Table 4 Overall performance on datasets.
From: Enhancing knowledge graph recommendations through deep reinforcement learning
Method | Amazon-Book | Yelp2018 | Alibaba-iFashion | |||
|---|---|---|---|---|---|---|
Recall@20 | Precision@20 | Recall@20 | Precision@20 | Recall@20 | Precision@20 | |
GRU4RC | 0.1382 ± 0.0023 | 0.0453 ± 0.0017 | 0.1256 ± 0.0028 | 0.0412 ± 0.0019 | 0.0983 ± 0.0021 | 0.0356 ± 0.0015 |
SR-GNN | 0.1425 ± 0.0028 | 0.0468 ± 0.0021 | 0.1298 ± 0.0032 | 0.0427 ± 0.0023 | 0.1027 ± 0.0026 | 0.0372 ± 0.0018 |
DKN | 0.1293 ± 0.0021 | 0.0421 ± 0.0016 | 0.1187 ± 0.0025 | 0.0389 ± 0.0018 | 0.0921 ± 0.0019 | 0.0332 ± 0.0014 |
KGAT | 0.1489 ± 0.0035 | 0.0489 ± 0.0028 | 0.1362 ± 0.0038 | 0.0448 ± 0.0031 | 0.1030 ± 0.0031 | 0.0378 ± 0.0025 |
KGIN | 0.1436 ± 0.0031 | 0.0472 ± 0.0024 | 0.1321 ± 0.0034 | 0.0435 ± 0.0026 | 0.1208 ± 0.0028 | 0.0413 ± 0.0022 |
KGRec | 0.1512 ± 0.0038 | 0.0508 ± 0.0031 | 0.1398 ± 0.0041 | 0.0461 ± 0.0034 | 0.1188 ± 0.0034 | 0.0429 ± 0.0028 |
PGPR | 0.1463 ± 0.0034 | 0.0481 ± 0.0027 | 0.1345 ± 0.0037 | 0.0442 ± 0.0030 | 0.1072 ± 0.0030 | 0.0387 ± 0.0024 |
MultiHopKG | 0.1441 ± 0.0032 | 0.0473 ± 0.0025 | 0.1328 ± 0.0035 | 0.0437 ± 0.0028 | 0.1058 ± 0.0029 | 0.0381 ± 0.0023 |
TPGR | 0.1492 ± 0.0036 | 0.0490 ± 0.0029 | 0.1371 ± 0.0039 | 0.0451 ± 0.0032 | 0.1097 ± 0.0032 | 0.0396 ± 0.0026 |
RKGR | 0.1528 ± 0.0041 | 0.0502 ± 0.0034 | 0.1406 ± 0.0043 | 0.0463 ± 0.0037 | 0.1163 ± 0.0037 | 0.0419 ± 0.0031 |
ReMR | 0.1562 ± 0.0043 | 0.0505 ± 0.0036 | 0.1415 ± 0.0045 | 0.0475 ± 0.0039 | 0.1172 ± 0.0039 | 0.0423 ± 0.0033 |
DRLRS | 0.1519 ± 0.0039 | 0.0499 ± 0.0032 | 0.1392 ± 0.0042 | 0.0458 ± 0.0035 | 0.1158 ± 0.0035 | 0.0417 ± 0.0029 |
EQGPR | 0.1558 ± 0.0045 | 0.0512 ± 0.0038 | 0.1445 ± 0.0047 | 0.0471 ± 0.0041 | 0.1191 ± 0.0041 | 0.0435 ± 0.0035 |
RKGnet (Ours) | 0.1592 ± 0.0028 | 0.0523 ± 0.0021 | 0.1468 ± 0.0031 | 0.0483 ± 0.0023 | 0.1224 ± 0.0025 | 0.0442 ± 0.0019 |
% Improv. | +1.92% | +2.15% | +1.59% | +1.68% | +1.32% | +1.61% |
