Multi-Task Learning for Improving Gender Accuracy in Neural Machine Translation
Carlos Escolano | Graciela Ojeda | Christine Basta | Marta R. Costa-jussa
Proceedings of the 18th International Conference on Natural Language Processing (ICON)
Machine Translation is highly impacted by social biases present in data sets, indicating that it reflects and amplifies stereotypes. In this work, we study mitigating gender bias by jointly learning the translation, the part-of-speech, and the gender of the target language with different morphological complexity. This approach has shown improvements up to 6.8 points in gender accuracy without significantly impacting the translation quality.