ADAPTIVE FUSION METHOD FOR USER-BASED AND ITEM-BASED COLLABORATIVE FILTERING

摘要

In many e-commerce sites, recommender systems, which provide personalized recommendationsnfrom among a large number of items, have recently been introduced. Collaborativenfiltering is one of the most successful algorithms which provide recommendationsnusing ratings of users on items. There are two approaches: user-based and item-basedncollaborative filtering. Additionally a unifying method for user-based and item-basedncollaborative filtering was proposed to improve the recommendation accuracy. The unifyingnapproach uses a constant value as a weight parameter to unify both algorithms.nHowever, because the optimal weight for unifying is actually different depending on thensituation, the algorithm should estimate an appropriate weight dynamically, and shouldnuse it. In this research, we first investigate the relationship between recommendationnaccuracy and the weight parameter. The results show that the optimal weight is differentndepending on the situation. Second, we propose an approach for estimation ofnthe appropriate weight value based on collected ratings. Then, we discuss the effectivenessnof the proposed approach based on both multi-agent simulation and the MovieLensndataset. The results show that the proposed approach can estimate the weight valuenwithin an error rate of 0.5% for the optimal weight