Researches On Association-Rule Algorithms Based On Boundary Idea

Following the big data to make the use of database technology widely, while the amount of data has been exponential growth. With the massive data, to elicit the usefully information seemed significant. Therefore, the technology of association rule mining emerges; its objectives are to identify the relationship between the data and useful knowledge generation. The algorithm of Apriori association rule mining is the most classic algorithms in the field, For many shortcomings in the algorithm so that the improved algorithm coming into being.The paper presents an algorithm which is based on boundary idea for association rule mining, then rough set theory has introduced to boundary idea making of association rule mining inquiry. Discussed in the following two parts:The first part:The principal task of association rule mining is to find all frequent patterns, and to calculate the frequency of each candidate frequent pattern occurrence. With integrate the boundary idea and project adjacency table, the paper proposed new viewpoint that the melioration of project adjacency table which is based on boundary idea and the algorithm of inner product of vectors. Consequently, the improved project structure which is named class project adjacency table will apply to association rule mining algorithm. According to analysis case and experimental results of the new algorithm, the new algorithm proved the feasibility and superiority.The second part:To explore the properties of the reduction on the basis of rough set theory in the association rule mining application. Through the basics of rough set, decision table information systems, reduction of attributes, attribute importance of research; and analysis and compare on rough set attribute reduction. Follow the Boundary idea act on properties reduction device, it will be applied association rule mining model of interaction between the new algorithm and the properties of the reduction on the basis of rough set.