Research On Collaborative Filtering Technologies Of Recommendation System For E-Commerce

As the internet and information technology rapidly develops, the electronic commerce has been widely used by enterprises and individual users. As the amount of merchandise information quickly increases, customers are easily confused by the huge number of goods. It is hard for customers to locate the merchandise they need. Sometimes it is even harder to know where to find the information about the merchandise. The electronic commerce recommendation systems are developed to help users to solve these problems.The collaborative filtering is one of the main technologies for the electronic commerce recommendation system. However, it has two problems: the lack of algorithm scalability and the sparsity of dataset. In order to solve these problems, we developed a hybrid recommendation system called Collaborative Filtering & Clustering based recommendation system. Firstly, the clustering algorithm is utilized to cluster items (merchandise) into several classes. The operations for one user following the clustering algorithm are limited within the interested classes of his/her own. This strategy improves the scalability of the recommendation algorithm and reduces the computation time. Secondly, the system begins the collaborative filtering process. In this step, an item-based algorithm is employed to compute predictive values and insert high values into the original matrix. A denser matrix is resulted from this process. Finally, a user-based algorithm is used to attain the final predictive value.An experiment has been designed to validate the effect of the proposed algorithm. The simulation results show that our method is more effective than traditional collaborative filtering algorithms.