An improved collaborative filtering recommendation method based on timestamp

The collaborative filtering is the traditional method in the recommender system. In order to improve the accuracy of recommender system, an improved collaborative filtering recommendation method based on the timestamp is proposed in this paper. Since the preferences of the user towards to items are a time series, so they are changing. Therefore, we choose user ratings that are related with the current user ratings through the sliding window technology. To accurately calculate the similarity between users, the timestamp information is employed to pre-filter the user preference. When recommending items for user, the timestamp is used again to filter the items that may be used by the target user. In this paper, the effectiveness of the proposed method is verified by Filmtipset datasets.