abstract

The purpose of this paper is to linguistically analyze the error patterns that appear in artificial intelligence translator. As the quality of artificial intelligence translation increases, errors still occur in machine translation despite being used in various ways in language education, text linguistics, and translation. This study examined the types and characteristics of errors that occur when translated from English to Korean by analyzing the errors in the results calculated by an artificial intelligence translator. In fact, by constructing the commencement speech text data, the types of errors were classified in terms of language, and the error patterns appearing in the commencement speech text were examined. In this study, a linguistic error analysis methodology for artificial intelligence translator outputs was presented to identify the form and semantic characteristics of translation text errors in English and Korean speeches. It is expected that this study will provide basic data for research related to error analysis using artificial intelligence translators and further help research on language engineering and applied linguistics.