A Case Study On The Output Of Machine Translation And The Post-editing Skills In Contract Translation

Since the focus of researches on machine learning shifted to Neural Machine Translation(NMT)from 2013,machine translation involving human language and computer language has been benefiting remarkably from such a trend and has seen significant improvement in the quality of its outcome.In 2014,Google released an article about its achievements in NMT.As one of the users of machine translation platforms,I found that the quality of machine translation outcome really improved a lot,and Google Translate is a typical example.Therefore,when I had to finish the translation in a week of a Purchase and Sale Contract(the Contract)whose word count was over 14,000,it hit me that I might adopt Machine Translation + Post-editing(MT+PE).I tested a few paragraphs in three platforms,Youdao Translate,Baidu Translate and Google Translate,and found that the three all could produce translation that sufficed to convey the meaning of the source text.I learned from the experience that MT with its improved translation capacity,can be a good helper in translating contract,a kind of text that does not require creativity in language,so I chose the Contract as a case I will study in the paper to prove the value of MT in contract translation.However,MT alone is not enough to produce a perfect translation,so PE is another focus of the paper.To be more clear,I'll list illustrative examples extracted from the Contract to analyze the advantages of MT in terms of word and paragraph translation,and also through examples to share PE skills I got from that experience,to provide more insights for translators into the value of MT and the development of PE skills,so that translators are willing to adopt MT+PE in contract translation or do more researches on the topic in the future.