Interactive Selection Recommendation Based on the Multi-head Attention Graph Neural Network

Shuxi Zhang, Jianxia Chen, Meihan Yao, Xinyun Wu, Yvfan Ge, Shu Li

Abstract: The click-through rate prediction of users is a critical task in the recommendation system. As a powerful machine learning method, graph neural networks have been favored by scholars to solve the task recently. However, most graph neural network-based click-through rate prediction models ignore the effectiveness of feature interaction and generally model all feature combinations, even if some are meaningless. Therefore, this paper proposes a Multi-head attention Graph Neural Network with Interactive Selection, named MGNN_IS in short, to capture the complex feature interactions via graph structures. In particular, there are three sub-graphs to be constructed to capture internal information of users and items respectively, and interactive information between users and items, namely the user internal graph, item internal graph, and user-item interaction graph correspondingly. Moreover, the proposed model designs a multi-head attention propagation module for the aggregation with an interactive selection strategy. This module can select the constructed graph and increase diversity with multiple heads to achieve the high-order interaction from the multiple layers. Finally, the proposed model fuses the features, and predicts. Experiments on three public datasets demonstrate that the proposed model outperformed other advanced models.