Attentive Social Recommendation: Towards User And Item Diversities

TL;DR

This paper introduces an attentive social recommendation system that leverages user and item diversities, social relations, and rating values through graph networks and disentangling strategies to improve recommendation accuracy.

Contribution

It proposes Rec-conv graph network layers for adaptive factor aggregation and a disentangling strategy for rating diversity, addressing gaps in existing social recommendation models.

Findings

ASR outperforms baseline methods on benchmark datasets.

Rec-conv layers effectively integrate social and rating factors.

Disentangling enhances the modeling of rating diversity.

Abstract

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.