Research Article
Network Embedding-Aware Point-of-Interest Recommendation in Location-Based Social Networks
Table 3
Top-
k recommendation performance on Yelp (
denotes the improvement of Geo-SRank over other methods).
| Overall performance comparison (Yelp) | | | Precision@k | | Recall@k | | MAP | |
| MostPopular | 0.00852 | 201 | 0.0095 | 172 | 0.00444 | 180 | WRMF | 0.02384 | 7.9 | 0.02354 | 9.77 | 0.01205 | 3.3 | GeoMF | 0.02464 | 4.4 | 0.02446 | 5.6 | 0.01229 | 1.3 | BPRMF | 0.01892 | 35.9 | 0.01855 | 39.2 | 0.0085 | 46.4 | Rank-GeoFM | 0.02521 | 2 | 0.02536 | 1.8 | 0.01234 | 0.8 | IRenMF | 0.02440 | 5.4 | 0.02430 | 6.3 | 0.01216 | 2.3 | SRank | 0.02448 | 5.1 | 0.02515 | 2.7 | 0.01221 | 1.9 | Geo-SRank | 0.02573 | — | 0.02584 | — | 0.01245 | — |
|
|