The SA-GNN model for Sequential Recommendation

The implementation of the paper:

Yansen Zhang, Chenhao Hu, Genan Dai, Weiyang Kong, and Yubao Liu, "Self-Adaptive Graph Neural Networks for Personalized Sequential Recommendation", in the 28th International Conference on Neural Information Processing (ICONIP) (ICONIP 2021)

Environments

• python 3.6.8
• PyTorch (version: 1.6.0)
• GPU (GeForce RTX 2080 Ti)

Dataset

The datasets and data preprocessing can refer to HGN.

Example to run the code

Train and evaluate the model:

python run.py

Some Baselines

• [MA-GNN] can be found in my another repos
• [-GAT] can be found in the file "gating_network_gat.py"

Comments

You can change the $L$ and $T$ according to your needs, but code should be adapted too

Acknowledgements

I found these repos useful (while developing this one):

• official GAT
• HGN

Issues/Pull Requests/Feedbacks

Feel free to send me an email or add issues if you have any questions.

Please cite our paper if you use our code. Thanks!

Author: Yansen Zhang ([email protected])