Options
Unveiling gender bias in AI translation: A corpus-assisted analysis of marriage-related text in the second sex
Author(s)
Date Issued
2024
Citation
Liu, J., & Yin, H. (17-19 May 2024). Unveiling gender bias in AI translation: A corpus-assisted analysis of marriage-related text in the second sex. The Japan Association for Language Teaching Computer Assisted Language Learning Conference 2024, Meijo University Nagoya Dome Campus, Japan.
Type
Conference Paper
Abstract
In the realm of artificial intelligence (AI) translation, concerns regarding gender bias have gained
prominence, warranting thorough investigation. This research endeavors to scrutinize the translations of marriage-related texts in AI’s English-Chinese translation through the lens of gender.
Drawing inspiration from Simone de Beauvoir’s seminal work The Second Sex, which delves into
the intricacies of gender dynamics, this study aims to compare its translations produced by AI with
those crafted by female and male translators. The original book, Le Deuxième Sexe, was first translated from French into English by Howard M. Parshley in 1953, which served as the sole source text
for Chinese translators in 20th century. The English version has been translated into Chinese by the
female translator Yang Meihui and published in 1973, and later by the male translator Tao Tiezhu
and published in 1998. Employing a corpus linguistics approach, this research will meticulously
extract and compare texts pertaining to marriage in ChatGPT-generated translations and male and
female translations of The Second Sex. By systematically comparing the translations, we aim to
uncover potential instances of gender bias inherent in AI translation systems. Through this comparative analysis, we seek to elucidate whether AI translations mirror or diverge from the gendered
nuances present in human-authored translations, particularly in the context of marriage-related discourse. This study not only contributes to the burgeoning discourse on gender bias in AI but also offers insights into the implications of automated translation systems on gender representation and
perception.
Availability at HKSYU Library
