Exploration Of Pre-editing Approaches For Machine Translation In Translating Abstracts Of Medical Papers

In today’s accelerated globalization,the market demand for translation has increased sharply.Machine translation,with its advantages of low cost and high speed,can greatly improve the efficiency of the work.However,due to the limitations of machine translation and the differences between Chinese and English,there still exist many problems in machine translation.In order to achieve a balance between translation efficiency and translation quality,the output often needs to be manually reviewed and revised a lot.Medical texts are characterized by their fixed language structures,particularity,and strong purpose,so the existing model of "Machine translation + Post-editing" is inefficient and time-consuming(Wang & Yun,2019).Therefore,the role of pre-editing stands out and the new translation model of "Pre-editing + Machine translation +Post-editing" is introduced.Pre-editing is a process whereby source text is revised before being input into translation software.Pre-editing only needs to revise one language,so it can greatly improve the machine’s recognition of the source text,enhance the readability of the output translation and alleviate the workload of post-editing,enabling the improvement both in translation quality and translation efficiency.There are evidences that the output quality of abstracts after pre-editing is improved,with less errors in translations.This thesis aims at the translation of medical abstracts,and,by analyzing the errors of medical abstracts in machine translation,finds out the pre-editing approaches of machine translation,in the hope that it would improve the output quality of machine translation for medical abstracts.