Code for WSDM 2022 paper, Contrastive Learning for Representation Degeneration Problem in Sequential Recommendation.

Usage

Download datasets from RecSysDatasets or their Google Drive. And put the files in ./dataset/ like the following.

$ tree
.
├── Amazon_Beauty
│   ├── Amazon_Beauty.inter
│   └── Amazon_Beauty.item
├── Amazon_Clothing_Shoes_and_Jewelry
│   ├── Amazon_Clothing_Shoes_and_Jewelry.inter
│   └── Amazon_Clothing_Shoes_and_Jewelry.item
├── Amazon_Sports_and_Outdoors
│   ├── Amazon_Sports_and_Outdoors.inter
│   └── Amazon_Sports_and_Outdoors.item
├── ml-1m
│   ├── ml-1m.inter
│   ├── ml-1m.item
│   ├── ml-1m.user
│   └── README.md
└── yelp
    ├── README.md
    ├── yelp.inter
    ├── yelp.item
    └── yelp.user

Run duorec.sh.

Cite

If you find this repo useful, please cite

@article{DuoRec,
  author    = {Ruihong Qiu and
               Zi Huang and
               Hongzhi Yin and
               Zijian Wang},
  title     = {Contrastive Learning for Representation Degeneration Problem in Sequential Recommendation},
  journal   = {CoRR},
  volume    = {abs/2110.05730},
  year      = {2021},
}

MISC

We have also implemented CL4SRec, Contrastive Learning for Sequential Recommendation. Change the --model="DuoRec" into --model="CL4SRec" in the duorec.sh file to run CL4SRec.

Our another sequential recommender model MMInfoRec, Memory Augmented Multi-Instance Contrastive Predictive Coding for Sequential Recommendation at ICDM 2021 is also available on GitHub, MMInfoRec.

Credit

This repo is based on RecBole.