RepPad

Official source code for RecSys 2024 paper: Repeated Padding for Sequential Recommendation

We provide implementations of the two most representative sequential recommendation models, GRU4Rec and SASRec. You can quickly apply RepPad to your sequential model based on the pseudo-code provided in the paper.

Run the Code

Go to the src folder in the SASRec or GRU4Rec directory, then run the following commands.

--aug_type=0 represents not using RepPad (traditional padding). --aug_type=1 represents using random(1,max) repeated padding. --aug_type=2 represents using random(1,max) repeated padding with delimiter 0.

python main.py --data_name=Toys_and_Games --aug_type=0 --model_idx=3
python main.py --data_name=Beauty --aug_type=0 --model_idx=3
python main.py --data_name=Sports_and_Outdoors --aug_type=0 --model_idx=3
python main.py --data_name=Home --aug_type=0 --model_idx=3
python main.py --data_name=Yelp --aug_type=0 --model_idx=3

python main.py --data_name=Toys_and_Games --aug_type=1 --model_idx=4
python main.py --data_name=Beauty --aug_type=1 --model_idx=4
python main.py --data_name=Sports_and_Outdoors --aug_type=1 --model_idx=4
python main.py --data_name=Home --aug_type=1 --model_idx=4
python main.py --data_name=Yelp --aug_type=1 --model_idx=4

python main.py --data_name=Toys_and_Games --aug_type=2 --model_idx=5
python main.py --data_name=Beauty --aug_type=2 --model_idx=5
python main.py --data_name=Sports_and_Outdoors --aug_type=2 --model_idx=5
python main.py --data_name=Home --aug_type=2 --model_idx=5
python main.py --data_name=Yelp --aug_type=2 --model_idx=5

Log Files

We also provide some log files and trained weights on these five datasets of SASRec in the src/output directory. xxxxx-1.txt is the performance of the original model, xxxxx-2.txt is the performance after adding RepPad.

Acknowledgement

• Training pipeline is implemented based on CoSeRec.
• SASRec model are implemented based on RecBole.

Thanks them for providing efficient implementation.

Reference

Please cite our paper if you use this code.

@inproceedings{dang2024repeated,
  title={Repeated Padding for Sequential Recommendation},
  author={Dang, Yizhou and Liu, Yuting and Yang, Enneng and Guo, Guibing and Jiang, Linying and Wang, Xingwei and Zhao, Jianzhe},
  booktitle={Proceedings of the 18th ACM Conference on Recommender Systems},
  pages={497--506},
  year={2024}
}