ChatGPT-based Research on the Influence of AI on English Translation

Abstract

Since its launch, ChatGPT has generated a great deal of conversation across a range of businesses and sparked an unparalleled technological revolution in the English translation space. In this work, we first establish and thoroughly discuss the English translation standard. We then measure the standard of English translation using HTER, the most widely used human editing distance and assess the quality of ChatGPT’s English translation. The impact of ChatGPT on the quality of English translation is investigated in the context of practice, and the working mechanism, capability characteristics, and performance of ChatGPT are analyzed to reveal the powerful performance of ChatGPT in improving the processing speed and efficiency of translation tasks. When the instances are compared, it can be shown that ChatGPT translations outperform manual translations in terms of linguistic variety. The BF values of word type lexical density and word item lexical indexes are as high as 1503.49 and 1386.72, respectively. Additionally, the BF values of the verb phrase complexity indicator and the ratio of parallel phrases indicate the syntactic complexity of ChatGPT translations, which are both significantly superior to human translations at 38497.02 and 4082.39, respectively.