h3dema/pyg-recomm-explain

Test with Explainer and a graph recommendation network

Recommendar explanability

This is just an example on how to try to explain the results from a recommendation.

Basic install on Windows

Install the python modules needed (for CPU environment).

# install environment
python -m venv venv
.\venv\Scripts\Activate.ps1
python -m pip install --upgrade pip

# install python modules
pip install -r .\requirements.txt

How to use

First, open Recommending Amazon Products-Model.ipynb to generate the model (gnn.pt) on the current directory. Then, open one of the other two notebooks to train and use the Explainer.

References

• Explainer

  • https://pytorch-geometric.readthedocs.io/en/latest/index.html
  • https://github.com/pyg-team/pytorch_geometric/blob/master/examples/explain/gnn_explainer.py

• Recommending Amazon Products using Graph Neural Networks in PyTorch Geometric

  • https://wandb.ai/manan-goel/gnn-recommender/reports/Recommending-Amazon-Products-using-Graph-Neural-Networks-in-PyTorch-Geometric--VmlldzozMTA3MzYw