Abstract:

The increasing prominence of machine translation (MT) and postediting in both industrial language business and everyday activities of the general public highlights a significant opportunity for translators to enhance productivity and efficiency in their professional duties. In the light of the evolving landscape of MT technologies over the past decade, the present study explores the human performance in postediting machine-translated production. To gain a more comprehensive understanding of postediting procedures and efforts required to postedit machine-translated texts, the study involved an analysis of keyboard activity collected during two postediting tasks. The study participants were seven postgraduate students of Translation at a technological university; their levels of expertise in translation varied. The machine translation system DeepL was employed to translate two paragraphs extracted from the official Apple website from English into Lithuanian. Importantly, the paragraphs represented two distinct text types, i.e. a research article and a user manual. The evaluation of the quality of MT output and the extent of postediting modifications was conducted using Translog-II, an automated tool for assessing user data activity (Carl, 2012). Specifically, user activity data identifying the temporal and cognitive efforts involved was analysed. The findings suggest that postediting a research article entailed a greater temporal effort, whereas post-editing a user manual required a higher level of cognitive effort on the part of posteditors. While more research is needed, the results of the present study point towards the posteditors’ greater degree of perceived competence when translating instructional manuals when compared to research articles. Studying the tendencies may provide helpful pedagogical implications as to the development of the postediting competence in translator training.

References:
[1] Carl, Michael (2012). "Translog - II: a Program for Recording User Activity Data for Empirical Reading and Writing Research", In Proceedings of the Eighth International Conference on Language Resources and Evaluation, European Language Resources Association (ELRA).

Keywords:

Machine translation, postediting, translator training, DeepL, Translog-II.