As the consumer purchasing behavior in real life is greatly influenced by his/her friends, it is of great importance to recommend products and sharing user information with others. In this paper, we propose a novel target customer selection method based on collaborative filtering, which is a typical data mining technique. The proposed E-commerce personalized recommendation system is made up of 1) user behavior saving, 2) recommendation algorithm and 3) recommendation output. Next, collaborative filtering is used to recommend suitable products to target users according to their neighbors purchasing behaviors. To test the effectiveness of the proposed algorithm, we compare the performance of our algorithm with SVD algorithm and nearest neighbor policy. Experimental results prove that the proposed algorithm can select suitable target customers for a given type of product with high accuracy.
