A Corpus-Based Evaluation of Google Translate Ability to Translate Economic Phrases from Persian into English

Nowadays, translation is an essential need for interactions around the globe and machine translation (MT) are desired to handle the huge volume of daily translations due to cost-cutting and time-saving. In this study we attempted to assess the bleeding-edge Google Translate to find out whether it can play the role of economic dictionaries. To this end, 100 Persian economic phrased were selected from the bilingual glossary of a book which was a source for translating economic texts. Then, the opted phrases were translated by Google Translate from Persian into English and the outcome was divided into three categorization; Google’s renditions which were: 1. Identical to the book translation, 2. Defective or extraneous translations and 3. Akin to the book translations. In additions, the mean of the frequencies of 40 economic phrases chosen from online Oxford Dictionary was calculated as a criterion to determine low and high frequencies. GloWbE was the corpus through which the frequencies were extracted. The results revealed that more than 60% of Google Translate output were perfect, i.e. identical or better than the book translations and approximately 30% were above the average frequency. Nevertheless, substitution of Google Translate for economic dictionaries in Persian language is a wishful thinking yet

Keywords:

Google Translate , Corpus-based evaluation , Economic Translation , Persian , English