A Study On The Corpus-based CAT Model Of Local Regulations

Recently, modern information and communications technology allows people to manipulate hugeamount of texts as well as other materials electronically, and the process of those texts further promotes thedevelopment of corpus linguistics itself. At the same time, the development of corpus linguistics and thedesign of modern corpus have provided new approaches to the research on interdisciplinary area such asforensic linguistics. Under the condition of increasingly tight international communications, parallelcorpora on national law or local regulations not only provide a powerful tool for forensic linguisticsresearch, but also play a positive role in legal practice. The global economic integration has promotedincreasingly frequent exchanges between China and other countries around the world. The translation andintroduction of Chinese laws and regulations overseas becomes more and more prominent in foreignexchange and trade activities. Although the Legal Affairs Office Attached to the State Council and itsbranches at provincial levels has set up recording review and translation divisions and some national lawsand local regulations have translated versions, both the quantity and quality are far from satisfying theactual needs. Yet at the same time, legal language is characterized by its definiteness and accuracy ofwording and terminologies, so a lot of repetitive work can be done by computer before human intervene. Asthe core part of computer-aided translation, translation memory combined with the corpus could have greatpotentials in the semi-automatic translation of laws and regulations.Based on the above background, the present empirical study attempts to design a rudimentalcorpus-based CAT model to improve the efficiency of translation of local regulations in Gansu provincewhile maintaining a decent accuracy. The study centers on issues about the building of corpus-based CATmodel, the quality of translation based on the model and the expandability of the model. The thesis is madeup of five chapters. The First Chapter presents the background, purpose and significance of the presentstudy. In Chapter Two reviews of previous studies at home and abroad are made. In Chapter Three, detaileddescriptions of the theory and method of corpus linguistics and computational linguistics and TQA theoryand cloud computing are made. In Chapter Four, research questions, research instruments and researchprocedure are presented. Major design of this study includes following aspects. Primarily several researchquestions about the study are put forward. In the next place, this study designs the corpus-based CAT model.Then it attempts to construct the basic translation memory of local regulations in Gansu province bymaking use of existing Internet resources. In Chapter Five, the application of the model is implemented with sample. Then the data are analyzed by TQA and the research findings are discussed. Chapter Six is theconclusion in which the existing limitations as well as potential breakthroughs in future are proposed.Compared to human translation and pure machine translation, CAT combines the advantages of speedwith exactness, especially it assists translator to achieve fast and efficient translation through the functionof memorizing and replacing, thus the current translation of local regulations in Gansu province can beeffectively improved. At the same time, this thesis put forward several strategies for quality control ofcorpus-based CAT. However, due to the limitations of subjective and objective factors, this study is only apreliminary attempt. There remain a lot of problems. For instance, the size of text materials is small, thesubject of texts varies a little and the process of TQA has large room to improve. All the limitations areexpected to improve and refine in future study.