数字人文视域下本草古籍知识组织的创新路径研究—以《植物名实图考》为例
|
欧阳天??
|
香港大学李嘉诚医学院中医药学院
|
摘要：数字人文的兴起为本草古籍的发掘与创新提供了新视角，推动传统文献研究从线性阅读走向知识挖掘与智慧服务。本草类古籍条目清晰、结构规范，在知识组织上具备天然优势，但当前数字化多停留在图像扫描和全文检索，缺乏知识元的精细标注与语义关联，难以构建深度知识网络。如何从“保存文献”转向“组织知识”，已成为该领域的关键问题。本文首先梳理本草古籍数字化的三个阶段：图像型、文本型和图文型数据库，并比较“中华医典”“中医典”“瀚堂典藏”等平台在规模、功能与技术上的差异，指出现有系统虽已实现图文对照与全文检索，但在语义关联、跨库整合和智能检索方面仍有明显不足。随后，以清代《植物名实图考》为例，分析其条目化、结构化的内容特征，揭示其在名称、性味、功效、主治、采收、禁忌等维度上的清晰边界，说明此类体例天然适合知识元的识别、抽取与重组。在此基础上，借鉴知识元理论与知识图谱技术，提出“化面为点—化点为面”的双阶段重构路径：先颗粒化为最小知识单元，再通过语义关联聚合为知识图谱，实现从线性文本到多维网络的范式转换。围绕这一路径，构建了包含载体层、内容层、应用层的三层创新框架，涵盖数据基础、知识元标引、图谱构建与智能服务等环节。并以《植物名实图考》“人参”条目为例，演示知识元抽取、语义关联与隐性知识发现的具体流程。研究表明，本草古籍的结构化体例与知识元方法高度契合，数字人文技术可有效推动其从“活下来”走向“活起来”—即从数字化存储转向知识重组与智慧服务。本研究为本草古籍的数字化保护与智能服务提供了理论参考与方法路径，也为数字人文与中医药文献研究的深度融合探索了可行方案。
关健词：数字人文；本草古籍；知识组织；知识图谱；《植物名实图考》 |
Research on Innovative Paths of Knowledge Organization of Ancient Materia Medica Texts from the Perspective of Digital Humanities：A Case Study of Illustrated Catalogue of Plants Names and Their Identities (Zhiwu Mingshi Tukao)
|
Au Yeung Tin Lui
School of Chinese Medicine, LKS Faculty of Medicine, The University of Hong Kong
|
Abstract: The rise of digital humanities offers a new perspective for the exploration and innovation of materia medica classics, advancing traditional philological research from linear reading toward knowledge mining and intelligent services. Materia medica classics, with their well-structured entries and standardized formats, possess inherent advantages for knowledge organization. How-ever, current digitization eff orts largely remain at the level of image scanning and full-text retrieval, lacking fi ne-grained annotation of knowledge elements and semantic linkages, thus failing to construct deep knowledge networks. How to shift from “preserving texts” to “organizing knowledge” has become a critical issue in this fi eld.This paper fi rst reviews the three phases of digitization for materia medica classics: image-based, text-based, and image-text hybrid databases. It compares several major platforms—including Zhonghua Yidian, Zhongyi Dian, and Hantang Diancang—in terms of scale, functionality, and technology, identifying that while existing systems have achieved image-text alignment and full-text search, they remain defi cient in semantic association, cross-database integration, and intelligent retrieval. Subsequently, taking the Qing-dynasty Zhiwu Mingshi Tukao(Illustrated Investigation of Plant Names and Realities) as a case study, the paper analyzes its itemized and structured content features, revealing clear boundaries across dimensions such as name, property, effi cacy, indication, harvesting, and contraindications. This demonstrates that such bibliographic formats are naturally suited for the identifi cation, extraction, and recombination of knowledge elements. Drawing on knowledge element theory and knowledge graph technology, the paper proposes a two-stage reconstruction pathway: “from surface to points — from points to surface.” The fi rst stage granularizes the text into minimal knowledge units; the second stage re-aggregates these units through semantic association into a knowledge graph, thereby realizing a paradigm shift from linear text to multidimensional networks. Around this pathway, a three-layer innovation framework is constructed—comprising the carrier layer, content layer, and application layer—covering data infrastructure, knowledge element indexing, graph construction, and intelligent services. Using the Ginseng(Renshen)entry in Zhiwu Mingshi Tukao as an example, the paper demonstrates the specifi c procedures of knowledge element extraction, semantic association, and implicit knowledge discovery.The study shows that the structured format of materia medica classics aligns well with knowledge element methodologies, and digital humanities technologies can eff ectively propel these texts from “surviving” to “revitalizing”—that is, from digital storage toward knowledge reorganization and intelligent services. This research provides theoretical references and methodological pathways for the digital preservation and intelligent services of materia medica classics, and also explores a feasible approach for the deep integration of digital humanities and traditional Chinese medical literature research.
Keywords : digital humanities；materia medica classics；knowledge organization；knowledge graph；Zhiwu Mingshi Tukao
参考文献
[1] 安波.融合思维链的中医药古籍多任务知识抽取方法研究[J].农业图书情报学报,2025(12):81-94.
[2] 沈旺,于琳,冯欣,等.基于动态知识图谱的中医疫病古籍知识演化研究[J].现代情报,2025,45(2):26-37.
[3] 张旭 , 赵彦辉 , 刘树春 . 本草古籍数字化及嵌入学术资源平台的探索与实践 [J]. 中国中医药图书情报杂志,2017,41(6):5-9.
[4] 周冉冉,杨凤,侯鉴宸,等.基于知识元理论的中医古籍深度标引思考与展望[J].甘肃中医,2026,39(04):32-37.
[5] 佟琳,张华敏,佟旭,等.基于命名实体识别的《神农本草经》知识图谱构建及可视化分析[J].中国中医药信息杂志,2024,31(8):37-43.
[6] 李贺,邵文诗,刘嘉宇,等.面向知识融合的本草典籍知识图谱实体对齐研究[J].现代情报,2026,46(3):30-43.
[7] 佟琳,张磊,曾子玲,等.本草领域本体构建及其应用[J].医学信息学杂志,2025,46(3):61-67.
[8] 杨凤,张艺,陶晓华,等.基于知识元理论与临床需求深度融合的中医古籍数智化研究路径和范式 [J]. 中医杂志,2024,65(12):1201-1207.
[9] 陈健,杨凤,任巧生,等.基于知识元理论的中风病古籍医案知识表示与古今双模知识图谱构建研究[J].中国中医药信息杂志,2024,31(11):24-32.
[10] 李贺,等.中医古籍数字化平台服务质量评价与提升策略研究[J].现代情报,2023,43(11):58-72.
[11] 吴其濬.植物名实图考[M].北京:中华书局,2018. |
|
|
 |
论文收录证明 / 文献检索报告
Document Retrieval Certificate / Proof of Publication Indexing
作者贡献声明 / 贡献确认书
Author Contribution Statement / Certificate of Authorship Contribution
同行评审报告 / 评审意见
Peer Review Report / Peer Review Comments
利益冲突
Conflict of Interest
作者声明不存在任何利益冲突。
The author declares no conflict of interest.
|
|
