Table 1 Inter-annotator agreement for annotations of PII between the annotators (A1 & A2) and with the curated gold standard (GS). The values are pair-wise F1 scores, and the support is the number of instances of each PII
From: A pseudonymized corpus of occupational health narratives for clinical entity recognition in Spanish
Entity class | Support | F1 score | ||
|---|---|---|---|---|
GS - A1 | GS - A2 | A1 - A2 | ||
Global (micro avg) | 5,460 | 0.92 | 0.92 | 0.86 |
Occupation | 1960 | 0.90 | 0.92 | 0.83 |
Full Date | 1105 | 0.98 | 0.97 | 0.95 |
Date Part | 859 | 0.96 | 0.92 | 0.88 |
Health Care Unit | 604 | 0.90 | 0.89 | 0.83 |
Company | 317 | 0.85 | 0.84 | 0.74 |
Age | 280 | 0.96 | 0.94 | 0.92 |
Last Name | 136 | 0.93 | 0.92 | 0.86 |
Location | 108 | 0.78 | 0.79 | 0.61 |
First Name | 65 | 0.92 | 0.87 | 0.79 |
ID | 13 | 0.96 | 0.69 | 0.67 |
Personal ID | 8 | 0.88 | 0.93 | 0.80 |
Phone Number | 3 | 1.00 | 0.80 | 0.80 |
2 | 1.00 | 0.00 | 0.00 | |