Stakhmych Y. Computer Modeling of Natural Languages (a Case Study of Translation Software Systems)

The thesis suggests a new approach to the investigation of computer modeling of English, Ukrainian and Russian in translation software systems Pragma 6.3 Business, Google Translate, Bing Translator, Babylon, Free Translation, Yandex. The methodology of quality assessment of translation software systems is improved by the detailed criteria used to identify lexical, grammatical (morphological and syntactic) and lexico-grammatical mistakes. The introduced English text samples allow to assess the efficiency of translation software systems in translating texts on different topics from English into Ukrainian and from English into Russian. The prevalence of Yandex over other systems in the number of mistakes in English-Ukrainian and English-Russian machine translations is determined. The optimization of translation software systems is defined to require the enrichment of their linguistic components, improvement of algorithms and further study on the basis of multilingual text corpora.