Table 4 Multi-label classification. Annotated data results with SCOVACLIS Score (\(SCOVACLIS_s\)) and filtered terms with SNOMED CT
From: Collecting specialty-related medical terms: Development and evaluation of a resource for Spanish
Original | SNOMED CT filter | ||||||
|---|---|---|---|---|---|---|---|
Classifier | Word representation | P (%) | R (%) | F1 (%) | P (%) | R (%) | F1 (%) |
Random forest | TF-IDF + \(SCOVACLIS_s\) | 70.0 | 17.5 | 28.7 | 69.1 | 14.0 | 22.6 |
Decision tree | TF-IDF + \(SCOVACLIS_s\) | 46.2 | 43.5 | 44.8 | 37.3 | 34.0 | 35.8 |
KNeighbors | TF-IDF + \(SCOVACLIS_s\) | 69.3 | 42.6 | 52.7 | 62.8 | 31.1 | 41.6 |
MLP | TF-IDF + \(SCOVACLIS_s\) | 74.7 | 57.4 | 64.9 | 70.1 | 56.1 | 60.5 |
Random forest | \(SCOVACLIS_s\) | 76.0 | 32.6 | 45.6 | 65.4 | 14.0 | 23.1 |
Decision tree | \(SCOVACLIS_s\) | 42.6 | 43.1 | 42.8 | 31.5 | 26.8 | 28.9 |
KNeighbors | \(SCOVACLIS_s\) | 69.4 | 42.1 | 52.4 | 57.9 | 29.0 | 38.7 |
MLP | \(SCOVACLIS_s\) | 75.8 | 43.5 | 55.3 | 73.4 | 24.3 | 36.6 |