Gender Bias in Machine Translation

Savoldi, Beatrice; Gaido, Marco; Bentivogli, Luisa; Negri, Matteo; Turchi, Marco

doi:10.1162/tacl_a_00401

Cited by 75 publications

(56 citation statements)

References 159 publications

(166 reference statements)

Supporting

Mentioning

Contrasting

Unclassified

Order By: Relevance

“…Basta et al [2020] show that this method improves the performance of machine translation with coreference resolution tasks. However, Savoldi et al [2021] note that this improvement might not be due to the added gender context, but for instance, a regularisation effect. 8.1.4 Balanced Fine-Tuning.…”

Section: Debiasing Using Data Manipulationmentioning

confidence: 93%

“…Costa-jussà and de Jorge [2020] use an inverse approach and train their model on a larger corpus and fine-tune it with a gender-balanced corpus showing that their approach successfully mitigates gender bias and increases performance quality even if the balanced dataset is coming from a different domain. However, Savoldi et al [2021] note that this approach does not account for the qualitative differences in how men and women are portrayed [Savoldi et al 2021].…”

Section: Debiasing Using Data Manipulationmentioning

confidence: 99%

“…In particular, previous research has confirmed gender bias to be prevalent in literature [Hoyle et al 2019], news [Wevers 2019], media [Asr et al 2021], and communication about and directed towards people of different genders [Fast et al 2016;Voigt et al 2018]. Further, prior studies have shown bias in underlying NLP algorithms such as word embeddings [Bolukbasi et al 2016] and language models [Nadeem et al 2021], as well as in the downstream tasks they are employed for, e.g., machine translation [Savoldi et al 2021], coreference resolution [Rudinger et al 2018;Webster et al 2018;Zhao et al 2018a], language generation [Sheng et al 2020], and part-of-speech tagging and parsing [Garimella et al 2019].…”

Section: Introductionmentioning

confidence: 99%

See 2 more Smart Citations

A Survey on Gender Bias in Natural Language Processing

Stanczak¹,

Augenstein²

2021

Preprint

View full text Add to dashboard Cite

Language can be used as a means of reproducing and enforcing harmful stereotypes and biases and has been analysed as such in numerous research. In this paper, we present a survey of 304 papers on gender bias in natural language processing. We analyse definitions of gender and its categories within social sciences and connect them to formal definitions of gender bias in NLP research. We survey lexica and datasets applied in research on gender bias and then compare and contrast approaches to detecting and mitigating gender bias. We find that research on gender bias suffers from four core limitations. 1) Most research treats gender as a binary variable neglecting its fluidity and continuity. 2) Most of the work has been conducted in monolingual setups for English or other high-resource languages. 3) Despite a myriad of papers on gender bias in NLP methods, we find that most of the newly developed algorithms do not test their models for bias and disregard possible ethical considerations of their work. 4) Finally, methodologies developed in this line of research are fundamentally flawed covering very limited definitions of gender bias and lacking evaluation baselines and pipelines. We see overcoming these limitations as a necessary development in future research.

show abstract

Section: Debiasing Using Data Manipulationmentioning

confidence: 93%

Section: Debiasing Using Data Manipulationmentioning

confidence: 99%

Section: Introductionmentioning

confidence: 99%

See 1 more Smart Citation

A Survey on Gender Bias in Natural Language Processing

Stanczak¹,

Augenstein²

2021

Preprint

View full text Add to dashboard Cite

show abstract

“…Although data are not the only factor contributing to generate bias (Shah et al, 2020;Savoldi et al, 2021), only few inquiries devoted attention to other technical components that exacerbate the problem (Vanmassenhove et al, 2019) or to architectural changes that can contribute to its mitigation (Costa-jussà et al, 2020b). From an algorithmic perspective, Roberts et al (2020) additionally expose how "taken-for-granted" approaches may come with high overall translation quality in terms of BLEU scores, but are actually detrimental when it comes to gender bias.…”

Section: Introductionmentioning

confidence: 99%

How to Split: the Effect of Word Segmentation on Gender Bias in Speech Translation

Gaido¹,

Savoldi²,

Bentivogli³

et al. 2021

Preprint

Self Cite

View full text Add to dashboard Cite

Having recognized gender bias as a major issue affecting current translation technologies, researchers have primarily attempted to mitigate it by working on the data front. However, whether algorithmic aspects concur to exacerbate unwanted outputs remains so far under-investigated. In this work, we bring the analysis on gender bias in automatic translation onto a seemingly neutral yet critical component: word segmentation. Can segmenting methods influence the ability to translate gender? Do certain segmentation approaches penalize the representation of feminine linguistic markings? We address these questions by comparing 5 existing segmentation strategies on the target side of speech translation systems. Our results on two language pairs (English-Italian/French) show that state-of-the-art subword splitting (BPE) comes at the cost of higher gender bias. In light of this finding, we propose a combined approach that preserves BPE overall translation quality, while leveraging the higher ability of character-based segmentation to properly translate gender.

show abstract

“…For a detailed survey of gender bias in machine translation, we refer readers toSavoldi et al (2021).3 https://translate.google.com…”

mentioning

confidence: 99%

Societal Biases in Language Generation: Progress and Challenges

Sheng¹,

Chang²,

Natarajan

et al. 2021

Preprint

View full text Add to dashboard Cite

Technology for language generation has advanced rapidly, spurred by advancements in pre-training large models on massive amounts of data and the need for intelligent agents to communicate in a natural manner. While techniques can effectively generate fluent text, they can also produce undesirable societal biases that can have a disproportionately negative impact on marginalized populations. Language generation presents unique challenges for biases in terms of direct user interaction and the structure of decoding techniques. To better understand these challenges, we present a survey on societal biases in language generation, focusing on how data and techniques contribute to biases and progress towards reducing biases. Motivated by a lack of studies on biases from decoding techniques, we also conduct experiments to quantify the effects of these techniques. By further discussing general trends and open challenges, we call to attention promising directions for research and the importance of fairness and inclusivity considerations for language generation applications.

show abstract

Gender Bias in Machine Translation

Cited by 75 publications

References 159 publications

A Survey on Gender Bias in Natural Language Processing

A Survey on Gender Bias in Natural Language Processing

How to Split: the Effect of Word Segmentation on Gender Bias in Speech Translation

Societal Biases in Language Generation: Progress and Challenges

Contact Info

Product

Resources

About