Application Of Association Rules And Ontology In The Treatment Of Diabetes Mellitus

Database technology is widely used in many aspects such as business,government and scientific research with the rapid development of information technology. Researchers need to solve the problem is how to obtain useful knowledge for decision support from these large amounts of data which stored in the database.Association rule mining is proposed,of which the goal is to find out the potentially meaningful connections and rules from the mass of data.However, apriori as the traditional association rule mining representative,it is less efficient used in the practical application. Because that apriori algorithm is based on confidence and support, frequent itemsets are produced by connecting and pruning,and association rule is obtained from frequent itemsets. Therefore,improving the efficiency of traditional association rule algorithm is still a key issue in the field.This paper studies the deficiency in traditional apriori algorithm,and came up with a new algorithm which improved the shortcomings in traditional apriori algorithm.In the end,this paper by means of the experiment proved the feasibility of the algorithm,and the new algorithm is compared with the traditional apriori algorithm on the efficiency.The number of diabetic patients in China ranked second in the world, there are a large number of patients with impaired glucose regulation(potential diabetes),the incidence of people with diabetes is 9.7% and the number of patients is nearly 100 million. All this suggests that outbreak of epidemic trend of diabetes is not over. At present,it is still a hot issue that how to use the information technology provide auxiliary support for the medication of diabetes. This paper combined association rule mining and the theory of ontology, semi-automaticly build the diabetes medication ontology by using the improved association rule algorithm to obtain the relationship between diabetes concept. In oder to assist in the drug treatment of diabetes,this article use the diabetes medication ontology to deductive.In the end,use the inference engine of Protégé for reasoning and testing of diabetes medication ontology.Therelults show that this model is feasible. And we can use this model provide auxiliary to support for the drug treatment of diabetes.The research significance of this paper is coming up with a highly efficient association rule mining algorithm.Using the new algorithm obtains the meaningful rules in the drug treatment of diabetes, providing a guideline for semi-automaticly building of diabetes medication ontology. At the same time,through improving and making drug treatments for the diabetes patients to assist in the drug treatment of diabetes. The main achievement of the study as follows:1. In order to solve the low efficiency of traditional association rule algorithm on time and space,this paper come up with a new algorithm based on Label Matrix. The whole process just scans the database once, and does not produce candidate itemsets.The efficiency of obtaining meaningful rules is increased to a great degree.2. This paper combined association rule mining and the theory of ontology,semi-automaticly build the diabetes medication ontology by using the improved association rule algorithm to obtain the relationship between diabetes concept. In oder to achieve semi-automaticly building of the diabetes medication ontology.3. In oder to assist in the drug treatment of diabetes,this article use the diabetes medication ontology to deductive.4. In the end,use the inference engine of Protégé for reasoning and testing of diabetes medication ontology.The relults show that this model is feasible. And we can use this model provide auxiliary to support for the drug treatment of diabetes.