Pre-editing In Chinese To English Machine Translation

Although machine translation is very popular at present, the readability of thetarget language texts is far from satisfaction. A large quantity of research has beendone aiming at optimizing machine translation performance. Yet, there are stillamounts of work remaining to be done to ensure machine translation engines acomprehensive and swifter development.This paper is a case study of Chinese to English translation in the Googlemachine translation engine. It intends to find out the circumstances in which errorswill occur in the English language texts generated by the Google machine translationengine. Then, the study makes an attempt to find out a series of pre-editing strategiesto meliorate the efficiency and accuracy of the Google machine translation engine.The data analyzed in the thesis comes from the Chinese to English translationexercises in Shanghai Advanced Interpretation Accreditation Examination (AdvancedWritten) and TEM-8. Both qualitative and quantitative analyses are conducted in theseven chapters of this thesis.Chapter One serves as the introduction of the whole paper and demonstrates theresearch purpose, questions and significance, as well as the organization of this thesis.Chapter Two is the literature review of previous relevant studies on machinetranslation, machine translation errors and pre-editing. In Chapter Three, the authorintroduces the research methodology, which deals with experiment design, samplechoosing and research methods. Chapter Four presents the circumstances in which theGoogle machine translation errors tend to arise and probes into the causes of thesedeficiencies. Chapter Five covers the pre-editing strategies categorized based on thesubject study of the thesis and the previous analysis of the former chapters. ChapterSix demonstrates the experiment, including data analysis and experiment results.Chapter Seven demonstrates the conclusion of the study.According to the analysis, the paper comes to the conclusion that pre-editing isof vital significance in enhancing machine translation performance. The pre-editing strategies proposed in the paper are as follows:1. re-ordering;2. adding;3. omitting;4. pre-translating;5. segmentation;6. rewriting;7. paraphrasing;8. simpliyfing;9,’re-punctuating. Whats more, these strategies are tested by BLEU and prove to beeffective in promoting the efficiency of the Google machine translation engine and thequality of texts it generates.