A Comparative Study of Productivity and Quality Gain Between Post-Editing and Translating From Scratch

Abstract

Using machine translation (MT) input represents a fundamental change in translators’ work mode. The issue of efficacy of MT uses is worth investigating since it is at the heart of understanding translators’ choices in post-editing MT results or translating from scratch. This study focuses on a comparative study of the impact of post-editing MT on productivity and translation quality of student translator subjects with different levels of translation experiences. This study also looks into the influence of translators’ translation experiences on their performances. The keylogging experiment results show that MT input contributes positively to productivity gain and time savings with some variations caused by translation experiences, and that the overall final text quality is significantly affected when translating with or without MT input though to a varying degree of quality gain. These findings suggest a positive role of post-editing MT in translator training.