摘要目的：探讨孟河名医马培之的用药规律。方法收集马培之治疗常见病的处方，在构建数据库的基础上，采用关联规则apriori算法、复杂系统熵聚类等数据挖掘方法，确定处方中药物使用频次、药物之间的关联规则和核心组合等。结果高频次药物包括茯苓、当归、白芍、陈皮、山药等，核心组合包括“生地黄-牡丹皮-麦冬”“续断-桑寄生-牛膝”“防己-独活-秦艽”“桑叶-蝉衣-桔梗”等，新处方包括“桑叶-蔓荆子-僵蚕-蝉衣-桔梗”“续断-桑寄生-牛膝-桑枝-五加皮”等。结论马培之用药经验丰富，多在经方基础上，灵活辨证，加减化裁。

abstractsObjective To analyze the principle of the drug use ofMenghe PhysiciansMa-Peizhi by using the Traditional Chinese Medicine(TCM)inheritance support system.Methods The prescriptions for the commonly encountered diseases that used byMa-Peizhi were collected to build a database, and analyze by the unsupervised data mining methods, such as apriori algorithm, entropy clustering complex systems, from which we could get the frequency of the drugs, the association rules between drugs, the core drug combinations, and so on.Results Based on the analysis of 745 prescriptions, the most frequently used drugs were tuckahoe, chiretta, paenoiae alba, dried orangepeel and dioscoreae. The core drug combinations were “radix rehmanniae recen- salivia chinensis-ophiopogon root”, “teasel root-viscum album-achyranthes”, “menispermaceae-heracleum hemsleyanum michaux-gentiana macrophylla”, and “mulberry leaf-periostracum cicadae-the root of balloon flower”. The new prescriptions were “mulberry leaf-viter rotundifolia-batryticated silkworm-periostracum cicadae-the root of balloon flower”, “teasel root-viscum album- achyranthes- ramulus mori- periplocae”, and so on.ConclusionMenghe PhysiciansMa-Peizhi was well experienced in treating the commonly encountered diseases by agile diagnosis and treatment, and addition or subtraction of changes based on the classical prescriptions.