Machine Translation Error Analysis And Post-Editing Strategies

Machine translation plays an increasingly important role in the language service industry because of its advantages in speed and efficiency.The proliferation of machine translation systems has already been an indisputable fact from which both language service enterprises and individuals can obtain great convenience.However,there are still some limitations in machine translation systems.The emergence of post-editing can effectively compensate for the congenital defects of Machine Translation.And over time,the post-editing of machine-translated output has become an indispensable part of the translation activities and is adopted by a growing number of language service providers.This thesis,taking the translation report on academic monographs for example,analyzes and summarizes the typical errors of machine translation,and provides post-editing strategies and methods accordingly based on the general principles of post-editing guidelines(TAUS)and existing research results made by both domestic and foreign scholars.The summary of typical errors of high frequency and strong regularity occurring in machine-translated output can,on the one hand,offer language material support for further processing of machine translation,namely,the automated processing of post-editing;on the other hand,provide feedback for the research and development of machine translation system,so as to improve the performance of machine translation systems.Besides,the significance of this translation mode on translating academic monographs as well as other types of text has also been discussed.