Building DartSpora Data Mining Platform And Chinese Medicine Formula Application

In the information era, with the development of world-wide-web, people accumulate magnanimity of data. There are a lot of important information under cover the proliferate data.Inorder to make use of data, people want to analysis the data on a higher view, that's why data mining is becoming more and more important. Data mining is a complex and demanding task. While a large number of methods have been established for numerous problems, many challenges remain to be solved. The rapidily changing requirement of data mining needs maximal re-use and innovative combinations of existing methods, as well as simple and quick integration of new ones.With the promotion of the TCM Informationalization Process, data mining is more and more wildly employed in TCM.With unremitting efforts of people in TCM and other related fields, several hundred thousands of TCM Prescriptions has been generalized into several prescription database. Among this TCM Prescriptions the ancient prescription database contains 80 thousand prescriptions. These data serve as the foundation for TCM Prescription Composition rules research.In this thesis, we design and implement the DartSpora data mining platform, cooperate with China Academy of Chinese Medical Sciences, apply DartSpora platform to TCM Prescription, study the TCM Prescription Composition rules.We focus on the following issues in this paper.1. Using Google Web Toolkits, GWT-EXT opensouce Ajax framework, andworld wild popular Data mining tool Rapid Miner to design and implement DartSpora Data Mining platform. It including experiment manage module, DartGrid module, Data base connection manage module, and user manage module.2. Inorder to provide access to distributed database which is composited basedon semantic, we integrate Dart Grid with DartSpora. User can get data they really need with their domain knowledge, without understanding the complex structure of distributed database. 3. Contrapose the character of TCM Prescription data, we design preprocess operator based on user-defined rules, speedup the efficiency and configuration of TCM Prescription data preprocess.4. Improving traditional Apriori algorithm by introducing data weight, and develop Weighted-Apriori algorithm. Using Web Well-known and history literature authorization as weight, apply Weighted-Apriori into spleen-stomach TCM Prescription Composition rules research. Transplant algorithms developed by CCNT Lab into DartSpora Platform.5. Apply DartSpora platform into TCM Prescription research.Here we show 3application cases: TCM Prescription data preprocess under user-defined rules, viral myocarditis TCM Prescription max-pattern mining, TCM Prescription Weight Frequent Pattern Mining based on data reliability.