Table 17 Detailed classification results (F1-score) using BERT. Results on WH+TW are averaged on ten samples
From: Analysis and classification of privacy-sensitive content in social media posts
Dataset | F1(sens.) | F1(non-sens.) | F1(macro) |
|---|---|---|---|
SENS2 | 0.73 | 0.83 | 0.78 |
SENS3 | 0.85 | 0.92 | 0.89 |
OMC | 0.75 | 0.61 | 0.68 |
WH+TW | 0.85 ± 0.01 | 0.91 ± 0.01 | 0.88 ± 0.01 |