Challenges of Machine-to-Machine Translation from Persian to Arabic and its Pathology

Abstract

With the advent of computers and information technology in almost all areas of human life and the increase in the efficiency of daily work, the field of translation is no exception. Despite the rapid pace of machine translation and the all-round development of new technologies, there are still many shortcomings in this area. The present study seeks to provide a descriptive-analytical approach to the challenging features and aspects of translating Persian into Arabic by machine translators, but this does not mean rejecting or opposing the types of machines and programs developed for this purpose, because this technology, despite its drawbacks and problems that we may face in translation, has also its advantages, including the ability to act quickly and save time and money. This study aims to take a step towards improving these machine-to-machine translations from Persian into Arabic by examining the case of these deficiencies and challenges, without questioning machine translation, by providing solutions to address these deficiencies and challenges. To do this, four of the most popular and powerful translation sites have been selected and reviewed. Then a solution is provided for each of the failures. The purpose of analyzing these challenges is to identify and address these deficiencies. The findings of the present study indicate that machine translation currently cannot replace human translation in terms of accuracy and efficiency, due to its inability to understand the context and adapt words quickly to the text, style and register, extract idiomatic and non-primary meanings and identify suprasegmental features affecting the text.