GDCCDR

The source code is official pytorch implementation of GDCCDR (Graph Disentangled Contrastive Learning with Personalized Transfer for Cross-Domain Recommendation) by Jing Liu, Lele Sun, Weizhi Nie, PeiGuang Jing and Yuting Su.

  title={Graph Disentangled Contrastive Learning with Personalized Transfer for Cross-Domain Recommendation},
  author={Liu, Jing and Sun, Lele and Nie, Weizhi and Jing, Peiguang and Su, Yuting},
  booktitle={Proceedings of the AAAI Conference on Artificial Intelligence},
  volume={38},
  number={8},
  pages={8769--8777},
  year={2024}
}

Requirements:

• Python == 3.8.13
• PyTorch == 1.11.0
• torch-sparse == 0.6.13
• numpy == 1.22.3

Datasets

We use four Amazon datasets (Sport&Phone, Sport&Cloth, Elec&Phone, Elec&Cloth) to evaluate our GDCCDR. We preprocess these datasets following BI-TGCF and filter out the cold-start items.

Training

You can use these commands to train the model:

• python main.py --dataset sport_phone --ecl_reg 0.2 --pcl_reg 0.001 --alpha 0.25 --beta 0.03 --layer 6
• python main.py --dataset sport_cloth --ecl_reg 0.05 --pcl_reg 0.05 --alpha 0.1 --beta 0.3 --layer 5