HypAR: Hypergraph with Attention on Reviews for Explainable Recommendation

This is the code for the paper:

Hypergraph with Attention on Reviews for Explainable Recommendation

[Theis E. Jendal](https://vbn.aau.dk/da/persons/141501), [Trung-Hoang Le](http://lthoang.com/), [Hady W. Lauw](http://www.hadylauw.com/), [Matteo Lissandrini](https://people.cs.aau.dk/~matteo/), [Peter Dolog](https://peterdolog.wordpress.com), and [Katja Hose](https://homes.cs.aau.dk/~khose/About_me.html)
Presented at [ECIR 2024](https://www.ecir2024.org/).

How to run

Install requirements. To use cuda, use 'requirements_cu116txt' instead.

pip install -r requirements.txt

HypAR uses https://github.com/evison/Sentires to preprocess the data. The processed data can be downloaded from: https://github.com/PreferredAI/seer.

After downloading file, simply run

python hypar_example.py

Contact

Questions and discussion are welcome: tjendal@cs.aau.dk

Changes

HypAR has been updated to accomodate Cornac>=2.3.1. If using an older version of Cornac use old requirements file.