Research And Application Of Apriori Algorithm In Association Rules Mining

Association rules technology occupies an important position in the field of data mining research,and its main purpose is to excavate the related knowledge between things.With the increasing scale and complexity of data sets,how to improve the efficiency of mining association rules for large data sets has become the theme of association rule mining.As a classical algorithm of mining association rules,Apriori algorithm has always been a hot spot of research by scholars.This paper systematically studied the classic Apriori algorithm,and introduced some existing improved algorithms with relatively high efficiency.Then,aiming at the shortcomings of Apriori algorithm in scanning database performance and producing invalid candidate item sets of large scale,a new IM-Apriori algorithm is proposed.The algorithm maps the database to a Boolean matrix by scanning and calculates the matrix directly.In the process of algorithm execution,the matrix is repeatedly compressed and pruned by using the relevant properties,so the efficiency of mining is greatly improved.After that,theoretical analysis and comparison between the original Apriori algorithm and the improved algorithm are made,and the superiority of the improved algorithm is obtained from several aspects of scanning times,space complexity and time complexity.Besides,the numerical experiments show that the mining performance of the IM-Apriori algorithm is much better than that of Apriori algorithm.Finally,the improved algorithm is used to analyze the data stored in the student performance database and finds out the potential rules.From the results of the mining,it can be found that there is a certain correlation between some courses.In addition,the order of course will also affect the students' achievements.These rules can provide some suggestions for the teaching management.