A Survey on Graph Neural Networks Method under Social Recommendation

With the exponential growth of online data, the social recommendation system conducts a comprehensive analysis of user-item interaction and user-user social relationships to understand the user’s interests and generate personalized recommendations. The integration of graph neural networks and social recommendation is a current hot topic in academia. Based on the survey, we reviewed the social recommendation system constructed using graph neural networks technology. We introduce the recommendation system and graph neural network framework, analyze the challenges faced by the graph neural networks in recommendations, and discuss possible solutions. Finally, we summarize and classify the coping methods in social recommendation based on existing research.