The Post-editing Of The E-C Machine-translated Output Of “Kandinsky Patterns”:A Summary Report

In recent years,machine translation(MT)quality has been significantly improved,and the production model of machine translation post-editing(MTPE)has been widely used in the global translation industry.Against this backdrop,this report is based on the project of post-editing the machine-translated output produced by the MT engine Niu Trans of “Kandinsky Patterns,” which was authored by Dr.Heimo Müller and his research team.The source text is an academic paper published on ar Xiv,an open source sharing platform for research results of Cornell University in the United States in 2019.As an informative text in artificial intelligence,the source text is clear,concise,and suitable for employing the MTPE model.The report’s author used several MT engines for trial and selected Niu Trans for this post-editing project.The project was completed through four steps.First,the author of this report extracted specialized terms from the source text and compiled a bilingual glossary.Second,the source text was converted from a PDF file into a DOCX one and fed into the selected MT engine Niu Trans for the machine-translated output.Third,the machinetranslated output was imported into Yi CAT translation management platform for postediting and proofreading.In this step,the author of this report acted as the post-editor to correct the mistakes or solve the problems in the machine-translated output.Finally,the post-edited output was revised and finalized.This report discusses the whole process,with an emphasis on the third step.This report only discusses selected mistakes or problems in the machine-translated output by Niu Trans of one academic paper in the field of artificial intelligence.Whether Niu Trans will exhibit similar problems when it is used to translate other academic papers in the same or other fields of study,or whether other MT engines have similar problems concerning the translation of the same source text,it cannot be predicted and will have to be explored in the future.Moreover,the post-editing skills for the machinetranslated output of other academic papers in artificial intelligence or other fields of study need further exploration.