This repository was created utilizing the survey report on self-supervised learning for recommender systems.
The existing library was taken, modified, and experimented with, and this repository is called SELFRec.

---

Requirements

  numba==0.53.1
  numpy==1.20.3
  scipy==1.6.2
  tensorflow==1.14.0
  torch>=1.7.0

---

Deployment

1. Docker Hub: https://hub.docker.com/repository/docker/dodo9249/thingcn/general
2. Wandb: https://wandb.ai/d9249/ThinGCN

---

Usage

1. conf, which contains the detailed parameters of the models utilized in this study. (You can adjust the weight of Weighted Forwarding.)
2. When you run run.sh, all the models in this study are executed.

---

Implemented Models

Model	Paper	Type	Code
LightGCN	He et al. LightGCN: Simplifying and Powering Graph Convolution Network for Recommendation, SIGIR'20.	Graph	PyTorch
NCL	Lin et al. Improving Graph Collaborative Filtering with Neighborhood-enriched Contrastive Learning, WWW'22.	Graph + CL	PyTorch
SGL	Wu et al. Self-supervised Graph Learning for Recommendation, SIGIR'21.	Graph + CL	TensorFlow & Torch
MixGCF	Huang et al. MixGCF: An Improved Training Method for Graph Neural Network-based Recommender Systems, KDD'21.	Graph + DA	PyTorch
SimGCL	Yu et al. Are Graph Augmentations Necessary? Simple Graph Contrastive Learning for Recommendation, SIGIR'22.	Graph + CL	PyTorch
XSimGCL	Yu et al. XSimGCL: Towards Extremely Simple Graph Contrastive Learning for Recommendation, TKDE'23.	Graph + CL	PyTorch
BUIR	Lee et al. Bootstrapping User and Item Representations for One-Class Collaborative Filtering, SIGIR'21.	Graph + DA	PyTorch

• CL is short for contrastive learning (including data augmentation); DA is short for data augmentation only

---

Leaderboard

General hyperparameter settings are: epoch: 300, batch_size: 2048, emb_size: 64, learning rate: 0.001, L2 reg: 0.0001.

Model	Hyperparameter settings
LightGCN	layer=3
NCL	layer=3, ssl_reg=1e-6, proto_reg=1e-7, tau=0.05, hyper_layers=1, alpha=1.5, num_clusters=2000
SGL	layer=3, λ=0.1, ρ=0.1, tau=0.2
MixGCF	layer=3, n_nes=64
SimGCL	layer=3, λ=0.5, eps=0.1, tau=0.2
XSimGCL	layer=3, λ=0.2, eps=0.2, l∗=1, tau=0.15
BUIR	layer=3, tau=0.995, drop_rate=0.2

The results are obtained on the dataset of

Yelp2018
douban-book
Amazon-Book
Amazon-kindle
FilmTrust
MovieLens-1m
iFashion
gowalla

---

Related Datasets

   

DataSet	Basic Meta					User Context
	Users	Items	Ratings (Scale)		Density	Users	Links (Type)
Douban	2,848	39,586	894,887	[1, 5]	0.794%	2,848	35,770	Trust
Yelp2018	19,539	21,266	450,884	implicit	0.11%	19,539	864,157	Trust
Amazon-Book	52,463	91,599	2,984,108	implicit	0.11%	-	-	-
Amazin-Kindle	0	0	0	a	0%	-	-	-
FilmTrust	0	0	0	a	0%	-	-	-
iFashion	0	0	0	a	0%	-	-	-
MovieLens-1m	0	0	0	a	0%	-	-	-
LastFM	1,892	17,632	92,834	implicit	0.27%	1,892	25,434	Trust

---

Reference

SELFRec is a Python framework for self-supervised recommendation (SSR) which integrates commonly used datasets and metrics, and implements many state-of-the-art SSR models. SELFRec has a lightweight architecture and provides user-friendly interfaces. It can facilitate model implementation and evaluation.
Founder and principal contributor: @Coder-Yu @xiaxin1998
Supported by: @AIhongzhi (A/Prof. Hongzhi Yin, UQ)

This repo is released with our survey paper on self-supervised learning for recommender systems. We organized a tutorial on self-supervised recommendation at WWW'22. Visit the tutorial page for more information.

If you find this repo helpful to your research, please cite our paper.

@article{,
  title={ThinGCN},
  author={Sangmin Lee},
  journal={},
  year={}
}

@article{yu2022self,
  title={Self-Supervised Learning for Recommender Systems: A Survey},
  author={Yu, Junliang and Yin, Hongzhi and Xia, Xin and Chen, Tong and Li, Jundong and Huang, Zi},
  journal={arXiv preprint arXiv:2203.15876},
  year={2022}
}