Application And Research Of Association Rulesin Respiratory Medicine Clinic Data

Association Rule Analysis is one of important technology in the Data Mining Technology. It generates the important relationships among the data by analyzing the relevance among itemsets within the data set. Generally, the expression of the rules is concise and easy to explain and understand. The research of mining association rules from a large database has become a hot topic in the research field of data mining in recent years.Hospital Information System (HIS) provides a convenient way of information management and on-line operation services for hospital management and medical treatment activities. During the use of HIS, a large number of medical information about the medical activities is recorded in fidelity. If some potential association rules among the illnesses can be extracted from the accumulated medical diagnosis data, and after the information discriminating and demonstration, valuable knowledge for diseases diagnosis and prevention can be extracted. However, the technologies of data processing and analyzing adopted by most of the HIS of hospitals have failed to fully extract the hidden knowledge in the huge amounts of data. Therefore, how to introduce the data mining techniques, such as association rule, into the HIS and how to make full use of the hospital diagnosis data to improve accuracy of disease diagnosis and enhance feasible of disease prevention are becoming a critical problem.On the basis of research of various kinds of data mining technologies, and the diagnosis and treatment data provided by the traditional Chinese respiratory department of Qingdao Haici hospital, this thesis focuses on the application of the association rule algorithm in respiratory department diagnosis and treatment data. The main research contents include:(1) Researching and analyzing the present domestic and overseas condition of data mining technology applied in the medical field.(2) Making an intensive study of processing mechanism of Apriori algorithm. By considering the real problems that the hospital diagnosis and treatment data are almost updated quickly and the clinic doctors mostly need to make rapid analysis of the relevance among the specific diseases, this thesis provides an improved Apriori algorithm.(3) Elaborating the application process of improved Apriori algorithm applied in the respiratory medicine diagnosis and treatment data of Haici hospital. At present, the Association Rule Analysis System developed by this thesis has been tested in the hospital outpatient service. Disease association rules which were generated by this System have provided reliable data support for the doctor’s clinical diagnosis.