An Improved Collaborative Filtering Algorithm Based on Expert Trust and Time Decay
Qingshuo Zheng · 2018
Collaborative Filtering (CF) is one of the most successful algorithms in recommendation systems. To solve the data sparsity and the cold-start problem in CF, an improved CF algorithm based on expert trust and time decay is proposed in this paper. The expert group is obtained by the expert trust, and the weight for the rating of the expert is calculated using the similarity between the expert and the target user. The predicted value of the target user is obtained by the weighted rating of the user and the time decay factor which used to reflect the change of the user interest. The simulation result shows that the proposed algorithm has significant higher prediction accuracy than KNN algorithm.