GNN在召回中的应用：SR-GNN——Session-based Recommendation with Graph Neural

Session-based Recommendation with Graph Neural

What This Paper is About

Each session graph is proceeded successively and the latent vectors for all nodes involved in each graph can be obtained through gated Graph Neural Networks.

We represent each session as a composition of the global preference and the current interest of the user in that session, where these global and local session embedding vectors are both composed by the latent vectors of nodes.

The Article presents the proposed method of session-based recommendation with Graph Neural Networks.

What You Can Learn

Similar to STAMP, on Yoochoose, NARM achieves good performance on the short group, but the performance drops quickly with the length of the sessions increasing, which is partially because RNN models have difficulty in coping with long sequences.

L and SR-GNN-ATT achieve the close-to-optimal performance.

The proposed method not only considers the complex structure and transitions between items of session sequences, but also develops a strategy to combine long-term preferences and current interests of sessions to better predict users’ next actions.