医学语言模型研究  

Research on medical language model

作　　者：赵立君 张军雁 何倩 庄严 郭锐 ZHAO Lijun;ZHANG Junyan;HE Qian;ZHUANG Yan;GUO Rui(China Mobile(Chengdu)Information and Communication Technology Co.,LTD.,Chengdu 610213,China;People's Liberation Army General Hospital,Beijing 100039,China)

机构地区：[1]中移(成都)信息通信科技有限公司,四川成都610213 [2]中国人民解放军总医院,北京100039

出　　处：《长江信息通信》2023年第11期1-7,11,共8页Changjiang Information & Communications

基　　金：科技创新2030——“新一代人工智能”重大项目(2021ZD0140408)。

摘　　要：随着医疗行业不断发展,医学数据呈现指数级增长趋势,如何高效处理和利用这些数据成为当前医疗领域面临的一个重要问题。自然语言处理(NLP)技术的不断发展使得医学语言模型的构建成为可能,这为医学数据的挖掘和分析提供了新的途径。文章首先梳理了医学语言模型构建的准备步骤,之后从无监督学习、监督学习和迁移学习等方面叙述了医学语言模型的具体构建方法,并探讨医学语言模型所面临的挑战和未来发展方向。With the continuous development of the medical industry,medical data shows an exponential growth trend.How to deal with and utilize these data efficiently has become an important issue facing the current medical field.The development of natural language processing(NLP)technology makes it possible to construct medical language model,which provides a new way for medical data mining and analysis.This paper reviews the construction methods of medical language model and its ap-plication in the medical field,including medical text classification,entity recognition,relationship extraction,knowledge map construction and so on.At the same time,this paper discusses the challenges facing the medical language model,including the complexity of language,the particularity of the medical field and data privacy issues.Finally,the future development direction of medical language model is prospected.

关 键 词：自然语言处理 医学语言模型 医学文本分类 实体识别 关系抽取 知识图谱 

分 类 号：TN915.63[电子电信—通信与信息系统]