| 结合用户长短期兴趣的深度强化学习推荐方法 |
| |
| 引用本文： | 阎世宏,马为之,张敏,刘奕群,马少平.结合用户长短期兴趣的深度强化学习推荐方法[J].中文信息学报,2021,35(8):107-116. |
| |
| 作者姓名： | 阎世宏 马为之 张敏 刘奕群 马少平 |
| |
| 作者单位： | 清华大学 计算机系 北京信息科学与技术国家研究中心,北京 100084 |
| |
| 基金项目： | 国家重点研发计划(2018YFC0831900);国家自然科学基金(61672311,61532011);清华大学国强研究院项目(2019GQG0004);中国博士后科学基金(2020M670339) |
| |
| 摘 要： | 结合强化学习(特别是深度强化学习)的推荐算法,在近年来相比已有方法取得了较大的提升.然而,现有绝大多数基于深度强化学习的推荐方法仅使用循环神经网络(RNN)等方法学习用户的短期兴趣,忽略了用户的长期兴趣,导致对用户的兴趣建模存在不足.因此,该文提出一种结合用户长期兴趣与短期兴趣的深度强化学习推荐方法(LSRL).首先,...
|
| 关 键 词： | 推荐系统 深度强化学习 长期与短期兴趣 协同过滤 门控循环单元 |
| 收稿时间： | 2020-10-30 |
Reinforcement Learning with User Long-term and Short-term Preference for Personalized Recommendation |
| |
| YAN Shihong,MA Weizhi,ZHANG Min,LIU Yiqun,MA Shaoping.Reinforcement Learning with User Long-term and Short-term Preference for Personalized Recommendation[J].Journal of Chinese Information Processing,2021,35(8):107-116. |
| |
| Authors: | YAN Shihong MA Weizhi ZHANG Min LIU Yiqun MA Shaoping |
| |
| Affiliation: | Department of Computer Science and Technology, Institute for Artificial Intelligence, Beijing National Research Center for Information Science and Technology, Tsinghua University, Beijing 100084, China |
| |
| Abstract: | Most of the existing recommendation methods based on deep reinforcement learning use recurrent neural network (RNN) to learn users' short-term preference, while ignoring their long-term preference. This paper proposes a deep reinforcement learning recommendation method combining both long-term and short-term user preference (LSRL). LSRL uses collaborative filtering to learn users' long-term preference representation and applies the gated recurrent unit (GRU) to learn user's short-term preference representation. The redesigned Q-network framework combines two types of representation and Deep Q-Network is used to predict users' feedback on items. Experimental results on MovieLens datasets show that the proposed method has a significant improvement according to NDCG and Hit Ratio compared to other baseline methods. |
| |
| Keywords: | recommender system deep reinforcement learning long-term and short-term preference collaborative filtering gated recurrent unit |
|
| 点击此处可从《中文信息学报》浏览原始摘要信息 |
|
点击此处可从《中文信息学报》下载全文 |
|
