A Report On The Post-Editing Of Machine Translation Based On The English-Chinese Translation Of The International Fencing Federation Handbook Of Regulations For World Champions

Based on the English-Chinese translation of the International Fencing Federation Handbook of Regulations for World Champions(Senior),this report introduces the postediting of machine translation.The task is released by a sports event company in Chengdu,requiring the translator to translate the handbook issued by International Fencing Federation in 2017 into Chinese.The Chinese version will be kept as the company’s internal file.With14063 English words,the handbook consists of the following 6 chapters: bidding,administrative and logistic obligations,infrastructures and venues,the competition,communication-marketing and TV and FIE medical handbook.Before post-editing,the author settles two main matters: the selection of machine translation platform and the building of termbase.To choose the most suitable platform,the author calculates the BLEU scores of 7 widely accepted platforms(Google,Bing,Baidu,Tencent,Sogou,Youdao,Niu Trans)respectively.The results show that Google Translate gets the highest score,becoming the first choice.Meanwhile,the author finds that even though Google Translate owns the highest BLEU score,its score is still lower than the average level.To make clear the reason,the author employs Google Translate to process Chapter One of the handbook and summarizes the error categories occurring in this part.Term mistranslation seems to be the most prominent problem.Therefore,the author builds a fencing termbase to improve the efficiency of correcting the mistranslated terms in the following work.In the process of post-editing,considering the features of the source text,the author summarizes 3specific methods(accuracy improvement,fluency polishing,DTP and document spec)applicable in this text type,with reference to Huawei Translation Services Center’s machine translation error categories and Translation Association of China’s machine translation categories.Several examples are provided.The author corrects mistranslation,omission &redundancy and ambiguity occurring in the translation produced by machine to guarantee the accuracy of the translation.Meanwhile,the author adjusts part of speech,sentence structure,expression and diction to improve the readability of the translation.Besides,in view of the special Desktop Publishing of the source text,the author places emphasis on punctuation,format and bulleted list in the post-editing process.Reviewing the whole translation task,the author insists that the utilization of machine translation in informative text is still of great significance.Nowadays,the model of “machine translation + post-editing” is still the trend in the field of translation.Translators shall embrace the machine translation and explore new methods to cooperate with it positively.What’s more,necessary preparations before post-editing,such as the building of termbase,have a direct effect on the efficiency of the post-editing.