An Improved Collaborative Filtering Recommendation Algorithm Based on Fusion Time Factor

Traditional collaborative filtering recommendation algorithms only focus on the user's historical behavior information, using similarity measurement method to obtain the user group with similar behavior to the target user and producing the recommendation result. However, the recommendation accuracy is low and the recommendation result is relatively simple and lacks novelty. In practice, the user's interest changes dynamically with time, and traditional collaborative filtering algorithms cannot reflect the user's interest changes in time. For this problem, this paper proposes an improved collaborative filtering algorithm based on traditional collaborative filtering algorithm. When analyzing the user's historical behavior data information and establishing user behavior characteristics, consider the influence of time factors on the user recommendation results. Experiments show that the improved collaborative filtering algorithm with fusion time factor has higher accuracy than the traditional collaborative filtering algorithm, and its recommendation results meet the needs of users, which proves the effectiveness of the algorithm.