A Triangular Personalized Recommendation Algorithm for Improving Diversity
Table 2
The values of the seven evaluation metrics after applying the different recommendation algorithms on the two datasets.
AUC
MAP
P
R
H
I
N
MovieLens-100K
GR
0.863
0.208
0.058
0.358
0.395
0.408
255
UCF
0.887
0.315
0.070
0.476
0.550
0.394
242
ICF
0.888
0.385
0.073
0.494
0.674
0.413
211
MD
0.898
0.325
0.075
0.527
0.618
0.355
230
CosRA
0.908
0.380
0.082
0.575
0.724
0.335
204
TR
0.6105
0.0482
0.0446
0.3196
0.8862
0.0024
138
MovieLens-1M
GR
0.856
0.144
0.053
0.222
0.403
0.415
1660
UCF
0.872
0.176
0.061
0.263
0.458
0.415
1640
ICF
0.885
0.289
0.072
0.314
0.629
0.404
1445
MD
0.885
0.188
0.066
0.297
0.504
0.403
1618
CosRA
0.895
0.223
0.074
0.350
0.598
0.387
1541
TR
0.5717
0.0447
0.0334
0.2940
0.8043
0.0017
531
Notes. The recommendation length is defined as . All the seven values are the average values of 10 independent realizations. The bold values indicate the best performance of the recommendation algorithm on a metric.
