Research Article
Applying Deep Learning Technologies to Evaluate the Patent Quality with the Collaborative Training
| Category | Train set | Test set | Precision | Recall | | Support | Precision | Recall | | Support |
| 1 | 0.98 | 0.99 | 0.98 | 19967 | 0.96 | 0.98 | 0.97 | 2232 | 2 | 0.93 | 0.92 | 0.92 | 10743 | 0.89 | 0.87 | 0.88 | 1227 | 3 | 0.93 | 0.87 | 0.90 | 11534 | 0.89 | 0.81 | 0.85 | 1209 | 4 | 0.77 | 0.82 | 0.80 | 5330 | 0.65 | 0.72 | 0.68 | 626 | 5 | 0.58 | 0.80 | 0.67 | 2502 | 0.53 | 0.65 | 0.58 | 293 | 6 | 0.54 | 0.50 | 0.52 | 1087 | 0.48 | 0.36 | 0.41 | 113 | 7 | 0.53 | 0.67 | 0.60 | 1382 | 0.44 | 0.59 | 0.50 | 144 | 8 | 0.69 | 0.74 | 0.72 | 930 | 0.57 | 0.53 | 0.55 | 98 | Micro | 0.89 | 0.90 | 0.89 | 53475 | 0.84 | 0.85 | 0.84 | 5942 | Macro | 0.74 | 0.79 | 0.76 | 53475 | 0.67 | 0.69 | 0.68 | 5942 |
|
|