LGRec:A Group Recommendation Method Based on Graph Convolutional Neural Networks

Group activities are very common in everyday life, where people often go out for meals with friends during holidays or travel with family. Recommending content to group users has become a complex and important task. It is challenging to recommend to groups based on the preferences of individual members. On one hand, the diversity of group members makes it difficult to aggregate their preferences. On the other hand, the scarcity of historical group activities makes it challenging to learn the preferences of the group as a whole. To address these issues, we propose a group recommendation model called LGRec, based on graph convolutional neural networks (GCNN). This model utilizes graph convolution on the interaction graph among group members to simulate multi-step diffusion between groups, thereby improving the performance of group recommendations.