基于信息瓶颈的社交推荐优化模型
|
|
蔡晓东,方晟,李婷
|
Information bottleneck-based social recommendation optimization model
|
|
Xiaodong CAI,Sheng FANG,Ting LI
|
|
| 表 3 各模型实验结果对比 |
| Tab.3 Comparison of experimental results of various models |
|
| Models | Epinions | | Yelp | | Douban-Book | | R@10 | N@10 | R@20 | N@20 | | R@10 | N@10 | R@20 | N@20 | | R@10 | N@10 | R@20 | N@20 | | GraphRec | 0.0436 | 0.0315 | 0.0681 | 0.0387 | | 0.0672 | 0.0485 | 0.1077 | 0.0607 | | 0.0971 | 0.1145 | 0.1453 | 0.1237 | | DiffNet++ | 0.0468 | 0.0329 | 0.0727 | 0.0406 | | 0.0707 | 0.0516 | 0.1114 | 0.0640 | | 0.1010 | 0.1184 | 0.1489 | 0.1270 | | SocialLGN | 0.0416 | 0.0307 | 0.0634 | 0.0371 | | 0.0681 | 0.0507 | 0.1059 | 0.0620 | | 0.1034 | 0.1182 | 0.1527 | 0.1274 | | ESRF | 0.0462 | 0.0329 | 0.0727 | 0.0406 | | 0.0718 | 0.0526 | 0.1123 | 0.0645 | | 0.1042 | 0.1199 | 0.1534 | 0.1301 | | GDMSR | 0.0461 | 0.0326 | 0.0721 | 0.0414 | | 0.0739 | 0.0535 | 0.1148 | 0.0658 | | 0.1026 | 0.1001 | 0.1538 | 0.1245 | | $\underline{{\rm{GBSR}}}$ | 0.0529 | 0.0385 | 0.0793 | 0.0464 | | 0.0805 | 0.0592 | 0.1243 | 0.0724 | | 0.1189 | 0.1451 | 0.1694 | 0.1523 | | IB-SRO | 0.0571 | 0.0412 | 0.0843 | 0.0493 | | 0.0862 | 0.0635 | 0.1318 | 0.0771 | | 0.1216 | 0.1481 | 0.1763 | 0.1561 | | Impro/% | 7.94 | 7.01 | 6.31 | 6.25 | | 7.08 | 7.26 | 6.03 | 6.49 | | 2.27 | 2.07 | 4.07 | 2.50 |
|
|
|