Pre-editing Strategies In C-E Machine Translation

In recent years,the rapid development of machine translation,especially the development of Neural Machine Translation System,has greatly improved the quality of machine translation.But it is undeniable that machine translation is still not comparable to high-level human translation.Therefore,the combination of machine translation and human editing has gained more and more recognition.At present,there are many studies on post editing but insufficient attention to pre-editing.Based on the case of a translation program named The Record of Global Cities Forum,this thesis first summarizes the problems of the source text,and explores which pre-editing strategies can effectively improve the quality of the machine translation output and reduce the workload of post editing by comparing the machine translation of the source text before and after pre-editing.There are problems in the readability and translatability of the source text.The problems in readability include content redundancy,necessary information deficiency,unclarity of structure,oral style,and etc.;In terms of the problems in translatability.Synonymous juxtaposition,omitted structure,unclear sentence information,and unclear logic.To solve these problems,the thesis proposes corresponding pre-editing strategies,such as deleting redundant information,adding necessary information,rearranging the structure of the source text,unifying the writing style,omitting repetitive words,adding subjects,function words,logical conjunctions,and rewriting.Through the study of pre-editing strategies,the thesis aims to help translators to improve text quality,avoid machine translation errors,and enable machines to better understand the original text,improve the quality of machine translation,and reduce the workload of post-editing,thus improving the efficiency of translation.