Application And Improvement Discussion About Apriori Algorithm Of Association Rules Mining In Cases Mining Of Influenza Treated By Contemporary Famous Old Chinese Medicine

Objective: To investigate and discuss the application and improvement about Apriori algorithm of association rules mining (ARM) in cases mining of influenza treated by contemporary famous old Chinese medicine. Methods: We analyzed the basic principles, processes and algorithms about the Apriori algorithm of ARM, then applied this algorithm to the cases mining of influenza treated by contemporary famous old Chinese medicine. SPSS Clementine12.0 statistical software was used to mine the association rules between Etiology and traditional Chinese medicine (TCM), Syndromes and TCM, Symptoms and TCM. Then the disadvantage of the algorithm was summarized, and several improved high efficiency algorithms were discussed. Results: The Apriori algorithm of ARM could extract the association rules between common Etiology (Evil of wind, cold, heat and fire) and TCM, Syndromes (Syndromes of wind-heat or wind-cold invading exterior) and TCM, Symptoms (chills, fever, cough, nasal congestion, runny nose) and TCM. However, it became inefficient when the sample was large. Therefore, it was necessary to search for an improved Apriori algorithm in order to enhance the efficiency of ARM of Chinese medicine cases. Conclusion: The classic Apriori algorithm is useful to mine cases of influenza treated by contemporary famous old Chinese medicine, and the improved Apriori algorithm may help to improve the efficiency of mining.