DLFS-Rec

Source code for our RecSys 2023 Paper "Distribution-based Learnable Filters with Side Information for Sequential Recommendation" .

If you find the code helpful for your research, please cite our paper:

@inproceedings{liu2023distribution,
  title={Distribution-based Learnable Filters with Side Information for Sequential Recommendation},
  author={Liu, Haibo and Deng, Zhixiang and Wang, Liang and Peng, Jinjia and Feng, Shi},
  booktitle={Proceedings of the 17th ACM Conference on Recommender Systems},
  pages={78--88},
  year={2023}
}

Architecture

Performance

The ground-truth item is paired with 99 randomly sampled negative items, and performance comparisons are presented in the following picture.

Training

The values of hyperparameters can be determined based on specific circumstances. For optimal parameters related to experimental results, please refer to the readme file in 'reproduction' folder.

python main.py \
--data_name data_name \
--hidden_size 128 \
--num_hidden_layers 2 \
--hidden_dropout_prob 0.5 \
--attribute_hidden_size 128 \
--lr 0.0001

Contact

If you have any questions about our paper or codes, please send email to zhixiang123.deng@gmail.com.