GCMC-Pytorch-dgl

Introduction

A Pytorch implementation of GCMC model with Deep Graph Library (DGL). The user-item bipartite graph is built using dgl-Heterogeneous Graph.

Dataset link: https://drive.google.com/file/d/1wLq3dkWxxVyymrNlqQLZB4nuj6PEeQlW/view?usp=sharing .

DGL official implementation: https://github.com/dmlc/dgl/tree/master/examples/pytorch/gcmc.

Paper

The gcmc model is proposed by the paper below:

@article{DBLP:journals/corr/BergKW17,
  author    = {Rianne van den Berg and
               Thomas N. Kipf and
               Max Welling},
  title     = {Graph Convolutional Matrix Completion},
  journal   = {CoRR},
  volume    = {abs/1706.02263},
  year      = {2017},
  url       = {http://arxiv.org/abs/1706.02263},
}

Run the code

Run the following command in the terminal:

python main.py --data_name ml-1m --gpu 0 --epoch 100 --embed_size 20 --lr 1e-2

Meaning of the arguments

--lr: learning rate
--gpu: gpu id
--epoch: number of training epoches
--embed_size: size of the hidden representations of nodes, should be able to be divided by  the number of possible rating values(5 in ml-1m).

There are also several optional arguments for this model, read parse.py for details.