The testing research of generative artificial intelligence nursing knowledge understanding based on multi-model comparison
XU Wenbo
YAO Yue
WANG Chao
CHEN Jie
HOU Hui
Abstract：Objective To assess the applicability and performance differences of generative artificial intelligence large models in nurs-ing education.Methods Select the real questions(400 questions)of the qualification examination for senior nurse practitioners.Based on subject classification(internal medicine,surgery,etc.),question stem information classification(knowledge memory questions,knowledge un-derstanding questions,etc.),option difficulty classification(simple option questions,general option questions,complex option questions),question type feature classification(non-disease example questions,disease example questions),and question stem requirements classifica-tion(positive option questions,negative option questions)Conduct multi-dimensional tests on ERNIE Bot,DeepSeek and GPT-4.Results All three models had advantageous subjects.ERNIE Bot had the highest accuracy rate(91.43％,88.24％,88.10％,82.81％)in nursing manage-ment,internal medicine nursing,pediatric nursing,and gynecological nursing.DeepSeek had the highest accuracy rate(84.85％,80.43％)in hospital infections and surgical care.GPT-4 had the highest accuracy rate(84.38％)in health education questions.In addition,the three ma-jor models had the highest correct rates in answering knowledge memory and clinical case questions(all＞86％).However,the correct rates of complex option questions(78.26％,73.91％,73.91％)were significantly lower than those of other question types.Conclusion Generative artificial intelligence can be used as an auxiliary tool for nursing education,but it needs combine with manual verification and multi-model collaborative mechanisms to optimize logical reasoning ability.In the future,reliability in complex scenarios should be enhance through the it-eration of specialized knowledge bases and the integration of clinical decision trees.
Keywords：large language modelnursing educationnursing teachingnursinggenerative artificial intelligence
Publication Date：2025-05-10
Online Publishing Date：2025-08-15(First online date of this platform, not the publication date of the document)
Pages：4( 388-391 )