On Machine Translation (MT) Quality Assessment: A Functional Approach

Machine translation(MT) is the application of computers to the task of text translation,which is a sub-field of computational linguistics that investigates the use of computer software to translate from one natural language to another.As machine translation is one of emerging technologies that affects our lives and work in revolutionary ways within a decade,many scholars have noticed its huge utility and presented in their papers criteria for the evaluation of machine translation systems to assess their quality,but few has related the MT quality to Functional Linguistics, while my main aim of this thesis is to evaluate the quality of the linguistic products of MT by employing the theories of functional linguistics.Machine translation is particularly effective in domains where formal language is used.Taking into consideration the limitation of Machine Translation,and to ensure a relatively high quality of the linguistic materials,I will narrow the subjects to formal languages.I adopt the Chinese version broadcasting news from the Xinhua Net (http://www.xinhuanet.com/) and employ the Google Online Translation (http://translate.google.com/) as the Machine Translation tool for research.By making discourse analysis of these of-sufficient-quality machine translation outputs by employing the basic theories of Functional Linguistics,I try to find the relationships of the three metafunctions through the analysis of the outputs of machine translation,hoping I can give some advice to both the machine translation users and the programming engineers of Machine Translation tools from the perspective of Functional Linguistics.