Research on the linguistic features of human translation and machine translation based on corpus:Taking Peter Pan's four translations as an example
Machine translation has a profound impact on various fields of translation today.To observe the differences in linguistic features between human translation and machine translation in the field of literature,the authors built Peter Pan's English-Chinese parallel corpus of the original English text and its two human translations and two machine translations.Tools such as Python,CRIE,Emeditor and ABBYY Aligner were used to conduct comparative research on 17 linguistic structures of the four translations from the lexical and syntactic levels.The research results show that in the field of literature,machine translation is not as good as human translation,with the former being less flexible than the latter in processing words and sentences,mainly inflected in features of evident"explicitation"of pronouns,and lack of logic in writing.It is found from the comparison of the two human translations that Ren Rongrong's version is full of rhythm and easy to read,while Yang Jingyuan's version is rich in language and fluent in text.It is found from the comparison of the two machine translations that the Google version has more diverse vocabulary expressions than the Youdao version,while the Youdao version is closer to Chinese expressions.
NETL
NSTL
万方数据
