The recommended algorithm is one of the most popular applications of today. Firstly, the original data is cleaned and processed, and then the association rules model and user value analysis model are established in this paper. Secondly, a Apriori algorithm is used to analyze the relationship between user history access records and the user group of K-means algorithm is used to divide value. Finally, the experimental results show that the results of the output of the association rules and the clustering ...