Evaluation of different English translations of Holy Koran in scope of verb process type

Access to a text which is the best translation from one language to another is one of the most important requirements of information societies. In other word, there may be many translations of a source text in which the most convenient one should be selected. The Holy Koran is one of the well-known books in the world which has been descended in Arabic and most of peoples would like to understand more about this book. As a result, there are more than 100 translations of Koran from Arabic to English. In this paper a system has been designed to evaluate different English translations of Koran using tools and concepts such as pos-tagging, natural language processing, computational linguistic and machine learning. To do such an evaluation, each verb process type in translated text should be compared with it´s of Koran text. The system uses Halliday Grammar, which is a useful theory for analyzing a formal text, to do this. It has been assumed that each verb in the text of Koran has been tagged manually. At the other hand, to detect each verb process type in translated English texts, a tagger is used to detect and tag each verb based on Halliday grammar. The best translation is one in which the number of similar verb process type in the source and translated text is maximum. This evaluation´s been performed on 5 English translations of Holy Koran (Qaraei, Pickthal, Arberry, Progressive Muslims, Yusuf Ali) in scope of Bashir and Nadhir Ayahs of Baqarah and Alelmran Surahs and the results show that Yusuf Ali translation is the best. In this paper, it has been tried to create not only a new way to analyze the quality of translating a text but also take a step to discover concepts and meanings of the Koran. Although it is seemed that some of Koran words are so meaningful that it is nearly impossible to put them in one class of verb process type.