Research Of Association Rules And Clustering Analysis And Its Application In On-line Shopping System

With the rapid development of Internet and the online shopping system, the technology of personalized recommendation has become an important field gradually. In order to provide best personalized services for customers, this paper established a personalized recommendation model which is based on the two important algorithms in data mining- Association Rules and Clustering Analysis.Association Rules is usually used to find interesting or relevant link between item sets in large amounts of data. This paper introduced the adoption of mining multi-dimensional association rule to generate recommendations, and put forward the algorithm of mining frequent predicate sets based the combination method.Clustering Analysis is used to divide data into different data sets with its own characteristics and without given the grouping rules. The requirements of Clustering Analysis are the data in different data sets should have significant differences, and the data within each group should be similar to each other as possible. This paper studies the adoption of item clustering to generate the recommendation, which based on the algorithm of improved Agglomerative of Hierarchical Clustering and the Fuzzy c-means clustering algorithm.Finally, this paper realizes the personalized recommendation model, and uses pictures to show the recommendation effects; In addition, the paper also verifies the algorithms'advancement and the personalized recommendation model's accuracy through experiments in this paper.