Artificial Intelligence and Translation Technology

Abstract

Many unprecedented changes in translation that can be seen today resulted from technological advances. In this thesis, the way in which translation technologies based on application of artificial intelligence (AI) make an impact on translation practice and translators is discussed. It is argued that translation technologies have increased significantly productivity and quality in some fields of translation. However, these technologies also lead to considerable uncertainties and challenges for translators. The thesis discusses in detail two machine translation systems – “PROMT” and “Google Translate”, which are currently the most popular among Internet users. The work of these systems is organized according to different principles: “PROMT” works according to the so-called principle of “translation according to the rules” and copies the human translation activity; Google Translate translates using technology based on the principle of statistics. During testing, typical mistakes made in machine translation have been identified, and the possibilities and areas of use of each of the translation systems have been determined. It has been established that these systems are unsuitable for working with texts containing a large number of complex sentences. These programs work mainly at the level of collocation, and they can be successfully used to translate formalized texts, such as technical documentation, consumer instructions, formal descriptions, etc., which are characterized by the use of simple common sentences and which do not contain sentences with complex syntactic constructions. It is concluded that professional translators can resort to the help of these systems in their practical activities when working with certain types of texts, which helps save time and optimize the translation process. However, a prerequisite for achieving high quality translations made using computer programs is their post-editing.