運用二階層同好群組解決合作式過濾推薦之稀疏問題

Abstract

Recommender systems have become dominant to reduce information overload and customize information access. The most successful recommender systems is collaborative filtering, which considers preferences of other users sharing similar interests. A major problem of collaborative filtering is the sparsity problem, which refers to a situation in which transactional data is sparse and insufficient to identify similarities in user interests. Accordingly, this research proposed two second-degree neighborhood methods to alleviate the sparsity problem in collaborative filtering. Extensive experiments using EachMovie data is used to analyze the characteristics of our methods. The results show that our approach contributes to the improvement of prediction quality of recommendations, especially the sparsity problem.

Recommender systems have become dominant to reduce information overload and customize information access. The most successful recommender systems is collaborative filtering, which considers preferences of other users sharing similar interests. A major problem of collaborative filtering is the sparsity problem, which refers to a situation in which transactional data is sparse and insufficient to identify similarities in user interests. Accordingly, this research proposed two second-degree neighborhood methods to alleviate the sparsity problem in collaborative filtering. Extensive experiments using EachMovie data is used to analyze the characteristics of our methods. The results show that our approach contributes to the improvement of prediction quality of recommendations, especially the sparsity problem.