Study And Application Of Quantitative Association Rule Mining Algorithm

With the rapid development of Internet technology,data information has grown explosively.In order to find hidden and valuable information from a large number of data,data mining technology becomes an important research field.As an important algorithm in data mining,association rule mining has achieved fruitful results in many aspects.However,with the emergence of a large number of data and various types of data,the mining technique of association rules is more demanding.Association rule mining technology can be divided into Boolean association rule and quantitative association rule according to the different processing objects.In the mining process of quantitative association rule,it is usually first to discrete the quantitative attribute,and then the Boolean association rule mining algorithm is used to deal with it.However,the discretization of quantitative attributes is usually accompanied by the problem of interval division,which can lead to the change of the result when the number is too large or too small.In this paper,quantitative attribute discretization is the starting point.The main research contents are as follows:(1)In view of the problems existing in the discretization of quantitative attributes,an improved clustering Algorithm of Iterative self-organizing Data Techniques Algorithm is proposed.It can automatically carry out interval division and select the initial clustering center based on the density automatically to ensure the stability of clustering results.At the same time,the algorithm control parameters and convergence conditions can be determined according to the clustering validity index.Experiments show that the algorithm has higher accuracy and stable performance than the original algorithm.(2)In the mining process of Boolean association rules,the algorithm of Apriori algorithm is low in efficiency,too many candidates are generated,and user demand is not considered,and the improved algorithm based on constraint and compression matrix is proposed.On the one hand,set rule constraints and mined for rules that interest users.On the other hand,the transaction database is transformed into a matrix to improve the efficiency of the algorithm.The results show that the algorithm is more efficient than the Apriori algorithm.(3)Based on the improved quantitative association rule mining algorithm,this paper studies the influence factors of student satisfaction.Using the survey data of " China College Student Survey",the main factors affecting student satisfaction were explored through data preprocessing,modeling and experimental results analysis.The results show that customer expectation has the greatest influence on student satisfaction,and other factors have different influence on student satisfaction.Finally,it puts forward practical suggestions on how to improve the service and improve college students' satisfaction.