Research On Multi-Dimensional Association Rules Mining

With the fast development of large database, huge amount of data have been stored in computers. But the existing database systems do not provide users with the necessary and effective tools to capture all stored information easily. Therefor, automatic knowledge discovery techniques have been developed to capture and use the voluminous information hidden in large database. Discovery of association rules is an important class of data mining whose aim is to capture the co-occurrences of itemsets, the most important thing to do is to find the large itemsets effectively, because this is time consuming and will finally decide the efficiency of algorithms. So now the main study is emphasized on how to find the large itemsets with more and more few time.In this paper, we propose a new structure, called Multi-dimensional Predicate Itemset Tree (MPIT), and a highly efficient multi-dimensional association rules mining algorithm based on this structure. MPIT stores the complete predicate information of each transaction in the database and divides it into a series of subsets according to different predicates. After that build itemsettree to process each of them. Experiments show that our algorithm has good performance in both time and space. Based on above work, we put forward an improved algorithm combined with meta-rule guiding, concept leveling and data cube techniques, which makes the mining algorithm more specific and faster.