数据挖掘技术在藏药方剂配伍理论研究中的应用进展

Research Progress on Compatibility Theory of Tibetan Medicine Formulas Based on Data Mining Technology

近年来,基于数据挖掘技术的藏药方剂配伍理论研究成为藏医药领域的研究热点,对藏药方剂的开发和名老藏医经验的继承发挥着重要作用。本文就基于文献和临床数据的方剂配伍理论研究现状进行系统梳理,发现目前使用的数据挖掘方法主要包括频数统计、关联规则、“味性化味”矢量结构模型、聚类分析、偏序结构、BP神经网络等方法,研究主要集中于治疗心血管疾病、胃肠疾病、肝胆疾病等藏医优势病种的方剂配伍规律研究,并从“味、性、效”层面探讨其配伍理论。在挖掘藏医药学术思想和配伍规律层面已经取得多项成果,为藏医药信息化、国际化奠定了基础。同时,也存在名词术语存在不统一性、方剂筛选缺乏严谨性、配伍理论研究缺乏深度与广度等科学问题待进一步攻克。最后根据目前的现状和存在的问题提出了今后发展的方向及研究思路。

Data mining is a process of discovering interesting and useful patterns and relationships in large volumes of data.In recent years,the research on the composition theory of Tibetan formulas based on data mining technology has becoming a research hotspot in the field of Tibetan medicine,which plays an important role in the inheritance of Tibetan famous and the experience of prescriptions great Tibetan doctors.In this paper,reviewed the researches of compatibility theory of Tibetan formulas based on medical texts and clinical data.Frequency,cluster,BP neural network were most likely used to analyze the compatibility theory of Tibetan formulas.The researches mainly focused on the compatibility theory of Tibetan formulas from the ancient texts,and clinical data of some Dominant diseases of Tibetan medicine,such as gastrointestinal diseases and angiocardiopathy.The compatibility theory were also discussed at "Ro,Nus-pa,Yontan" theory level.The disadvantages and problems,such as the inconsistency of terms,the rigor lacks of Selecting of formulas,and not enough in-depth were typed out as well.At the end,the research direction and methods were discussed.