Study On Data-Mining Technology And Its Application In Tranditional Chinese Prescription Analysis

Data mining is a new multi-disciplinary field, drawing work from areas including statistics, machine learning, database technology, artificial intelligence, pattern recognition and so on. The primary work of data mining is to extract useful but implicit information from enormous data. Data of Traditional Chinese Medicine (TCM) is a unique treasury of China, but the data is not well dealt with and utilized, Mining patterns and knowledge from data of TCM is necessary and important.This thesis, combined with the feature of Traditional Chinese Prescription (TCP), studies several primary technologies of data mining and brings forward appropriate methods of data mining, and analyzes the compatibility of TCP by using these methods. At the beginning the thesis present, the idea of linear association rules. Association rules is an important technology of data mining, but the traditional association rules only deal with whether there are some relations between two variables, not about the quantities of variables. This thesis brings forward linear association rules combined with practical application so as to deal with TCP and then make medicine have some relations with their dose.Bring forward the method of Jaccard coefficient method to calculate the distance of the clustering between two variables. Clustering of Traditional Chinese Medicine is based on binary variables of characteristics and taste of TCM, but general methods of calculating distance such as Euclidean distance cannot show the characteristics of binary variables effectively when processing them, but the method of Jaccard Coefficient can do well, in addition, the method is simple enough to put into practice.After the dose of TCM is transformed from text to numeric, the mean, variance of dose and the percent of medicine in each dosecan also be calculated with the normative and coherent data.Finally, an introduction of the design of analysis system of TCP is given, which includes the preprocessing of TCP data and the functions of some modules in the system.