HG-GNN

The implementation of Heterogeneous global graph neural networks for personalized session-based recommendation.

Usage

Data Processing

• The data preprocessing code is following: https://github.com/twchen/lessr/blob/master/utils/data/preprocess.py and https://github.com/CRIPAC-DIG/SR-GNN/blob/master/datasets/preprocess.py

Training & Testing

• Run python main.py
• You can modify the basic.ini configuration file to set the dataset details and modify the config/model.ini file to set the hyper-parameters of target model.
• update: You can get train_seq.pkl using the data_processor code: data_helper.py

Requirement

• pip isntall -r requirements.txt