Study on Knowledge Representation System of Ancient Medical Cases of Stroke and Construction of Dual-mode Knowledge Graph Research Based on the Knowledge Element Theory
10.19879/j.cnki.1005-5304.202404191
- VernacularTitle: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
- Author:
Jian CHEN
1
;
Feng YANG
;
Qiaosheng REN
;
Ying LI
;
Xiaohua TAO
;
Tao LUO
;
Ying GAO
;
Jingling CHANG
Author Information
1. 北京中医药大学东直门医院,北京 100700
- Keywords:
knowledge element theory;
knowledge graph;
knowledge representation system;
stroke;
ancient medical cases
- From:
Chinese Journal of Information on Traditional Chinese Medicine
2024;31(11):24-32
- CountryChina
- Language:Chinese
-
Abstract:
Objective To explore the research path to construct knowledge representation,knowledge analysis,knowledge presentation and knowledge reasoning of ancient stroke medical cases based on knowledge element theory and knowledge graph.Methods A multi-layered knowledge representation system of medical cases was constructed,which included"knowledge body-knowledge element-semantics and semantic relationships",and knowledge element indexing system was used to parse the text into structured data.Combined with the attribute graph model for graph design,Neo4j was used to build a knowledge graph,and visual presentation and analysis were performed.Results A total of 517 medical cases from 86 ancient books were included,forming a knowledge representation system including 2 types of knowledge bodies,19 types of knowledge elements,27 types of semantics and 9 types of semantic relationships.The original knowledge graph had 19 node types(8 virtual nodes)and 15 relationship types,totaling 6363 nodes and 14697 relationships;the integrated knowledge graph had 11 node types and 11 relationship types,totaling 1229 nodes and 11456 relationships.Hierarchical subgraph visualization and node connectivity analysis methods completed implicit knowledge exploration and reasoning.Conclusion The knowledge representation system can guide the structured expression and semantic level analysis of ancient text,and provide systematic data support for the construction of knowledge graphs.The original knowledge graph of ancient and modern dual-mode knowledge graph focuses on the visual presentation of complex diagnosis and treatment ideas,which is conducive to improving the efficiency of medical cases review;integrating the knowledge graph to build a scale-free network and using group cases as the object can realize the tacit knowledge of stroke from a macro dimension excavation.
