Source code for Knowledge-refined Denoising Network for Robust Recommendation
- Ubuntu OS
- Python >= 3.8 (Anaconda3 is recommended)
- PyTorch 1.7+
- A Nvidia GPU with cuda 11.1+
We use three processed datasets: Alibaba-iFashion, Yelp2018 and Last-FM.
- We follow the paper "Learning Intents behind Interactions with Knowledge Graph for Recommendation" to process data.
- You can find the full version of recommendation datasets via Alibaba-iFashion, Yelp2018 and Last-FM.
- Alibaba-iFashion dataset
python main.py --dataset alibaba-ifashion --lr 0.0001 --context_hops 3 --num_neg_sample 200 --margin 0.6 --max_iter 2
- Yelp2018 dataset
python main.py --dataset yelp2018 --lr 0.0001 --context_hops 2 --num_neg_sample 400 --margin 0.8 --max_iter 1
- Last-FM dataset
python main.py --dataset last-fm --lr 0.0001 --context_hops 2 --num_neg_sample 400 --margin 0.7 --max_iter 2