摘要
量化中药功效相似度是方剂功效规约中一个基本而重要的问题。采用数据挖掘技术对该问题进行了探讨,即从现有的药物记载功效中挖掘出功效之间的相似关系。该方法首先通过关联规则算法获得强双向关联功效对,以去除相关程度较小的功效对,然后采用SimRank迭代算法求取功效之间的相似。基于这样的思路,获得了比较合理与客观的功效相似度。与人工方法计算功效相似度相比,既大大降低了工作量,又减少了主观性。为方剂功效规约研究打下了基础。
Quantifying the similarity between effects of Chinese Traditional Medicine （TCM） is a fundamental and important issue for effect summarization of prescription. In this paper, Data Mining technologies are used to explore this issue. First, pairs of effects with strong association are got by association rule algorithm so as to remove the effects pair with weak association, then, SimRank method is adopted to compute the similarities between each pair of effects. Based on this method, reasonable and objective quantified similarities are obtained. Compared with manual computation, the proposed method can not only reduce the computing workload but also the subjectivity. This research establishes the foundation for automatic summarization analysis of prescription in TCM.
出处
《辽宁中医杂志》
CAS
北大核心
2011年第3期402-405,共4页
Liaoning Journal of Traditional Chinese Medicine
基金
"十一五"国家科技支撑计划(2007BAI10B06)
关键词
中药功效
相似度
关联规则
SIMRANK
effects of traditional chinese medicine
similarity
association rule
SimRank
