Machine Translation Quality Assessment of Literary Texts from English to Persion and Persian to English Based on Waddington’s Model (Method A)

This research described the quality of Machine Translation (MT) in translating literary texts from English to Persian and Persian to English. The type of MT system applied for this purpose was Google Translate, and the model used to assess the quality of the MT system was Waddington’s method A model. The main purpose of this study was to specify which translation was evaluated to be more qualified, English to Persian or Persian to English, and in what points of view the texts were translated inappropriately based on Waddington’s model. For this purpose, ۴۰ paragraphs were selected from ۳ literary books; half of the paragraphs were in English, and half of them were in Persian. The English paragraphs were translated into Persian through Google Translate, and the Persian ones were translated into English through the same system. The types of errors in each of the translated texts were found based on the selected model, and in order to determine whether the Persian translation or the English translation was more qualified, the types and the number of errors were assessed through Chi-square test. The results demonstrated no significant difference between the qualities of Google Translate from English to Persian and Persian to English.

Keywords:

Translation Quality Assessment (TQA) , Machine Translation (MT) , Google Translate (GT) , Waddington’s model , Literary texts.