A Comparison Analysis Of Chinese-English Translation Quality In Urban Design Description Based On Translator-computer Interaction

Computer science and machine translation(MT)platforms have been greatly and radically improved in the past decades.Their development has made the translator-computer interaction a better approach for translation than traditional human translation.After both MT platform and human translators check and edit the translation,ideal outputs can be generated to satisfy customer’s needs.There are commonly two ways of translator-computer interaction,namely "MT+post-editing",and "pre-editing+MT".Both methods are widely used in different circumstances,but few studies before give comparison and analysis of these methods.The thesis focuses on the comparison of quality in Chinese-English translation in the field of urban planning through translator-computer interaction.The experiment generates outputs through two methods,"MT+post-editing" and "pre-editing+MT".A scoring sheet generated from Multidimensional Quality Metrics(MQM)framework is used for quality assessment.All of these procedures will analyse the advantages and shortages of two ways of translator-computer interaction in translating professional documents.The research shows that "MT+post-editing" generates higher quality outputs than"pre-editing+MT" in the Chinese-English translation of urban planning document,but errors cannot be fully avoided.The "pre-editing+MT" method is easier and quicker,but its output quality cannot be easily controlled.By comparing the translation of Chinese and English native speaker,it can be observed that English native speaker prevails in the translation of cultural knowledge and English grammar.In the following questions,"pre-editing+MT" group spends less time in translation than the other group,and it suggests a review process to be added after MT.Both groups regard translator-computer interaction as a time-saving method,and they believe human involvement is indispensable in this process.