MACHINE TRANSLATION FOR POST-EDITING PRACTICES

It has been proven in many quality-focused studies that machine translation output some language pairs is still far from publishable (Koponen, 2016). Even so, post-editing become a daily practice among translators and service providers, especially with where demonstrates good human parity. The fast development of its quality improvement have led to growing demand post-editors. This study attempts evaluate the most popular tools for Lithuanian order find correlation between results automatic estimation (i.e., BLUE score), / manual evaluation following multidimensional metrics (MQM) common engines used by freelancers providers.
 conclusions are based on findings survey vs analysis. demonstrate support previous research may not be taken granted. Human better indicator whether tool fits purpose translation. brings fore insightful beneficial translator post-editor trainers pedagogical perspective as well industry practical perspective.