The Research & Application Of Mining Association In TCM Diagnosis

How to take full advantage of the valuable medical information resources to offer the scientific decisions for diagnosis and treating of disease and promote the medical research has already become the focus that people have paid close attention to as the information capacity of the hospital database is expanding constantly. Especially the traditional Chinese medicine in our country, it is very difficult to keep a large amount of the valuable Chinese medicine experience only by simple text with age of old docter of TCM of our country.Data mining is the course of finding the hidden, unknown, interesting information from the database or the data warehouse and it is a kind of interdisciplinary subjects which is rised by the fusion of artificial intelligence and database technology in recent years, it is devoted to find the knowledge or law about the essence and development trend of things which is implied in the data and offers the support for the decisions of the experts.Data mining is a kind of the research directions which has broad prospect and is full of challenging in the field of traditional Chinese medicine. However, so far, the academic thought and the research of the dialectical experience of the famous old docters of TCM still remain at the stage of sorting out and summing up and there are certain subjective compositions. The main purpose of this subject is to mine the laws of using medicine of the typical cases by data mining and collecting numerous medicine cases of the typical cases from the inconsistent or even intact database to offer the references for the clinical treatments of Chinese medicine, the teachings of traditional Chinese medicine and the reseachs of Chinese medicine.The concept, technology, current research situation and basic theories of association rules are introduced in this thesis, and the classical mining algorithm is analyzed. Finally, frequent pattern growth algorithm(FP-growth) to mime association rules in the knowledge and the experiences of Chinese medicine is applied, which is faster an order of magnitude less efficient than Apriori algorithm because of avoiding generating candidate itemsets. By combining the real medical materials - chronic gastritis Chinese medicine materials, the thesis utilizes FP-tree algorithm to mine association rules in Chinese medicine data, does the research from analyzing the materials of the chronic gastritis patients and mines the association rules of symptom and dialectic, dialectic and prescription, symptom and prescription, analyzes the association rules of being mined. This thesis makes a try of the data mining in the applications of Chinese medicine.