A Corpus-Based Evaluation of Human and Machine Translation Quality for the Literary Classic The Analects

Abstract

This study is based on a self-built English translation corpus of the Analects, which includes two human translations and two artificial intelligence (AI) translations. By integrating research methods from corpus linguistics and quantitative linguistics, and employing a translation quality assessment model, this study conducts a comparative analysis of the translation quality of human and AI translations from two major dimensions: lexis and syntax. The results show that in the lexical dimension, human translation demonstrates greater flexibility in word selection and lexical richness, and is also more versatile in part-of-speech conversion. In the syntactic dimension, AI translation systems can effectively identify and reproduce the logical structure of the original text, generating logically coherent and structurally complete translations, but they tend to exhibit a certain degree of word-for-word translation. Based on these findings, this study points out that both human and AI translations have their own strengths and limitations. Therefore, this paper proposes strategies to optimize the human-computer collaborative translation model in order to enhance the application efficiency of artificial intelligence in the translation of classical works.