Identifying Grey Sheep Users By The Distribution of User Similarities In Collaborative Filtering
Yong Wei Zheng, Mayur Agnani, Mili Singh · 2017
Recommender Systems have been successfully applied to alleviate the information overload problem and assist the process of decision making. Collaborative filtering, as one of the most popular recommendation algorithms, has been fully explored and developed in the past two decades. However, one of the challenges in collaborative filtering, the problem of "Grey Sheep" user, is still under investigation. "Grey Sheep" users is a group of the users who have special tastes and they may neither agree nor disagree with the majority of the users. The identification of them becomes a challenge in collaborative filtering, since they may introduce difficulties to produce accurate collaborative recommendations. In this paper, we propose a novel approach which can identify the Grey Sheep users by reusing the outlier detection techniques based on the distribution of user-user similarities. Our experimental results based on the MovieLens 10M rating data demonstrate the ease and effectiveness of our proposed approach.