Disentangled Contrastive Learning for Social Recommendation
This work addresses the challenge of improving recommendation accuracy in social networks by disentangling user behaviors, representing an incremental advancement over existing social recommendation models.
The paper tackles the problem of modeling users' heterogeneous behavior patterns in social recommendation by proposing a disentangled contrastive learning framework, DcRec, which learns separate user representations for item and social domains and transfers knowledge between them, achieving superior performance on real-world datasets.
Social recommendations utilize social relations to enhance the representation learning for recommendations. Most social recommendation models unify user representations for the user-item interactions (collaborative domain) and social relations (social domain). However, such an approach may fail to model the users heterogeneous behavior patterns in two domains, impairing the expressiveness of user representations. In this work, to address such limitation, we propose a novel Disentangled contrastive learning framework for social Recommendations DcRec. More specifically, we propose to learn disentangled users representations from the item and social domains. Moreover, disentangled contrastive learning is designed to perform knowledge transfer between disentangled users representations for social recommendations. Comprehensive experiments on various real-world datasets demonstrate the superiority of our proposed model.