman", resulting in a triplet 3 We leave out the word pair ('guy', 'gal'), as we have noticed better results without the word pair. Ethayarajh et al (2019) and Du et al (2021) warn that including low-frequency words can negatively impact the bias measure, which we suspect is the case here.…”

Section: From Gender Representation To Gender Biasmentioning

confidence: 67%

The Birth of Bias: A case study on the evolution of gender bias in an English language model

Wal¹,

Jumelet²,

Schulz³

et al. 2022

Proceedings of the 4th Workshop on Gender Bias in Natural Language Processing (GeBNLP)

View full text Add to dashboard Cite

Section: From Gender Representation To Gender Biasmentioning

confidence: 67%

The Birth of Bias: A case study on the evolution of gender bias in an English language model

Wal¹,

Jumelet²,

Schulz³

et al. 2022

Proceedings of the 4th Workshop on Gender Bias in Natural Language Processing (GeBNLP)

View full text Add to dashboard Cite

“…Following the advise from [21], and to assess the quality of the gender direction obtained, we further perform PCA starting from an extended list of 50 pairs of gender words, taken from [7], and compare the result with � ⃗ g . From the full list of pairs available on the author's repository, 1 we select only those consisting of words present in GloVe.…”

Section: Gender Directionmentioning

confidence: 99%

Improving Gender-Related Fairness in Sentence Encoders: A Semantics-Based Approach

2023

View full text Add to dashboard Cite

The ever-increasing number of systems based on semantic text analysis is making natural language understanding a fundamental task: embedding-based language models are used for a variety of applications, such as resume parsing or improving web search results. At the same time, despite their popularity and widespread use, concern is rapidly growing due to their display of social bias and lack of transparency. In particular, they exhibit a large amount of gender bias, favouring the consolidation of social stereotypes. Recently, sentence embeddings have been introduced as a novel and powerful technique to represent entire sentences as vectors. We propose a new metric to estimate gender bias in sentence embeddings, named bias score. Our solution leverages semantic importance of words and previous research on bias in word embeddings, and it is able to discern between neutral and biased gender information at sentence level. Experiments on a real-world dataset demonstrate that our novel metric can identify gender stereotyped sentences. Furthermore, we employ bias score to detect and then remove or compensate for the more stereotyped entries in text corpora used to train sentence encoders, improving their degree of fairness. Finally, we prove that models retrained on fairer corpora are less prone to make stereotypical associations compared to their original counterpart, while preserving accuracy in natural language understanding tasks. Additionally, we compare our experiments with traditional methods for reducing bias in embedding-based language models.

show abstract

“…For instance, Jacobs and Wallach (2021) argue for applying psychometrics to study algorithmic fairness -a discussion we now extend to NLP bias measures. In section 6 we will consequently position our paper in the literature and compare our contributions to those of related works (Bommasani & Liang, 2022;Du et al, 2021;Jacobs & Wallach, 2021, i.a. ).…”

Section: Introductionmentioning

confidence: 99%

Undesirable Biases in NLP: Addressing Challenges of Measurement

Van der Wal,

Bachmann,

Leidinger

et al. 2024

jair

View full text Add to dashboard Cite

As Large Language Models and Natural Language Processing (NLP) technology rapidly develop and spread into daily life, it becomes crucial to anticipate how their use could harm people. One problem that has received a lot of attention in recent years is that this technology has displayed harmful biases, from generating derogatory stereotypes to producing disparate outcomes for different social groups. Although a lot of effort has been invested in assessing and mitigating these biases, our methods of measuring the biases of NLP models have serious problems and it is often unclear what they actually measure. In this paper, we provide an interdisciplinary approach to discussing the issue of NLP model bias by adopting the lens of psychometrics — a field specialized in the measurement of concepts like bias that are not directly observable. In particular, we will explore two central notions from psychometrics, the construct validity and the reliability of measurement tools, and discuss how they can be applied in the context of measuring model bias. Our goal is to provide NLP practitioners with methodological tools for designing better bias measures, and to inspire them more generally to explore tools from psychometrics when working on bias measurement tools. This article appears in the AI & Society track.

show abstract

Assessing the Reliability of Word Embedding Gender Bias Measures

Cited by 12 publications

References 32 publications

The Birth of Bias: A case study on the evolution of gender bias in an English language model

The Birth of Bias: A case study on the evolution of gender bias in an English language model

Improving Gender-Related Fairness in Sentence Encoders: A Semantics-Based Approach

Undesirable Biases in NLP: Addressing Challenges of Measurement

Contact Info

Product

Resources

About