Chapter 4. Post-editing neural machine translation in specialised languages

Abstract

This study focuses on phraseology in specialised texts and on students’ difficulties pertaining to phraseology in post-editing neural machine translation output. It is undertaken within the corpus-based methodological framework that we have developed for several purposes, one of which being to assess the impact of corpus use on translation and post-editing. The objective of the study is to propose a descriptive analysis of typical student errors related to phraseology in order to design tailored pedagogical materials. We aim to show that, with consistent training in querying corpora and in interpreting results in an appropriate manner, students can manage to improve their productions when translating specialised texts or when post-editing machine translation output.