Logo
ScienceStack
Table of Contents
Fetching ...
Sign In

Current State-of-the-Art of Bias Detection and Mitigation in Machine Translation for African and European Languages: a Review

Catherine Ikae, Mascha Kurpicz-Briki

TL;DR

It is shown how the majority of the work in this field of bias detection and mitigation methods concentrates on few languages, and that there is potential for future research to cover also the less investigated languages to contribute to more diversity in the research field.

Abstract

Studying bias detection and mitigation methods in natural language processing and the particular case of machine translation is highly relevant, as societal stereotypes might be reflected or reinforced by these systems. In this paper, we analyze the state-of-the-art with a particular focus on European and African languages. We show how the majority of the work in this field concentrates on few languages, and that there is potential for future research to cover also the less investigated languages to contribute to more diversity in the research field.

Current State-of-the-Art of Bias Detection and Mitigation in Machine Translation for African and European Languages: a Review

TL;DR

It is shown how the majority of the work in this field of bias detection and mitigation methods concentrates on few languages, and that there is potential for future research to cover also the less investigated languages to contribute to more diversity in the research field.

Abstract

Studying bias detection and mitigation methods in natural language processing and the particular case of machine translation is highly relevant, as societal stereotypes might be reflected or reinforced by these systems. In this paper, we analyze the state-of-the-art with a particular focus on European and African languages. We show how the majority of the work in this field concentrates on few languages, and that there is potential for future research to cover also the less investigated languages to contribute to more diversity in the research field.

Paper Structure

This paper contains 16 sections, 2 figures, 1 table.

Table of Contents

  1. Introduction
  2. Methods
  3. Results
  4. Papers that focus on bias detection using pre-defined test suites, corpora, or experiments to measure gender bias in machine translations.
  5. Bias mitigation in machine translation, introducing innovative methods to reduce gender bias in translations.
  6. Addressing specific gender-related challenges such as pronoun translation, coreference resolution, and the handling of relationships in translations.
  7. Efforts to tackle exposure bias in neural machine translation (NMT)
  8. Data collection plays a critical role in bias mitigation, with some researchers developing gender-balanced corpora for training and evaluation, aiming to improve datasets for detecting and mitigating gender bias in machine translation.
  9. Various evaluation benchmarks and methods have also been developed to detect bias in machine translation.
  10. Pedagogical frameworks.
  11. Discussion
  12. Limitations of the State-of-the-Art
  13. Outlook for Future Work
  14. Limitations
  15. Queries and Lists of Languages
  16. ...and 1 more sections

Figures (2)

  • Figure 1: ACL Query African Languages.
  • Figure 2: ACL Query European Languages.