Can You Translate that into Man? Commercial Machine Translation Systems Include Stylistic Biases

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

Language (Technology) is Power: A Critical Survey of “Bias” in NLP (2020, neutral): “…Sap et al. (2019); Sheng et al. (2019); Sun et al. (2019); Sweeney and Najafian (2019); Vaidya et al. (2019); Gaut et al. (2020); Gencoglu (2020); Hovy et al. (2020); Hutchinson et al. (2020); Kim et al. (2020); Peng et al. (2020); Rios (2020); Sap et al. (2020); Shah et al. (2020); Sheng et…”
Multi-Dimensional Gender Bias Classification (2020, background): “are judged as having been produced by older and more male speakers than the original was (Hovy et al., 2020).”
Disembodied Machine Learning: On the Illusion of Objectivity in NLP (2021, background): “cluding stylistic (Hovy et al., 2020) and gender bias (Vanmassenhove et al.”
Theories of “Gender” in NLP Bias Research (2022, background): “002 [61] Dirk Hovy, Federico Bianchi, and Tommaso Fornaciari.”
Conversational Assistants and Gender Stereotypes: Public Perceptions and Desiderata for Voice Personas (2020, background): “For example, in the area of Natural Language Processing (NLP), it has been shown that word embeddings (Bolukbasi et al., 2016), co-reference resolution (Zhao et al., 2018) and machine translation systems (Hovy et al., 2020) are likely to reflect and even amplify social biases in the data.”