Construction of Chinese Medicine Prescription Data Mining Based on Association Rules of the System

Abstract

Background and objectives: Traditional Chinese medicine (TCM) prescriptions are complex compositions according to certain compatibility rules to treat different diseases. Therefore, understanding the superposition, inhibition, and interaction between the effective components is essential. This study aimed to use modern information technology and the established TCM databases to analyze and explore the effectiveness of TCM compositions in terms of prescription and regulation.Methods: We adopted data mining technology to analyze the composition rules and active components among the spleen and stomach prescriptions that mainly includes the core algorithms of the apriori method and the frequent pattern (FP)-growth method.Results: In the present work, a new algorithm was developed that uses the modified FP-tree and the head table structure, producing only the FP-tree once and the head table structure at each recursion. Conclusions:The new algorithm can be used to obtain the same results as the original algorithm of frequent item set mining, but it is at least two times faster than the FP-growth method. Thereafter, it can be used to screen and preprocess other agent data as well as to create a database by using the improved algorithm, thus establishing a simulation system for effective component and computer-composing principles of TCM that is characterized by frequent item sets, association rules, and the clustering analysis algorithm and consequently undergoes analysis for the frequency of medicinal herbal ingredients, frequency of symptoms with certain compounds, medicine-sickness connection, and pharmaceutical composition clustering to determine the related effective ingredients to treat disorders of the spleen and stomach.