1) 两层attention model
2) 实验设置：
one day a session;
the last session as current short-term, and the session before that session viewed as long-term;
one random item in last session are viewed as predict one (positive example), one random item in long-term sessions is verelative negative example.
| 文献题目 | 去谷歌学术搜索 | ||||||||||
| Sequential Recommender System based on Hierarchical Attention Network | |||||||||||
| 文献作者 | Haochao Ying; Hui Xiong | ||||||||||
| 文献发表年限 | 2018 | ||||||||||
| 文献关键字 | |||||||||||
| Sequential Recommendation, Long-term, Short-term, Dynamic, Next recommendation, Small, Gowalla, one day a session | |||||||||||
| 摘要描述 | |||||||||||
| With a large amount of user activity data accumu- lated, it is crucial to exploit user sequential behav- ior for sequential recommendations. Convention- ally, user general taste and recent demand are com- bined to promote recommendation performances. However, existing methods often neglect that user long-term preference keep evolving over time, and building a static representation for user general taste may not adequately reflect the dynamic char- acters. Moreover, they integrate user-item or item- item interactions through a linear way which lim- its the capability of model. To this end, in this paper, we propose a novel two-layer hierarchical attention network, which takes the above proper- ties into account, to recommend the next item user might be interested. Specifically, the first attention layer learns user long-term preferences based on the historical purchased item representation, while the second one outputs final user representation through coupling user long-term and short-term preferences. The experimental study demonstrates the superiority of our method compared with other state-of-the-art ones. | |||||||||||
