Reproduce “Challenging the Myth of Graph Collaborative Filtering: a Reasoned and Reproducibility-driven Analysis” on OpenResearch
Turn your favorite coding agents into research agents that review literature, develop hypotheses, and run experiments.
Universal app for Apple silicon and Intel. Requires macOS 11 or later.
About this paper
Vito Walter Anelli, Daniele Malitesta, Claudio Pomo, Alejandro Bellogín, Tommaso Di Noia, Eugenio Di Sciascio
The authors re-implement six graph-based collaborative-filtering methods and compare them with tuned traditional baselines across five datasets, using shared offline evaluation protocols and user-group analyses. On BookCrossing, EASE_R reaches nDCG@20 of 0.0847 versus 0.0712 for the best graph method, while RP3β leads Recall@20 at 0.0941 versus 0.0819. Rankings remain dataset-dependent.
Summary by alphaXiv.
Tip: change arxiv.org to autoarxiv.org in any arXiv abstract link to open the paper here.