Attentive Social Recommendation: Towards User And Item Diversities
This work addresses the need for better diversity handling in social recommendation systems, which is incremental as it builds on existing methods by incorporating attention and disentangling strategies.
The paper tackles the problem of underutilized user and item diversities in social recommendation systems by proposing an attentive social recommendation (ASR) model that aggregates social and rating factors with automatic weighting and disentangles rating values, achieving effectiveness and advantages in benchmarks.
Social recommendation system is to predict unobserved user-item rating values by taking advantage of user-user social relation and user-item ratings. However, user/item diversities in social recommendations are not well utilized in the literature. Especially, inter-factor (social and rating factors) relations and distinct rating values need taking into more consideration. In this paper, we propose an attentive social recommendation system (ASR) to address this issue from two aspects. First, in ASR, Rec-conv graph network layers are proposed to extract the social factor, user-rating and item-rated factors and then automatically assign contribution weights to aggregate these factors into the user/item embedding vectors. Second, a disentangling strategy is applied for diverse rating values. Extensive experiments on benchmarks demonstrate the effectiveness and advantages of our ASR.