深度学习技术在中医方剂文本挖掘中的应用与挑战

打开文本图片集

中图分类号：TP18文献标识码：A

文章编号：2096-4706（2026）04-0073-08

Application and Challenges of Deep Learning Technology in Text Mining of Traditional Chinese Medicine Formulas

LIYiwen，LI Zhibiao，LIU Zixuan （SchoolofInteligentedicineandfoationEngineg，angxiUiversityofnseedcine，ncang0ia）

Abstract： Traditional Chinese Medicine （TCM） formulas have developed over thousands of years and embody the precious experienceandintellectualcrystalizationofmedicalpractitionersarossalldynasties.Inrecentyears，therapiddevelopntof DeepLearing technologyhasprovidedanefectiveapproachforexploringtheirinherentlogic.Thispapersortsoutandanalyzes the applicationsofDeeLearning inTCMformularesearchandclasifies itbytechnicalcategory.Facedwith theinherently complex knowledge systemofTCMformulas，DepLearing technologyexhbits itsunique advantages，especiallyineficacy classifiation，forulaompositionaalysisndnoveldrugdiscoveryHowever，curentresearchillfceschalengesuchas limitedandunbalanceddata，complexfeatureextraction，greatdificultyinmodelconstructionandinsufcientintegrationof TCMknowledge.Inthe future，researchcanadvance aroundtheconstructionofstandardized formuladatasets，multimodal data fusion，thediingofutig-dgeArtifcialIntellgenethologiesandinterdisiplinaryitegation，oastofutreploe the mechanism of action and scientific connotation of TCM formulas.

Keywords： Traditional Chinese Medicine formula; text mining; Deep Learning;Large Language Model; Artificial Intelligence

0 引言

方剂是中医药学体系中的核心组成部分，医者根据患者的具体病情、体质等因素，在辨证审因之后，选择合适的药物，并按照一定的组方结构要求，妥善配伍而成[]。（剩余17081字）