Quangui Zhang, Longbing Cao, Chengzhang Zhu, Zhiqiang Li and Jinguang Sun. CoupledCF: Learning Explicit and Implicit User-item Couplings in Recommendation for Deep Collaborative Filtering, IJCAI2018
Recent work on incorporating specific user/item attributes and their characteristics, such as user demographics, item description, and user’s review on items, into recommendation has shown its potential in improving recommendation quality and addressing issues such as sparsity and cold start. However, existing work usually treats users/items as independent while ignoring the rich coupling relationships within and between users and items, leading to limited performance improvement. In reality, users/items are related with various couplings existing within and between users and items, which may better explain how and why a user has personalized preference on an item. In this work, we propose a neural user-item coupling learning collaborative filtering, CoupledCF. CoupledCF jointly learns both explicit and implicit couplings within/between users and items w.r.t. user/item attributes and deep features for deep CF recommendation. Empirical results on two real-world large datasets show that CoupledCF significantly outperforms two latest neural recommenders: neural matrix factorization (up to 40%) and Google’s Wide&Deep network (over 15%) on these datasets.