Error Types Of Machine Translation And Post-editing Methods

With the improvement and development of computer technology,the application of machine translation is more and more extensive.Machine translation can improve translation efficiency and reduce translation cost,but the quality of machine translation is still far from that of manual translation and cannot meet the requirements of customers.However,post-editing,which has emerged in recent years,can effectively make up for the shortcomings of machine translation,so the model of machine translation and post-editing(MTPE)is widely used and adopted by many language service providers.This thesis is a translation report on machine translation and post-editing,with a source text from the book A History of Costume and Fashion-Medieval World.Since the source text is an informative text,with precise words,clear logic,and repeated professional terms,it is suitable to be translated under machine translation with post-editing.After comparing the results of several MT engines,the author chooses Niu Trans to pre-translate the selected text,and then imports it into YiCAT auxiliary translation platform to edit and proofread the MT output for several times.Based on the translation practice and relevant studies at home and abroad,the machine translation errors in this practice can be divided into three categories-lexical,syntactic and textual,with the reference of the types of machine translation errors by Li Mei,Zhu Ximing(2013),Cui Qiliang and Li Wen(2015).Under the guidance of memetics,this thesis summarizes the post-editing methods from the above three aspects through analyzing ample examples of MT error types in this translation practice,so as to provide some references for the research and practice of machine translation and post-editing.