%A WEN Wen, LIN Ze-tian, CAI Rui-chu, HAO Zhi-feng, WANG Li-juan %T Predicting User’s Dynamic Preference Based on Embedding Learning %0 Journal Article %D 2019 %J Computer Science %R 10.11896/jsjkx.180901801 %P 32-38 %V 46 %N 10 %U {https://www.jsjkx.com/CN/abstract/article_18548.shtml} %8 2019-10-15 %X Traditional methods for capturing user preferences mainly focus on user’s long-term preferences.However,user interests always change over time in real-world applications.As a result,how to capture user’s dynamic prefe-rences still remains a big challenge.This paper proposed an embedding-based approach for predicting user’s dynamic preferences.Firstly,an improved embedding method is used for learning the low-dimensional vector representations of items from user’s click sequences.Then,based on the learned item vectors and user’s short-term click behaviors,user’sdynamic preferences are obtained and used for predicting the next click.Experiments were conducted on two real-world datasets and the proposed method was compared with state-of-the-art methods.The results demonstrate the significant superiority of the proposed method in prediction accuracy compared with other algorithms.