CAJ | 학술논문

基于中医传承辅助系统软件，采用改进的互信息法、复杂系统熵聚类、无监督的熵层次聚类等无监督数据挖掘方法，对630个藏兽医验方和纤毛婆婆纳的组方规律进行分析。结果表明，630个方剂中使用频次最高的单味药物、药物归类、四气、五味和归经分别是诃子、菊科、温性、苦味和肝脏；从26个含有“纤毛婆婆纳”处方中筛选得到2个新处方。
기우중의전승보조계통연건，채용개진적호신식법、복잡계통적취류、무감독적적층차취류등무감독수거알굴방법，대630개장수의험방화섬모파파납적조방규률진행분석。결과표명，630개방제중사용빈차최고적단미약물、약물귀류、사기、오미화귀경분별시가자、국과、온성、고미화간장；종26개함유“섬모파파납”처방중사선득도2개신처방。
To analyze the composing principles of the prescriptions of 630 Tibetan veterinary proved recipes and Veronica ciliata ,unsupervised data mining method such as revised mutual information,complex sys-tem entropy cluster and unsupervised hierarchical clustering were used based on the traditional Chinese medicine inheritance system software.The results showed that the single medicine,classification,four properties,five tastes channel-tropism with a higher use frequency in 630 proved recipes were Terminalia chebula,Asteraceae,warm,bitter and liver individually;and two new prescriptions were mined from the 26 proved recipes including Veronica ciliata .