A Research on the Compatibility Principles of Formulas for Spleen Stomach Health from Thousand Golden Prescriptions Based on Clustering Mining and Association Rule Mining Technology

摘要

As the earliest clinical medical encyclopedia ever found in China, Thousand Golden Prescriptions describes a great many health preserving formulas of the traditional Chinese medicine (TCM). From TCM's point of view, many symptoms are correlated, and hence the corresponding curative formulas are also correlated. It is much scientific and subjective to do research about correlation of herbs from the aspect of attribution of herbs such as character, taste and its function by using the method of clustering mining. This paper screens and collates the spleen & stomach health preserving formulas from Thousand Golden Prescriptions to develop a formula database containing such fields as herb, formula name and indication, so as to analyze herb application frequency and carry out the mining of herb-application association rules, herb-classification association rules and clustering mining. This paper leverages the data mining approach to discuss the herb application rules of spleen & stomach health preserving formulas from multiple perspectives, and eventually obtains the commonly-used herb groups and herb pairs through association rule mining which have been confirmed by pharmacology experts. This study draws to a conclusion that the compatibility rules of the spleen & stomach health preserving formulas as described in Thousand Golden Prescriptions can be discovered by applying data mining.