Artificial Intelligence: Translating SoftwareAccuracy in Translating English-Persian Texts inand out of Context

Machine translation as artificial intelligence (AI) is globally applied increasingly and one of the popular Iranian translatingsoftware is called Padideh which is not merely applied to obtain data from different languages but also as a medium tocommunicate and interact. This descriptive paper aimed to study machine translation functionality accuracy in the English-Persian translation of texts in and out of context. The sample of in-context texts was selected from the novel “The Old Man andthe Sea” while the out-of-context texts were a group of sentences with various grammatical features based on the book series“Headway”. The research outcomes revealed that the semiotic-based error analysis associated with machine translation inPadideh followed the translation process from the source language forms (signifier) to the target language forms (signifier) byhighlighting the concept equivalence (signified) of both languages. Padideh applied not only words, and phrases, but sentences.The study results also showed that AI had some ambiguities in sentential and contextual analysis on certain grammatical,lexical-semantic, and pragmatic features. The findings have practical implications not only for greater consistency and higherproductivity but also for greater flexibility and better quality which can have some innovations for better AI-enhancedtranslation devices.

Keywords:

machine translation , Padideh , accuracy , artificial intelligence