On Case Study On Post-editing Of Google Neural Machine Translation

The application of Neural Machine Translation has improved the quality of machine translation since 2016.Given the circumstances,the cooperation between human translators and machine translation is accepted by translators again.From the viewpoint of the author,some of translation work will involve human translators revising the rough results provided by machine translation.The thesis examines how Google Neural Machine Translation misleads the translators and in which way it indeed helps them to work.In other words,this thesis deals with post-editing procedure rather than simple comparison between human translation and machine versions.It explores how Neural Machine Translation influences post-editing by selecting four different materials with officially-released translations and aims to rank Google's translation quality in different text types.In conclusion,the author finds that Google Neural Machine Translation performs the best in the content-focused text and it also works for subtitle translation.