An Efficient Collaborative Filtering Algorithm with Item Hierarchy
Guanghua Cheng, Songjie Gong · 2008
Recommender systems are becoming increasingly popular with the evolution of the Internet, and collaborative filtering that using explicit ratings on items from users is the most successful technology for building recommendation systems. But traditional collaborative filtering algorithm is not suitable for itempsilas multiple content and multiple level recommendations. So, a new concept hierarchy methodology improving user-item matrix and integrating items of similar users and those multiple level association is presented, which not only overcomes the difficulty of data sparsity, but also solves the item's multiple content and level problem. Experimental results indicate that the algorithm can achieve better prediction accuracy and provide better recommendation results than with the traditional collaborative filtering algorithms.