Table 2 Best threshold values found by the grid search method
| Parameter | Value | |
|---|---|---|
| Training Set | Number of true positives | 417 |
| Number of false positives | 31 | |
| Number of false negatives | 25 | |
| Precision | 0.931 | |
| Recall | 0.943 | |
| F1-score | 0.937 | |
| Validation Set | Number of true positives | 477 |
| Number of false positives | 39 | |
| Number of false negatives | 19 | |
| Precision | 0.924 | |
| Recall | 0.961 | |
| F1-score | 0.942 | |
| Thresholds | Phonetic similarity | 0.844 |
| String similarity | 0.831 |