Multi-interaction fusion collaborative filtering for social recommendation

GNNs(Graph Neural Networks) use graph structure to make recommendations, receiving more and moreattention. Firstly, existing work focuses on aggregating social interaction information, ignoring users who rateon the same items. Secondly, the existing recommendation methods cannot dynamically reflect changes inuser interests. Thirdly, existing methods do not take into account the interaction of subgraphs in GNNs andinteractions between user and item factors. In this paper, a graph social fusion recommendation (GSFR) methodis proposed. GSFR captures multiple social information simultaneously, based on which it can dynamicallyadjust user interest weight. Specifically, GSFR captures the interactions in subgraphs with a dynamic attentionmechanism, which can represent changes in user interests in heterogeneous networks. A mutualistic mechanismis combined to simulate the mutually reinforcing relationship between social behavior and virtual behavior.User and item latent factors are obtained based on space vectors from the aggregation part. Recommendationsare made from inherent characteristics of space vectors interaction behaviors. Comprehensive experimentalresults on three public datasets show the effectiveness of the proposed model.

Automated summary

This paper proposes a graph social fusion recommendation method, which can capture multiple social information simultaneously and dynamically adjust user interest weights. It utilizes a dynamic attention mechanism to capture interactions in subgraphs, representing changes in user interests in heterogeneous networks. A mutualistic mechanism is combined to simulate the mutually reinforcing relationship between social behavior and virtual behavior.