Reproducibility Report for UltraGCN: Ultra Simplification of Graph Convolutional Networks for Recommendation
Hartmann, Fabian, Hepp, Sebastian, Yifan Mayr, Matthias Moik · Zenodo (CERN European Organization for Nuclear Research) · 2023
We report on the reproducibility of the results of UltraGCN by Mao et al. published during CIKM ’21, November 2021, Virtual Event, QLD, Australia. UltraGCN is an ultra simplified graph convolutional network for recommendation. In this work, the reproducibility of the original paper is assessed in terms of its experimental design, data/code availability.