Personalized Machine Translation: Preserving Original Author Traits
This addresses the need for personalized machine translation to better reflect original author characteristics, though it is incremental as it focuses on a single trait.
The paper tackled the problem of preserving author gender traits in machine translation, showing that these traits are obfuscated in translations, and proposed domain-adaptation techniques that retain gender traits without harming translation quality.
The language that we produce reflects our personality, and various personal and demographic characteristics can be detected in natural language texts. We focus on one particular personal trait of the author, gender, and study how it is manifested in original texts and in translations. We show that author's gender has a powerful, clear signal in originals texts, but this signal is obfuscated in human and machine translation. We then propose simple domain-adaptation techniques that help retain the original gender traits in the translation, without harming the quality of the translation, thereby creating more personalized machine translation systems.