A Personalized Recommendation Fusing Tag Feature and Temporal Context
Ling Li · 2022
The rapid growth of users and items provides enormous potential for users to find their interested information. This has attracted a lot of attentions how to use both tag feature and temporal context to improve recommendation accuracy. In this paper, we calculate users’ similarity by using user-tag bipartite, so as to construct user-tag feature vector. Then, we take the temporal context into consideration to dynamically discover neighbors which have higher effect weights. Finally, we fuse the neighbor sets to collaborative filtering algorithm based on the neighborhood. We evaluate the proposed algorithm using a real-world data set and compare the performance with classical baseline methods, showing the improvements in terms of different evaluation.