A Report On The English-Chinese Translation Of Life In Lego (Excerpts)

With the continuous upgrading of information technology and natural language processing tools,the speed and efficiency of machine translation have been greatly improved.However,due to the inadequacy of machine translation in semantic recognition and contextual understanding,machine translation can hardly meet the delivery standard in translation industry,which makes human post-translation editing a necessity.The translation model combining machine translation and post-editing has been widely used in the language service industry.This translation practice uses the “MT+PE” mode to carry out machine translation and postediting of the discipline biography text Life in Lego through Youdao and Memo Q platforms.In this thesis,the errors of machine translation are classified based on the Multidimensional Quality Metrics(MQM)model,and the errors of machine translation are analyzed at the dimensions of accuracy,fluency and terminology and the corresponding post-editing strategies are proposed.In this practice,machine errors mainly include accuracy errors such as mistranslation,untranslated and omission,fluency errors such as part of speech,word order and cohesion,as well as terminology errors.Through practice and analysis,the translater proposed that translators should identify polysemous words,supplement missing information,transform part of speech,adjust word order,clarify referential relations,add conjunction words,and revise terms in post-editing stage.