Application Of Data Mining Technology To Traditional Chinese Medicine Formula Composition

TCM Compound Prescription (i.e. TCM Prescription) plays an important role in TCM science. Hundreds of thousands of TCM Compound Prescriptions accumulated over thousands of years and many established prescription databases has become the most precious resource and wealth of TCM.Under the prerequisite of dialectics, guided by the TCM theory, TCM Prescription bases on the medicinal theory and follows TCM Prescription Composition rules to the "principal-assistant-adjuvant-guiding(PAAG)"composition., aiming at the key point of diseases. Thus against the diseases, drugs has become the "Army with strict discipline" so as to get comprehensive accommodation. "Prescription composition" means compounding the medcines according to specific rules, making the goal of "compounding all the medcines and putting each medicine in the proper sites", and "adjusting the more and the less", implementing Therapy System for comprehensive accommodation against disease. Composition is the core of TCM Prescription, also the key problem in prescription research.Applying the research results of data mining and related technology to the intelligent analysis of historical TCM Prescription Composition data, this paper studies the basic rules between the TCM diseases and Prescription Compositions and provides theoretical support for the effective prescription simplification and reasonable composition, which is one of the important projects for TCMinformationalization.With the promotion of the TCM Informationalization Process and unremitting efforts of people in TCM and other related fields, TCM Prescriptions has been generalized into several prescription database. TCM prescriptions amount to several hundred thousand, among which the ancient prescription database contains 80 thousand prescriptions. Therefore, the paper emphasizes the ancient prescription database.It contributed to the areas as follows:1. advocate and implement the Apriori Algorithm, also achieve a large amount of experimental results based on three dataset(including TCM ancient prescriptiondatabase etc.);2. establish a integrating platform for TCM prescription data mining;3. apply the machine learning technology to the standardization, index and structurization of effective field in Chinese Medicine table, classify, index and structurize the compound effective fields in TCM ancient prescription database;4. classify, correct, standardize and index the compound prescription ingredients in TCM ancient prescription database for the structurization;5. develop research system for the TCM Prescription Composition rules according to the concept of prescription tree.The paper is supported by project fund as follows: the constructing and sharing of TCM scientific and technical information database (2002DEA30042), basic research in key scientific problem of 973 project (2003CB317006) .