From: Network-theoretic information extraction quality assessment in the human trafficking domain
Metric | Set1 | Set2 | Set3 | Set4 | Set5 | Set6 | Set7 | Set8 |
---|---|---|---|---|---|---|---|---|
No. of unique extractions | 2,298 | 1,756 | 3,668 | 1,697 | 1,716 | 2,355 | 4,248 | 2,206 |
No. of extractions per advertisement | 1.0850 | 0.9639 | 1.3004 | 0.9490 | 0.9649 | 1.0999 | 1.4206 | 0.9638 |
No. of ads with no extractions | 194 | 651 | 367 | 708 | 531 | 181 | 132 | 1,623 |
No. of ads with at least 1 extraction | 11,336 | 10,879 | 11,163 | 10,822 | 10,999 | 11,349 | 11,398 | 9,907 |
Precision | 1.0 | 0.9845 | 0.7419 | 1.0 | 1.0 | 0.9864 | 0.7638 | 0.9845 |
Recall | 1.0 | 0.8746 | 0.8892 | 0.8746 | 0.6796 | 1.0 | 1.0 | 0.8745 |
F-score | 1.0 | 0.9263 | 0.8089 | 0.9414 | 0.8092 | 0.9931 | 0.8661 | 0.9263 |