Table 8 Results for multi-label SSVM model with different feature sets
From: Disorder recognition in clinical texts using multi-label structured SVM
Feature set | Precision | Recall | F 1-Score |
|---|---|---|---|
SSVM + BOW | 0.7626 | 0.3329 | 0.4635 |
SSVM + BOW + POS | 0.7953 | 0.3857 | 0.5195 |
SSVM + BOW + POS + capitalization | 0.8417 | 0.5702 | 0.6799 |
SSVM + BOW + POS + capitalization + case pattern | 0.8398 | 0.5839 | 0.6889 |
SSVM + BOW + POS + capitalization + case pattern + word representation | 0.8244 | 0.6620 | 0.7343 |