Association Rule Mining Research Based On Rough Set Theory

With the development of technique of database, quantity of data in database increases rapidly. When data is stored into database continuously, people more and more show interesting in the association among item sets in database, which is called association rule. Nowadays, association rule has been applied to all kinds of fields, such as medical treatment, market analysis. However, the efficiency of traditional algorithms of association rule mining is lower, and people may show interesting in only some of association rules mined. Thus, It is important to mine association rules effectively which people are interesting in.We focus on following several aspects: (1)Rough sets which is one of theoretical tools of data mining;(2) Generalization of decision table which serves as a preprocessing for data mining;(3) Association rule mining based on Rough set theory.Rough set theory, which is a powerful tool in dealing with vagueness and uncertainty, forms an appealing foundation for data mining. Based onthe research of rough set theory, we discussed the arithmetic of attribute reduction.Generalization of decision table is important for the data mining. We study the technology of discretization of continuous-valued attributes and attribute-oriented concept tree ascension of nominal attributes. When the agriculture domain is targeted, we combine these two technologies.We improve on the algorithm Apriori which is classical arithmetic of association rule mining based on rough set theory. A prototype system was constructed for association rule mining within the framework of rough set theory, and applied to a database. The result shows the effectiveness of prototype system.