TIE-DGNN

Paper and code

This is the code for the Paper: Transition Information Enhanced Disentangled Graph Neural Networks for Session-based Recommendation. We have implemented our methods in Pytorch.

Usage

The code contains the datasets that has been processed, i.e., Tmall,Nowplaying,Last.fm etc. Then you can run the file TIE-DGNN\main.py to train the model.
For example: First, python build_graph.py --dataset Tmall, then python main.py --dataset Tmall The datasets we have uploaded have completed the build_graph steps.

Requirements

Python 3 ；PyTorch 1.9.1