Association Rules Algorithm Based on the Intersection

Mining association rules in the database is one of important study in data mining research. Traditional association rules consist of some redundant information, and need scan database many times and generate lots of candidate item sets. Aiming at low efficiency in association rules mining using traditional methods, this paper proposes the algorithm (ISMFP), which is based on intersection for mining the maximum frequent patterns. Firstly, applying the intersection theory of mathematics, put forwards a number of concepts and definitions. Then gives the process of association rules mining, and analyzes its performance. After that, the example describes the process of implementation of the algorithm. Finally, the experimental results show that the algorithm ISMFP is efficient on mining frequent patterns, especially there exists low threshold of support degree or long patterns.