Simple and Behavior-Driven Augmentation for Recommendation with Rich Collaborative Signals

Introduction

This is the official repository for the PyTorch implementation of our framework Simple Collaborative Augmentation for Recommendation (SCAR).

Overview

We propose augmentation techniques (COLADD, COLREP) for contrastive learning-based graph collaborative filtering, which minimize the loss of core interactions between nodes and provide multiple collaborative signals.

Acknowledgments

This implementation is based on the open-source SSL-based recsys framework, SSLREC.

If you use our code or the processed dataset, please cite the following paper as a reference.

@inproceedings{Ren_2024, series={WSDM ’24},
   title={SSLRec: A Self-Supervised Learning Framework for Recommendation},
   url={http://dx.doi.org/10.1145/3616855.3635814},
   DOI={10.1145/3616855.3635814},
   booktitle={Proceedings of the 17th ACM International Conference on Web Search and Data Mining},
   publisher={ACM},
   author={Ren, Xubin and Xia, Lianghao and Yang, Yuhao and Wei, Wei and Wang, Tianle and Cai, Xuheng and Huang, Chao},
   year={2024},
   month=mar, collection={WSDM ’24} }

Requirements

pytorch==1.13
python==3.10
numpy==1.22.3
scipy==1.7.3
dgl==1.1.1
pyyaml==6.0.1
tensorboard

How to Run

python main.py --model scar --cuda (if using a GPU) GPU_NUM

For more details, please refer to the SSLRec paper or docs/User Guide.md.