RECOMMENDER SYSTEMS BASED ON COLLABORATIVE FILTERING IN E-COMMERCE
Collaborative Filtering is one of the most important technology of the recommender sys tems in E-Commerce. The recommender systems based on Collaborative Filtering(CF)employ statistical techniques to find a number of customers known as nearest neighbors. Based on the rating of the nearest neighbors to the product items, the systems produce the recommended list for the target customer. In our paper, we present the entire process of CF- based recommender system, which includes three sub -tasks, namely representation, neighborhood formation, and recommendation generation. Then we analyze the limitation of traditional CF-based recommender systems, such as scalability, scarcity and quality. Finally, we will introduce the current research in this area.
E-commerce Collaborative filtering Recommender System Similarity Nearest neighbors
Guo Liming Luo Xiangzheng
Southwest University of Finance and Economics, Chengdu, 610071, China MAIPU Communication Technology Co. LTD, Chengdu, 610041, China
国际会议
The Second China and U.S. Advanced Workshop in Electronic Commerce 2004
成都
英文
136-139
2004-06-20（万方平台首次上网日期，不代表论文的发表时间）