Non-lineal gradual forgetting collaborative filtering algorithm capable of adapting to users' drifting interest

Since the traditional Collaborative Filtering(CF)algorithm often leads to bad recommendation when users' interests are changed,and inspired by the forgetting law of psychology,a non-lineal gradual forgetting collaborative filtering algorithm is developed which diminishes the importance of each rate with time and then determines similarities between different users.The experiment using MovieLens dataset shows that the algorithm is more accurate than the traditional collaborative filtering algorithms.