基于时序图神经网络和对比学习的会话推荐

Session-based recommendations are designed to predict the next click item using an anonymous user\'s current ongoing session. Since the user\'s information is unknown, the information available is limited. In recent years, because of the excellent performance of graph neural network in many applications, many work has applied graph neural network to session based recommendation. However, this paper finds that converting session data into graph-structured data is a lossy graph coding approach, which results in the loss of item-order information in the session. In addition, session-based recommendations can only be recommended using one session. The problem of sparse data is more serious than other suggestions. Self-supervised learning can discover real samples and has great potential to solve the problem of data sparsity. Therefore, this paper proposes a session-based recommendation model based on graph neural network and contrast learning to solve the problem of information loss and data sparsity in graph coding.