摘要
目的/意义探索医学信息学跨学科主题演化路径识别方法,为该领域跨学科研究布局与科研管理提供参考。方法/过程首先利用BERTopic和大语言模型(large language model,LLM)在获取的医学信息学文献中识别全局主题和阶段主题;然后基于主题影响力与学科多样性二维分析框架,确定跨学科主题;最后分析跨学科主题的演化路径。结果/结论基于Topic-LLM的分析方法,发现医学信息学跨学科主题的学科多样性和主题影响力呈现稳步增长趋势。
Purpose/Significance To explore the identification method of interdisciplinary topic evolution path in medical informatics,so as to provide references for the interdisciplinary research layout and scientific research management in this field.Method/Process Firstly,global and staged topics are identified from medical informatics literature using BERTopic and large language models(LLM).Secondly,interdisciplinary topics are determined based on a two-dimensional analysis framework integrating topic influence and disciplinary diversity.Finally,the evolution paths of these interdisciplinary topics are analyzed.Result/Conclusion An analysis method based on Topic-LLM is proposed,and it is found that the disciplinary diversity and topic influence of interdisciplinary topics in medical informatics exhibit a steady growth trend.
作者
苏中琪
董淼
崔雷
SU Zhongqi;DONG Miao;CUI Lei(Library,China Medical University,Shenyang 110122,China;Department of Finance,China Medical University,Shenyang 110122,China;School of Health Management,China Medical University,Shenyang 110122,China)
出处
《医学信息学杂志》
2026年第1期55-61,69,共8页
Journal of Medical Informatics
关键词
BERTopic
大语言模型
跨学科
主题识别
主题演化
BERTopic
large language model(LLM)
interdisciplinary
topic identification
topic evolution
