Application of content-based user profiling in hybrid recommender approach
Lei Ren · Microcomputer Information · 2010
User profiling is an important component of recommender,user profiling can mine and represent user preference knowledge.Aiming at the issues of the user profiling granularity and the sparsity of user-item matrix in the classical collaborative filtering,this paper proposed a hybrid recommender approach based on collaborative filtering and content-based user profiling which is implemented by Widrow-Hoff learning.The experimental evaluation demonstrated that the proposed approach outperforms other similar ones in convergence and accuracy.