论文部分内容阅读
目的:应用关联规则对近20年中国知网数据库中关于糖尿病性周围血管病变的文献进行数据挖掘,分析其用药规律。方法:通过主题词检索方式,检索1995年至2015年中国知网数据库,通过纳入和剔除标准进一步筛选文献,利用Note express 3.0和Epidata 3.1软件纳入文献信息,借助SPSS Modeler14.1对清洗转换后的数据进行频数统计、关系网络模型、Apriori关联规则算法等数据挖掘分析。结果:共纳入符合要求的文献28篇,得到方剂33首;通过频数分析得出用药频次最高的依次是当归、黄芪、川芎、桂枝、鸡血藤、牛膝、地龙;频率大于25%,药物功效排在前两位的分别是补虚药、活血化瘀药;归经排在前两位的分别是肝经、脾胃经;关联分析得到符合最小支持度和置信度的药对和药组分别有9条和4条。结论:益气养血、活血通络法是治疗糖尿病性周围血管病变的首选。
OBJECTIVE: To use the association rules to mine the literature about diabetic peripheral vascular disease in CNKI database in the past 20 years, and to analyze its medication rules. Methods: Retrieve the CNKI database from 1995 to 2015 by using the keyword search method, further screening the documents by inclusion and exclusion criteria, incorporating the document information with Note express 3.0 and Epidata 3.1 software, and using SPSS Modeler 14.1 to clean the converted Data frequency statistics, relational network model, Apriori association rules algorithm data mining analysis. Results: A total of 28 literatures were selected and 33 prescriptions were obtained. According to frequency analysis, the highest frequency of medication was Angelica, Radix Astragali, Chuanxiong, Guizhi, Millettia, Achyranthes, earthworm; the frequency was more than 25% , Drug efficacy in the top two are the tonic drugs, blood circulation drugs; ranked in the top two are the liver, spleen and stomach; correlation analysis obtained with the minimum support and confidence of the drug and Drug groups were 9 and 4 respectively. Conclusion: Yiqi Yangxue and Huoxuetongluo are the first choice for the treatment of diabetic peripheral vascular disease.