A Comparative Study Between Machine Translation And Human Translation

The development of globalization,the establishment of the Belt and Road Initiative,and the deepening of communication and cooperation among countries have greatly enlarged the translation volume.Human translation(HT)can never meet such a huge demand,and more and more people have devoted themselves to the studies on machine translation(MT).Since the birth of MT,its fast speed and decreasing cost have made it popular.To make MT more practical,computer linguists have been trying to improve the accuracy of MT.With the development of globalization and internet technique,MT has played a key role in enhancing political,economic and cultural communication.The thesis has put a stress on Google Translate,comparing the translation versions between Google Translate and human translation based on different text types.The text types in this thesis mainly include the informative text,the expressive text and the operative text mentioned in Reiss' s text typology.The total linguistic data has covered the typical text types such as news reports,literature works,the inauguration speeches and advertisements.The versions of HT are chosen from official versions or recognized versions by professionals while the versions of Google translation are translated by Google Translate directly.The author has compared two kinds of translations in terms of five error types,including word errors,word orders,sentence structures,information redundancy and others errors.As a study of MT from the perspective of linguistics,the thesis has focused on different translation effects of MT on different text types.It's proved in the thesis that MT is good at dealing with the informative text most,then the expressive text and finally the operative text.The results will help people to use MT reasonably according to different text types and the usage of translation versions,and they also give some suggestions on the specialized development direction for MT.