A Contrastive Analysis Of Google Translate And Youdao Translate

MT(machine translation)has been developing rapidly in recent years.With the combination of neural network technique,MT is becoming increasingly intelligent,so it is of necessity to study MT.The paper aims to compare the translation quality of two MT tools in the text of popular science by analyzing the disadvantages of MT tools and put forward some suggestions for further improvement.The study compares two MT tools——Google Translate and Youdao Translate by using popular science work Darwin’s Unfinished Symphony as the analysis text.150 sentences from the second chapter are selected as the example.The Chinese version of the book published by Citic Press Group serves as the translation reference.The paper marks the translation errors of two MT tools and determines error types.SPSS(statistical package for the social sciences)is used to make a quantitative comparison and analysis between and within two MT tools.Possible causes that might affect the efficiency of MT will be further speculated.The findings suggest that there is a substantial similarity between Google Translate and Youdao Translate in lexical,syntactic and orthographic errors.The statistics show that noun errors,noun phrase errors and verb errors appear with the highest frequency among the 18 error types both in two MT tools,while errors of abbreviation and passive voice measure the least.Lexical errors of two MT tools accounts for more than 60% of the total errors,which significantly higher than their syntactic errors.Furthermore,it is found that Youdao Translate performs better than Google Translate regarding lexical and syntactic errors.Specifically,Youdao made fewer mistakes in errors of noun,verb,adjective,conjunction and word permission,and its errors of the noun phrase,preposition phrase and participle are significantly less than its counterpart.The following conclusions are drawn after the analysis and summary of the problems in the practice of MT.Compared with Google Translate,Youdao Translate does a better job in fidelity and comprehensibility.MT performs poorly in polysemy at the different context,long sentences,and translation of participle compound words.So according to this situation,two suggestions are put forward for improvement: the translation mode should be based on “machine translation + human translation” and on “corpus + annotation system.”...