Blurring-Sharpening Process Models for Collaborative Filtering

This repository provides an efficient and clarified PyTorch implementation of the Blurring-Sharpening Processing Model (BSPM). The original and official implementation of BSPM can be found here. The paper can be found here.

Setup

conda create -n bspm python=3.10
conda activate bspm
conda install pytorch==1.12.1 cudatoolkit=11.3 -c pytorch
conda install torchdiffeq==0.2.2 numpy==1.22.4 scipy==1.9.3 pyyaml tqdm

Run

# Execution with command-line arguments
python main.py --dataset gowalla --sharp_solv rk4 --sharp_step 1 --sharp_time 2.5 --ideal_cutoff 448 --ideal_weight 0.2

# Execution with a configuration file
python main.py --config bspm_gowalla