This repository includes the implementation for Deconfounded Causal Collaborative Filtering
Paper: Deconfounded Causal Collaborative Filtering
Paper Link: https://dl.acm.org/doi/full/10.1145/3606035
Environment requirements can be found in ./requirement.txt
-
Electronics and CDs and Vinyl: The origin dataset can be found here.
-
Yelp: The origin dataset can be found here.
-
The data processing code can be found in
./src/data_preprocessing/
For example:
# DCCF on Electronics dataset
> cd ./src/
> python ./main.py --rank 1 --model_name DCCF --optimizer Adam --lr 0.001 --dataset Electronics --metric ndcg@5,recall@5,precision@5 --gpu 0 --epoch 100 --test_neg_n 1000
@article{xu2023deconfounded,
title={Deconfounded causal collaborative filtering},
author={Xu, Shuyuan and Tan, Juntao and Heinecke, Shelby and Li, Vena Jia and Zhang, Yongfeng},
journal={ACM Transactions on Recommender Systems},
volume={1},
number={4},
pages={1--25},
year={2023},
publisher={ACM New York, NY}
}