基于古籍文字缺省补全策略的胡希恕经方知识图谱构建的方法学研究  

Methodological study on the construction of HU Xishu Jingfang knowledge graph based on the ancient text default completion strategy

作　　者：孙超[1] 佟旭[2] SUN Chao;TONG Xu(School of Traditional Chinese Medicine,Capital Medical University,Beijing 100069,China;Institute of Basic Theory of Chinese Medicine,China Academy of Chinese Medical Sciences,Beijing 100700,China)

机构地区：[1]首都医科大学中医药学院,北京100069 [2]中国中医科学院中医基础理论研究所,北京100700

出　　处：《中华中医药杂志》2025年第3期1247-1250,共4页China Journal of Traditional Chinese Medicine and Pharmacy

基　　金：中国中医科学院科技创新工程项目(No.CI2021A00409);首都卫生发展科研专项“青年优才”项目(No.首发2022-4-4282);北京市教委科技计划一般项目(No.KM202110025021);北京中医药“薪火传承3+3工程”崔锡章中医文化传承工作室(No.京中医科字【2018】119号);首都医科大学第二课堂创新项目(No.D2KT2023183)。

摘　　要：经方蕴含丰富的辨证论治经验,但由于其“寓理于事,因事明理”表述方式,使我们在构建经方知识图谱、探索计算机辅助诊疗时,必须“于无字处求文”,解决其由于文字缺省导致的信息不全,大量知识隐而未现的问题。基于此,文章提出“症状体征串”所代表的“病机”是“无字”之处所求的关键,以此为基础构建补全策略,以胡希恕先生对经方病机解读为例,构建推理规则,搭建基于胡希恕先生学术思想的仲景经方知识图谱。最终构建的知识图谱可自动补全经方文字缺省的信息,基于胡老诊疗思维进行“症状-方证”推理。本研究有助于中医经典的传承和学术经验的推广,未来可进一步补充多家论述形成经方信息学大数据,为多元化、模糊化的中医诊疗思维下探索古籍辅助临床诊断奠定基础。The ancient books of Jingfang contain rich experience in diagnosis and treatment.However,due to its expression of'embed principles in narratives,elucidate truths through events',when constructing the Jingfang Knowledge Graph and exploring computer-aided diagnosis and treatment,we must'searching for text in the absence of words'to solve the problem of incomplete information and a large amount of knowledge hidden due to missing words.Based on this,we proposed that the‘disease mechanism'represented by the'string of symptoms and signs'is the key to what is sought in the‘wordless'place,and use this as the basis for constructing a completion strategy,and use Mr HU Xishu's interpretation of the scriptures as an example to construct a strategy for the disease mechanism of the scriptures.We constructed inference rules to build a knowledge graph of Zhongjing's Jingfang based on Mr HU's academic thought,and explored the logic of Chinese medicine thinking in the prescription.The final knowledge graph can perform the basic reasoning of symptom-formula evidence and automatically complete the default information in the text of the scriptures based on HU's academic thought.This study contributes to the inheritance of traditional Chinese medicine(TCM)classics and the promotion of academic experience,and in the future,it can further supplement multiple expositions to construct the big data of scripture informatics,which lays the foundation for exploring the clinical diagnosis assisted by ancient books under the diversified and fuzzy TCM diagnosis and treatment thinking.

关 键 词：中医古籍 经方 胡希恕 知识图谱 知识补全 

分 类 号：R249[医药卫生—中医临床基础]