Speed-to-Quality Ratio in Fully Human Translation vs. Post-Editing of Machine Translation Output
Ana Sabljak · Repozitorij Filozofskog fakulteta u Zagrebu' at University of Zagreb (University of Zagreb) · 2014
Bearing in mind an ever-growing amount of texts to be translated, often under very tight deadlines, there is growing pressure in the industry to use translation technology to speed up the translation process.More and more often, translators are called upon to post-edit machine translation (MT) output rather than translate a text from the very beginning.The primary purpose of this text is to examine whether post-editing of MT output indeed leads to greater speed and quality in translation.This study compares the speed and quality of fully human translations to those of post-edited MT output for three different text types: a novel, a news report, and a legal text.Quality is assessed along four parameters, namely the morphosyntactic features, the semantic features, style, and general impression.The MT system used for the purposes of the study is Google Translate.