mlsanityData to 13 Oct 2026

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

arXiv 2105.13782 ↗First posted 28 May 2021cs.CLcited by 11

Not on the map. Why?

Abstract (arXiv)

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 sub-word 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.

Key facts. Cited by 11 papers in the map (data through 13 Oct 2026). Builds on 49 of its 99 references that are in the map. First posted on arXiv 28 May 2021.

Builds on

49 of 99 references.

References · 49

  1. 2017Attention Is All You Needcited by 60,106
  2. 2015Rethinking the Inception Architecture for Computer Visioncited by 5,457
  3. 2014Neural Machine Translation by Jointly Learning to Align and Translatecited by 4,539
  4. 2014Sequence to Sequence Learning with Neural Networkscited by 2,981
  5. 2015Neural Machine Translation of Rare Words with Subword Unitscited by 2,910
  6. 2018SentencePiece: A simple and language independent subword tokenizer and detokenizer for Neural Text Processingcited by 1,979
  7. 2018A Call for Clarity in Reporting BLEU Scorescited by 1,793
  8. 2019fairseq: A Fast, Extensible Toolkit for Sequence Modelingcited by 1,487
  9. 2019SpecAugment: A Simple Data Augmentation Method for Automatic Speech Recognitioncited by 1,359
  10. 2020Language (Technology) is Power: A Critical Survey of "Bias" in NLPcited by 951
  11. 2015Sequence Level Training with Recurrent Neural Networkscited by 573
  12. 2016Google's Multilingual Neural Machine Translation System: Enabling Zero-Shot Translationcited by 467
37 more
  1. 2019Mitigating Gender Bias in Natural Language Processing: Literature Reviewcited by 287
  2. 2019Evaluating Gender Bias in Machine Translationcited by 247
  3. 2019Counterfactual Data Augmentation for Mitigating Gender Stereotypes in Languages with Rich Morphologycited by 200
  4. 2019Predictive Biases in Natural Language Processing Models: A Conceptual Framework and Overviewcited by 137
  5. 2017Sequence-to-Sequence Models Can Directly Translate Foreign Speechcited by 121
  6. 2018Assessing Gender Bias in Machine Translation -- A Case Study with Google Translatecited by 121
  7. 2021Gender Bias in Machine Translationcited by 116
  8. 2019Getting Gender Right in Neural Machine Translationcited by 98
  9. 2020Reducing Gender Bias in Neural Machine Translation as a Domain Adaptation Problemcited by 86
  10. 2019Toward Gender-Inclusive Coreference Resolutioncited by 84
  11. 2018Modeling Language Variation and Universals: A Survey on Typological Linguistics for Natural Language Processingcited by 63
  12. 2019On Measuring Gender Bias in Translation of Gender-neutral Pronounscited by 59
  13. 2020Gender in Danger? Evaluating Speech Translation Technology on the MuST-SHE Corpuscited by 57
  14. 2018Tied Multitask Learning for Neural Speech Translationcited by 54
  15. 2019Lost in Translation: Loss and Decay of Linguistic Richness in Machine Translationcited by 45
  16. 2020Mitigating Gender Bias in Machine Translation with Target Gender Annotationscited by 42
  17. 2020Neural Machine Translation Doesn't Translate Gender Coreference Right Unless You Make Itcited by 40
  18. 2016How Grammatical is Character-level Neural Machine Translation? Assessing MT Quality with Contrastive Translation Pairscited by 39
  19. 2020Dynamic Programming Encoding for Subword Segmentation in Neural Machine Translationcited by 38
  20. 2019On the Linguistic Representational Power of Neural Machine Translation Modelscited by 34
  21. 2016Character-based Neural Machine Translationcited by 33
  22. 2020Automatically Identifying Gender Issues in Machine Translation using Perturbationscited by 28
  23. 2019Filling Gender & Number Gaps in Neural Machine Translation with Black-box Context Injectioncited by 27
  24. 2017Linguistically Motivated Vocabulary Reduction for Neural Machine Translation from Turkish to Englishcited by 23
  25. 2019Gender Representation in French Broadcast Corpora and Its Impact on ASR Performancecited by 23
  26. 2016A Character-Level Decoder without Explicit Segmentation for Neural Machine Translationcited by 21
  27. 2019On Using SpecAugment for End-to-End Speech Translationcited by 19
  28. 2020Evaluating Gender Bias in Speech Translationcited by 18
  29. 2020Breeding Gender-aware Direct Speech Translation Systemscited by 16
  30. 2020Decoding and Diversity in Machine Translationcited by 15
  31. 2018XNMT: The eXtensible Neural Machine Translation Toolkitcited by 14
  32. 2019From Bilingual to Multilingual Neural Machine Translation by Incremental Trainingcited by 14
  33. 2020Consistent Transcription and Translation of Speechcited by 14
  34. 2020On Target Segmentation for Direct Speech Translationcited by 12
  35. 2020Gender Bias in Multilingual Neural Machine Translation: The Architecture Matterscited by 11
  36. 2019On the Importance of Word Boundaries in Character-level Neural Machine Translationcited by 5
  37. 2021How to Write a Bias Statement: Recommendations for Submissions to the Workshop on Gender Bias in NLPcited by 5

Built on by

Cited by year

Citing papers per year: 2021: 2; 2022: 2; 2023: 4; 2024: 3.

11 papers.

  1. 2021Between words and characters: A Brief History of Open-Vocabulary Modeling and Tokenization in NLPcited by 131
  2. 2022Under the Morphosyntactic Lens: A Multifaceted Evaluation of Gender Bias in Speech Translationcited by 23
  3. 2021Don't speak too fast: The impact of data bias on self-supervised speech modelscited by 20
  4. 2023Tokenization Matters: Navigating Data-Scarce Tokenization for Gender Inclusive Language Technologiescited by 16
  5. 2023Recent Advances in Direct Speech-to-text Translationcited by 15
  6. 2024What the Harm? Quantifying the Tangible Impact of Gender Bias in Machine Translation with a Human-centered Studycited by 15
  7. 2022Tokenization Consistency Matters for Generative Models on Extractive NLP Taskscited by 7
  8. 2023Exploring the Impact of Training Data Distribution and Subword Tokenization on Gender Bias in Machine Translationcited by 7
  9. 2024Enhancing Gender-Inclusive Machine Translation with Neomorphemes and Large Language Modelscited by 7
  10. 2023Test Suites Task: Evaluation of Gender Fairness in MT with MuST-SHE and INEScited by 6
  11. 2024Data Bias According to Bipol: Men are Naturally Right and It is the Role of Women to Follow Their Leadcited by 2