Selected publications below. Full list of publications available on Google Scholar.

Post-editing Effort of a Novel with Statistical and Neural Machine Translation. 2018. Frontiers in Digital Humanities.First publication in which professional literary translators worked with machine translation. It showed that regardless of their seniority or familiarity with machine translation, they translated faster when assisted by a literary-adapted machine translation system than when they translated on their own.
Post-editese: an Exacerbated Translationese. 2019. MT Summit.First comprehensive quantitative analysis of the characteristics of machine translations, machine-assisted translations, and human translations, demonstrating that the usage of machine translation produces translations that are lexically poorer and have more interference from the source language. Best paper award.
A Multifaceted Evaluation of Neural versus Phrase-based Machine Translation for 9 Language Directions. 2017. EACL.This article analyses the characteristics of translations produced by the previously mainstream statistical paradigm and the current neural paradigm in machine translation systems. Its publication coincided with the growing popularity of neural machine translation and showed how its translations were different from statistical machine translations, e.g., more fluent but weaker for long sentences.
Attaining the unattainable? Reassessing claims of human parity in neural machine translation. 2018. WMT.This article refutes the first claim of human parity in machine translation, showing methodological problems in the way that evaluation was conducted. When these were resolved and the evaluation was conducted in a fair way, it demonstrated that parity had not been achieved. The article proposed a set of recommendations for machine translation evaluation, which have been adopted by the research community.
The impact of post-editing and machine translation on creativity and reading experience. 2020. Translation Spaces.First work that measured the impact of machine translation on the creativity of the resulting translations and on their reading experience. It showed that the usage of machine translation leads to less creative translations than those produced by professional translators. However, the reading experience is not necessarily affected.