Research And Implementation Of E-commerce Recommendation System Based On Multi-minimum Support Degree Association Rules

With the rapid development of Internet technology,the data began exponential growth,it is more and more difficulty for users to find their own information.Therefore,how to quickly find the information needed in the massive data will become an important research content of e-commerce recommendation systemThis paper starts from the background of the research,the significance of the research and the present situation of the research of the recommendation system of the commodity.We studied the algorithms commonly used in the recommendation system and the problems they face.Because the association rule algorithm is more successful in the E-commerce recommendation system,we adopted the association rule algorithm as the recommendation algorithm of the recommendation system.However,in the actual application of the case,the traditional association rules mining support threshold set a single,set difficult,so the algorithm is running inefficient.In this paper,an improved scheme for FP-growth algorithm,an association rule mining algorithm based on multi-project support tree structure and support degree array,is proposed,giving the implementation process of improved association rule algorithm and the concrete realization of code.At the same time,this paper also aims at a single association rule algorithm which is difficult to meet the practical application,and proposes a recommendation system based on data stratification and user-based Top-N recommendation strategy.Due to the large amount of data,the data processing takes a long time cost,the traditional data processing technology is difficult to quickly complete large-scale data processing.Therefore,we used Hadoop as a technical means to achieve efficient e-commerce recommendation system.In this paper,the data of the mall is taken as the data source,and the system is tested by using the accuracy and coverage evaluation.The experimental results show that the proposed model has a significant improvement in the accuracy of the proposed results.The coverage rate of each commodity category is over 90%.Finally,based on theoretical research and experimental testing,we discussed the architecture and implementation of e-commerce recommendation system,and used JavaEE framework and B / S mode to build the recommendation system.The system includes the two modules of the background management system and the online recommendation system,which respectively realize the data processing and the online recommended function,thus verifying the feasibility and application of the system.