What Will You Like? Ask People Who Are Like You: Past and Future Research on Collaborative Filtering in Recommender Systems

Recommender systems have been well recognized and deployed as an effective tool for automatically recommending items to users, particularly in e-Commerce. As the most successful approach used in those systems, collaborative filtering (CF) technology receives increasing attentions from both research and commercial communities. This technology has quickly evolved and penetrated from the research world to popular applications. However, the architectural issues such as cold-start, sparse ratings, and scalability continue to dominate and plague the field. This paper dissects major research problems and issues related to CF in recommender systems and introduces state-of-the-art solutions to different problems