传统中医药作为中华历史的瑰宝,在经历时间淘洗后沉淀下大量的珍贵经验和价值数据。数据挖掘和复杂网络作为数据处理及知识发现的有效手段,分别从统计性和复杂性角度描述数据,并被广泛应用于现代中医药学研究中。数据挖掘偏向于发掘表层统计规律,缺乏对于系统内在机制的深入讨论;复杂网络主要通过网络角度描述系统结构,长于在功能复杂性中发掘普适规律,但在大数据处理中存在局限。分析了数据挖掘及复杂网络在中医药领域的应用及各自存在的问题,并给出2种模式整合的方法及框架,以脑血管疾病中医用药为例,构建其中医用药属性拓展网络,并结合改进的重叠社区发现(COPRA)算法挖掘中医用药重叠社团,探讨中药方剂配伍规律。该方法可在原有基础上有效发掘非频繁项集的药性关联规律,为中药配伍提供依据。

The traditional Chinese medicine (TCM) as a treasure of the Chinese history, has deposited lots of precious experience and valuable data. As an effective mean of data processing and knowledge discovery, data mining and complex networks are used to describe the data from the statistics and complexity, and are widely used in the research of modern TCM. Data mining bias to explore the surface of the statistical law, lack for an in-depth discussion of the inner mechanism of the system; While the complex network mainly describes the system structure through the perspective of network and good at exploring the universal law in the functional complexity, but there are limitations in the large data processing. To analyze on the applications and problems of data mining and complex networks in the field of TCM, give two methods and frame of paradigms integration, construct Chinese medicine property development network in case of cerebral vascular disease, and excavate the regularity of compatibility of TCM combined with pharmaceutical association and improved COPRA algorithm. The experimental results show that the method can effectively explore the association rules of the infrequent itemsets, and provide the basis for the compatibility of Chinese herbal medicine.

陕西省科技厅课题(2014K14-02-02)