Abstract This study examined the definitions of sample words in both the large language model (ChatGPT) and the Chinese learner’s dictionary ( Juzi Chinese ), comprehensively evaluating the accuracy, comprehensibility, practicality, and diversity of word definitions and example sentences, while exploring ChatGPT’s advantages and limitations in assisting the Chinese learner’s dictionary compilation. In terms of definition quality, ChatGPT frequently displayed issues such as missing senses, inaccurate or erroneous definitions, part-of-speech labeling errors, redundant senses, and circular explanations based on near-synonyms. This is particularly evident in words with complex semantics and multiple meanings, where its definition accuracy was relatively low. In terms of example sentence quality, while GPT-generated examples were practical, they still lacked accuracy and diversity. Moreover, ChatGPT’s definitions and examples lacked metalinguistic awareness and showed poorer comprehensibility than Juzi . At present, ChatGPT may serve as a supportive tool in lexicographic work, with its potential value likely to increase through enhanced human-AI collaboration.
Rongyan Liu (Fri,) studied this question.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: