An Analysis Of The Mistranslation Of C-J Machine Translation Of Political Statements

Machine translation,also known as automatic translation,is the process of using a computer to convert a natural language into another natural language.It generally refers to the translation of sentences and full texts between natural languages.In order to continuously improve the quality of translation,based on rules,statistics,and examples of traditional machine translation,R&D personnel added artificial intelligence technologies such as voice recognition,image processing,and deep neural networks to machine translation.However,the complexity of the language makes it appear in the output process of a variety of translation errors,through the actual use of the summary of the problem,to provide reference for the R&D personnel to help the machine become more "clever".The existing literature and literature mostly summarizes the shortcomings of machine translation.In terms of post-translation analysis,most of them are studies of Chinese-English machine translation.There are very few researches on machine translation between Chinese and Japanese.This article selects the finalized speeches of General Political Secretary Xi Jinping's speech at the 19 th CPC Central Committee Political Bureau Standing Committee Meeting with Chinese and Foreign Journalists,and “China Premier Zhu Rongji's Direct Talk”(Chinese and Japanese bilingual).It mainly analyzes the translated version of "Youdao Translation Officer" and analyzes mistranslations of typical wrong translations.At the same time,supplemented by some examples of “Baidu Translation”and“Google Online Translation”,we will strive to make the study of machine translation issues more objective.Through the way of mutual translation between Chinese and Japanese,the typical errors of translation are classified from the perspective of source language,and the relevant problems still need to be improved are discussed.This article consists of five parts.The first chapter clarifies the research background,significance and prior research.Through practical use,the problems of typical errors in Chinese-Japanese machine translation are summarized,and the causes of mistranslations are analyzed.This can provide a basic analysis of the language level for Chinese-Japanese machine translation research.The second chapter discusses related research on machine translation methods and development history;Chapters 3 and 4 are the main research contents,and discuss the typical errors of machine translation in Chinese and Japanese from lexical,syntactic and other aspects.Among them,typical errors in vocabulary are the problems of proper noun errors,ambiguities caused by polysemy,missing translations of lexical mistranslations,and improper word segmentation.Syntactic and other errors are caused by differences in language order and expression habits of the two languages,such as Chinese and Japanese.Based on this,the thesis discusses the difficulties of machine translation in the two source languages: Chinese and Japanese.Finally,in the conclusion part,this paper comprehensively analyzes the deficiencies of Chinese-Japanese machine translation in terms of thesauri capacity,word order adjustment,matching context,etc.,and proposes an improved machine that expands the vocabulary capacity,pays attention to structural differences,improves context recognition capabilities,and improves human-machine collaboration.Translating translation quality and responding to heavy translation tasks.