Yelp2018 Top-K (Recall@20) yelp2018-recall20 Leaderboard
Top-K collaborative-filtering ranking on the Yelp2018 implicit-feedback benchmark (the standard LightGCN/SGL split). Recall@20 over all users; higher is better. Transcribed from the SELFRec leaderboard, which evaluates every method under one training harness (batch 2048, dim 64). · Metric: Recall@20 (higher is better)
| # | Model | Recall@20 | Paper |
|---|---|---|---|
| 1 | XSimGCL | 0.07 | link |
| 2 | SimGCL | 0.07 | link |
| 3 | DirectAU | 0.07 | link |
| 4 | MixGCF | 0.07 | link |
| 5 | SGL | 0.07 | link |
| 6 | NCL | 0.07 | link |
| 7 | LightGCN | 0.06 | link |
| 8 | MF (BPR) | 0.05 | link |