Discussion on Classical Formula Artificial Intelligence Research Path Based on Graph Theory
-
摘要: 分析了经方理论与图论的相关性, 提出基于图论的经方人工智能(AI)研究路径，从逻辑推理和命题逻辑角度分析图论与经方理论、思维的相关性, 提出可行的研究路径。图论的应用能解决经方智能化研究中的知识表示问题, 图的属性和度量方法有助于从中医思维出发进行知识发现, 图的矩阵表示和图连通性能为经方智能辅助诊疗模型融入领域知识, 提升模型效率。图论可为经方AI研究提供理论指导, 基于图论的知识图谱等研究技术可为经方AI研究中的难题提供解决方案。Abstract: This article analyzes the correlation between classical formula theory and graph theory, puts forward the research path of classical formula AI based on graph theory, introduces the related contents of graph theory and classical formula AI research, analyzes the correlation between graph theory and classical formula theory and thinking from the perspective of logical reasoning and propositional logic, and proposed feasible research path. The application of graph theory can solve the problem of knowledge representation in the intelligent research of classical formula. The graph properties and their measurement methods are helpful for knowledge discovery from the perspective of TCM thinking. The matrix representation and connectivity of graph integrate the domain knowledge of classical formula into AI-assisted diagnosis and treatment model and improve the model efficiency. Graph theory can provide theoretical guidance for the research of classical formula AI, and knowledge graph based on graph theory can provide solutions for the difficult problems in the research of classical formula AI.
-
图 3 “症-证图”与“药-证图”相应(举例)
Figure 3. Correspondence between symptom-syndrome diagram and medicine-syndrome diagram (example)
图 4 通过“症-证图”分类判断六经病证
Figure 4. Determining the syndromes of the six meridians through the classification of symptom-syndrome diagram
图 5 通过“药-证图”与“症-证图”匹配筛选“方证相应”路径
Figure 5. Filtering the formula-syndrome correspondence path by matching the medicine-syndrome diagram with the symptom-syndrome diagram
图 6 六经分类模型与GCN提取图的空间特征示意图
注: Gi为医案映射到知识图谱中获得的子图, Gj为其中一个六经病证标准子图, l为L的列向量, U是以为单位特征向量的矩阵, λ为特征值, f为图中任意N维向量。
Figure 6. Schematic diagram of spatial features between the six-meridian classification model and GCN extraction diagram
-
[1] 白景瑄, 胡晓娟, 许家佗. 基于复杂网络技术的中医诊疗规律研究进展[J]. 时珍国医国药, 2020, 31(9): 2207-2209. https://www.cnki.com.cn/Article/CJFDTOTAL-SZGY202009051.htmBAI JX, HU XJ, XU JT. Research progress of TCM diagnosis and treatment law based on complex network technology[J]. Lishizhen Med Mater Med Res, 2020, 31(9): 2207-2209. https://www.cnki.com.cn/Article/CJFDTOTAL-SZGY202009051.htm [2] 王曦廷, 卢涛. 中医药认知计算: 概念、框架与路径[J]. 中华中医药杂志, 2022, 37(1): 35-40. https://www.cnki.com.cn/Article/CJFDTOTAL-BXYY202201006.htmWANG XT, LU T. Cognitive computation of Chinese medicine: Concept, framework and pathway[J]. China J Tradit Chin Med Pharm, 2022, 37(1): 35-40. https://www.cnki.com.cn/Article/CJFDTOTAL-BXYY202201006.htm [3] 卜月华, 王维凡, 吕新忠. 图论及其应用[M]. 2版. 南京: 东南大学出版社, 2015: 2.BU YH, WANG WF, LYU XZ. Graph Theory and Its Application[M]. 2nd ed. Nanjing: Southeast university press, 2015: 2. [4] FORTUNATO S. Community detection in graphs[J]. Phys Rep, 2010, 486(3/4/5): 75-174. [5] 许进, 张雷. DNA计算机原理、进展及难点(Ⅰ): 生物计算系统及其在图论中的应用[J]. 计算机学报, 2003, 26(1): 1-11. https://www.cnki.com.cn/Article/CJFDTOTAL-JSJX200301000.htmXU J, ZHANG L. DNA computer principle, advances and difficulties (Ⅰ): Biological computing system and its applications to graph theory[J]. Chin J Comput, 2003, 26(1): 1-11. https://www.cnki.com.cn/Article/CJFDTOTAL-JSJX200301000.htm [6] STAVRAKAS V, MELAS IN, SAKELLAROPOULOS T, et al. Network reconstruction based on proteomic data and prior knowledge of protein connectivity using graph theory[J]. PLoS ONE, 2015, 10(5): e0128411. doi: 10.1371/journal.pone.0128411 [7] KEMPE D, KLEINBERG J, TARDOSE. Maximizing the spread of influence through a social network[C]//Proceedings of the ninth ACM SIGKDD international conference on Knowledge discovery and data mining. August 24-27, 2003, Washington, D.C. . New York: ACM, 2003: 137-146. [8] 樊新荣. 《伤寒论》三阳三阴病证的证素辨证研究[J]. 湖南中医药大学学报, 2015, 35(1): 41-43. https://www.cnki.com.cn/Article/CJFDTOTAL-HNZX201501013.htmFAN XR. Syndrome-element study of the three-Yang and three-Yin differentiation in "treatise on febrile diseases"[J]. J Tradit Chin Med Univ Hunan, 2015, 35(1): 41-43. https://www.cnki.com.cn/Article/CJFDTOTAL-HNZX201501013.htm [9] 贾春华. 基于命题逻辑的《伤寒论》方证理论体系研究[D]. 北京: 北京中医药大学, 2006.JIA CH. Research on the theoretical system of prescription and syndrome in treatise on febrile diseases based on propositional logic[D]. Beijing: Beijing University of Chinese Medicine, 2006. [10] 王鑫, 邹磊, 王朝坤, 等. 知识图谱数据管理研究综述[J]. 软件学报, 2019, 30(7): 2139-2174. https://www.cnki.com.cn/Article/CJFDTOTAL-RJXB201907016.htmWANG X, ZOU L, WANG CK, et al. Research review on knowledge graph data management[J]. J Softw, 2019, 30(7): 2139-2174. https://www.cnki.com.cn/Article/CJFDTOTAL-RJXB201907016.htm [11] 王松, 李正钧, 杨涛, 等. 中医药知识图谱研究现状及发展趋势[J]. 南京中医药大学学报, 2022, 38(3): 272-278. doi: 10.14148/j.issn.1672-0482.2022.0272WANG S, LI ZJ, YANG T, et al. Current status and development trend of knowledge graph research in traditional Chinese medicine[J]. J Nanjing Univ Tradit Chin Med, 2022, 38(3): 272-278. doi: 10.14148/j.issn.1672-0482.2022.0272 [12] 虞红蕾, 曹灵勇, 瞿溢谦, 等. 消渴病经方知识图谱构建与知识发现[J]. 浙江中医药大学学报, 2022, 46(2): 113-119, 125. https://www.cnki.com.cn/Article/CJFDTOTAL-BHON202202001.htmYU HL, CAO LY, QU YQ, et al. Knowledge graph construction and knowledge discovery of classic prescriptions of diabetes disease[J]. J Zhejiang Chin Med Univ, 2022, 46(2): 113-119, 125. https://www.cnki.com.cn/Article/CJFDTOTAL-BHON202202001.htm [13] SAXENA A, CHAKRABARTI S, TALUKDAR P. Question answering over temporal knowledge graphs[EB/OL]. [2023-03-01]. https://arxiv.org/abs/2106.01515 .[14] 魏泽林, 张帅, 王建超. 基于知识图谱问答系统的技术实现[J]. 软件工程, 2021, 24(2): 38-44. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGC202102009.htmWEI ZL, ZHANG S, WANG JC. Implementation of question answering based on knowledge graph[J]. Softw Eng, 2021, 24(2): 38-44. https://www.cnki.com.cn/Article/CJFDTOTAL-ZGGC202102009.htm [15] 郭超峰, 施学丽. 基于复杂网络理论的经方"方证相应"研究: 以桂枝汤为例[J]. 辽宁中医杂志, 2013, 40(3): 438-440. https://www.cnki.com.cn/Article/CJFDTOTAL-LNZY201303024.htmGUO CF, SHI XL. A study on the correspondence of prescriptions and syndromes in classical Chinese medicine based on complex network theory: Taking Guizhi Decoction as an example[J]. Liaoning J Tradit Chin Med, 2013, 40(3): 438-440. https://www.cnki.com.cn/Article/CJFDTOTAL-LNZY201303024.htm [16] 刘礼荣. 基于复杂网络的《金匮要略》病传规律研究[D]. 杭州: 浙江中医药大学, 2019.LIU LR. Study on the law of disease transmission of synopsis of golden chamber based on complex network[D]. Hangzhou: Zhejiang Chinese Medical University, 2019. [17] BLONDEL VD, GUILLAUME JL, LAMBIOTTE R, et al. Fast unfolding of communities in large networks[J]. J Stat Mech, 2008, 2008(10): P10008. [18] BAI YS, DING H, BIAN S, et al. SimGNN: A neural network approach to fast graph similarity computation[EB/OL]. [2023-03-01]. https://arxiv.org/abs/1808.05689 .[19] SHUMAN DI, NARANG SK, FROSSARD P, et al. The emerging field of signal processing on graphs: Extending high-dimensional data analysis to networks and other irregular domains[J]. IEEE Signal Process Mag, 2013, 30(3): 83-98. [20] 杨涛, 朱学芳. 中医辨证智能化研究现状及发展趋势[J]. 南京中医药大学学报, 2021, 37(4): 597-601. doi: 10.14148/j.issn.1672-0482.2021.0597YANG T, ZHU XF. Discussion on the status and development trend of research on intellectualization of Chinese medicine syndrome differentiation[J]. J Nanjing Univ Tradit Chin Med, 2021, 37(4): 597-601. doi: 10.14148/j.issn.1672-0482.2021.0597 [21] 刘震, 赵壮, 林祺, 等. 基于图论的智能针灸机器人取穴原理研究[J]. 世界中医药, 2018, 13(8): 1992-1996. https://www.cnki.com.cn/Article/CJFDTOTAL-SJZA201808043.htmLIU Z, ZHAO Z, LIN Q, et al. Research on the principle of intelligent acupuncture and moxibustion robot acupoint selection based on graph theory[J]. World Tradit Chin Med, 2018, 13(8): 1992-1996. https://www.cnki.com.cn/Article/CJFDTOTAL-SJZA201808043.htm [22] XU Q, GUO Q, WANG CX, et al. Network differentiation: A computational method of pathogenesis diagnosis in traditional Chinese medicine based on systems science[J]. Artif Intell Med, 2021, 118: 102134. [23] 尹丹, 周璐, 周雨玫, 等. 中医经方知识图谱"图搜索模式"设计研究[J]. 中国中医药信息杂志, 2019, 26(8): 94-98. https://www.cnki.com.cn/Article/CJFDTOTAL-XXYY201908019.htmYIN D, ZHOU L, ZHOU YM, et al. Study on design of graph search pattern of knowledge graph of TCM classic prescriptions[J]. Chin J Inf Tradit Chin Med, 2019, 26(8): 94-98. https://www.cnki.com.cn/Article/CJFDTOTAL-XXYY201908019.htm -
下载:
点击查看大图
图(7)
计量
- 文章访问数: 190
- HTML全文浏览量: 23
- PDF下载量: 23
- 被引次数: 0
