User-Weight Model for Item-based Recommendation Systems

Nowadays, item-based Collaborative Filtering (CF) has been widely used as an effective way to help people cope with information overload. It computes the item-item similarities/differentials and then selects the most similar items for prediction. A weakness of current typical itembased CF approaches is that all users have the same weight in computing the item relationships. In order to improve the recommendation quality, we incorporate users’ weights based on a relationship model of users into item similarities and differentials computing. In this paper, a model of user relationship, a method for computing users’ weights, and weight-based item-item similarities/differentials computing approaches are proposed for item-based CF recommendations. Finally, we experimentally evaluate our approach for recommendation and compare it to typical item-based CF approaches based on Adjusted Cosine and Slope One. The experiments show that our approaches can improve the recommendation results of them.