| With the rapid development of translation technology today,in order to improve the quality and efficiency of translation,human-machine coupling has become the general trend of the translation industry in this era,and machine translation has even become an indispensable tool for translators.Good enough post-editing and terminology management skills have become a new standard for translators.Compared to the initial stage,the intelligence of machine translation has been amazingly developed owing to the research breakthroughs in the field of natural language processing of artificial intelligence.The processing of text is no longer a simple "word-to-word translation".However,there are still many limitations.There are many terms and vocabularies that machines cannot identify by virtue of their original memory and neural network in government affairs texts,which have caused great obstacles to the machine translation process.This thesis takes the English translation project of the "Shanghai Anti-Monopoly Compliance Guide for Undertakings" as an example,analyzes the common types of errors in machine translation in government affairs texts,and proposes some solutions to these errors.In practicing,the author used the Yi CAT platform to pre-process the original text,then explains how these terms are extracted manually,and then reviews the types of machine translation errors and analyze them,and finally propose the corresponding types of post-editing strategies.This thesis is divided into three chapters.The first chapter introduces the basic situation of the "Shanghai Anti-Monopoly Compliance Guide for Undertakings" project,including the basic introduction of the project,the platform used and the overall translation process.The second chapter introduces the text characteristics of the project and also combines multiple examples to classify and analyze machine translation errors.The third chapter analyzes the principles of extracting terms manually and the selection of corpus and finally summarizes and proposes corresponding post-editing solution. |
