Gender Bias in Multilingual Neural Machine Translation: The Architecture Matters
By Marta Ruiz Costa-jussà, Carlos Escolano, Christine Basta, Javier Ferrando, Roser Batlle Roca, Ksenia Kharitonova. 2020
20 citations recorded by the source corpus.
Multilingual Neural Machine Translation architectures mainly differ in the amount of sharing modules and parameters among languages. In this paper, and from an algorithmic perspective, we explore if the chosen architecture, when trained with the same data, influences the gender bias accuracy. Experiments in four language pairs show that Language-Specific encoders-decoders exhibit less bias than the Shared encoder-decoder architecture. Further interpretability analysis of source embeddings and the attention shows that, in the Language-Specific case, the embeddings encode more gender information, and its attention is more diverted. Both behaviors help in mitigating gender bias.
Subfields: Natural Language Processing, Safety, Robustness & Interpretability
How researchers see this paper
The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.
In Natural Language Processing
- Nanyun Peng
- Kai-Wei Chang
- Emily Sheng
- Premkumar Natarajan
- Yejin Choi
- Adina Williams
- Eric Michael Smith
- Percy Liang
In Safety, Robustness & Interpretability
- Emily Sheng
- Nanyun Peng
- Kai-Wei Chang
- Premkumar Natarajan
- Adina Williams
- Sunipa Dev
- Yejin Choi
- Jieyu Zhao
Citations in the corpus
- Societal Biases in Language Generation: Progress and Challenges (2021, background): “, 2021) MT (Vanmassenhove et al., 2018; Elaraby et al., 2018; Prates et al., 2019; Stanovsky et al., 2019; Escudé Font and Costa-jussà, 2019; Cho et al., 2019; Moryossef et al., 2019; Saunders and Byrne, 2020; Saunders et al., 2020; Kocmi et al., 2020; Costa-jussà and de Jorge, 2020; Costa-jussà et ”