基于CART算法的医疗隐性知识挖掘研究——以中医医案为例

Research on Medical Tacit Knowledge Mining Based on CART Algorithm——Taking Traditional Chinese Medicine Cases as an Example

【目的/意义】中医医案中包含的隐性知识具有巨大的价值,通过数据挖掘技术对中医医案进行分析,挖掘临床医案中的隐性知识,有利于知名中医个人知识经验的传承和中医理论知识的积累和拓展。【方法/过程】收集中医医案中患者的主诉和现病史,并对主诉和现病史进行数据清洗,通过基于cart算法的数据挖掘技术挖掘中医医案中的隐性知识,探究病症与患者症候各个属性之间的关系。【结果/结论】本文以胃脘痛为例,发现了胃脘痛与患者症候各个属性之间的相关程度,为数据挖掘技术在中医医案隐性知识挖掘研究提供借鉴。【创新/局限】本文采用的cart算法判断胃脘痛与患者属性之间的相关性与医案中确诊结果进行比较,得出该方法相关性准确率高于ID3算法、C4.5算法和SLIQ算法,判断相关性最高。

【Purpose/significance】The tacit knowledge contained in Traditional Chinese Medicine medical records is of great value.The analysis of Traditional Chinese Medicine medical records through data mining technology and the mining of tacit knowledge inclinical medical records are beneficial to the personal knowledge and experience of well-known Traditional Chinese Medicine practi-tioners. Inheritance and accumulation and expansion of theoretical knowledge of Traditional Chinese Medicine.【Method/process】Col-lect the chief complaints and current medical history of patients in Chinese medical records, and clean the data of the chief complaintsand current medical history. Use the data mining technology based on the cart algorithm to mine tacit knowledge in traditional Chinesemedical records to explore the disease and patient symptoms the relationship between various attributes.【Result/conclusion】Takingstomachache as an example, this paper finds the correlation between stomachache and various attributes of patients’ symptoms, whichprovides a reference for data mining technology in the research of tacit knowledge mining in Traditional Chinese Medicine medical re-cords.【Innovation/limitation】The cart algorithm used in this paper to judge the correlation between gastric pain and patient attributesis compared with the diagnosis results in medical records, and the correlation accuracy of this method is higher than that of ID3 algo-rithm, C4.5 algorithm and SLIQ algorithm, and the correlation is the highest.