Personalized Recommendation Based on Collaborative Filtering with Social Network Analysis

Abstract

Collaborative Filtering (CF) is recommendations technique that provides personalized recommendation of services and products to customers by understanding preference through similarity between customers. The network of customers which can be made based on the information about customer's visit information can be used to increase the effect of recommendation. In this study, it suggests a CF based recommendation method that uses network centrality measures of customers with similarities of customers in CF. The usefulness of the proposed method is tested using user visiting log data of a representative UCC (User Created Contents) site. The experimental results show that the combined usage of SNA (Social Network Analysis) measures with similarity measures provides better recommendation performance than traditional CF method.