Code for paper [Disentangled Contrastive Learning for Cross-Domain Recommendation].
Accepted by DASFAA 2023.
- Python 3.6
- PyTorch 1.10.2
- Numpy
OurModel.py
: our implementation of modelrun.py
: model training and testingdata/
: four cross-domain recommendation tasks based on two widely used datasets Amazon and DoubanAmazon/Cell_Elec/
Amazon/Movie_Music/
Douban/Movie_Book/
Douban/Music_Book/
utils/
load_data.py
: auxiliary functions constructing training set and testing set for cross-domain scenarioparser.py
: some parameters concerned with the modelhelpers.py
: functions to save the model
-
default: python run.py (Task: Amazon-Movie-Music)
If you want to choose a certain task: python run.py --dataset=[chosen dataset] --domain_1=[chosen domain 1] --domain_2=[chosen domain 2]