摘要
为明确中医治疗抑郁症用药规律,融合Apriori优化算法与Relim算法,采用数据挖掘技术进行分析。针对传统Apriori算法频繁扫描数据库从而生成大量候选项集的缺点,改变其原有剪枝方式以减少扫描次数。将改进后的Apriori算法与无需产生候选项集的Relim算法就中医治疗抑郁症的方剂数据进行关联规则分析,并绘制两个算法时间效率图。结果发现,两种算法在挖掘药物频繁项集与关联规则的结果基本相同,通过分析发现,中医常以疏肝、理气、补肾、滋阴等药物为主治疗抑郁症。改进后的Apriori算法可降低数据库扫描次数,较传统Apriori算法运行效率有所提高,Relim算法在空间利用率和时间执行率上均略优于改进后的Apriori算法。两种算法挖掘结果体现出中医治疗抑郁症注重疏肝理气、补肾滋阴、调理气血等特点。基于关联规则的方法可作为中医用药规律分析的重要工具。
Improved Apriori Algorithm and Relim Algorithm were combined and data mining technology was employed to identify the medication rules for depression.Aiming at the shortcomings of frequently scanning the database in the traditional Apriori algorithm to generate a large number of candidate itemsets,the original pruning method was changed to reduce the number of scans.The improved Apriori algorithm and Relim algorithm without generating candidate itemsets were used to mine the prescription data of traditional Chinese medicine for treating depression and compare the performance of the two algorithms by time efficiency graphs.The two algorithms are basically the same in mining frequent itemsets and association rules of drugs.Through analysis,it is found that traditional Chinese medicine often treats depression by treating drugs such as soothing liver,regulating qi,nourishing kidney and nourishing yin.The improved Apriori algorithm can reduce the number of database scans,which improves the operating efficiency of the traditional Apriori algorithm.The Relim algorithm is slightly better than the improved Apriori algorithm in terms of space utilization and time execution rate.The mining results of the two algorithms both reflect the characteristics of traditional Chinese medicine in the treatment of depression,such as dispelling liver and regulating qi,nourishing kidney-Yin,and regulating qi and blood.The method based on association rules can be an important tool for the analysis of traditional Chinese medicine.
作者
王慧敏
龚庆悦
胡孔法
邵荣强
陈燕
WANG Hui-min;GONG Qing-yue;HU Kong-fa;SHAO Rong-qiang;CHEN Yan(School of Artificial Intelligence and Information Technology in Nanjing University of Chinese Medicine,Nanjing 210023,China)
出处
《软件导刊》
2020年第3期190-193,共4页
Software Guide
基金
国家自然科学基金项目(81674099)
江苏省中医药管理局项目(YB2017008)。
