Table 7.
Comparison of 2006 and 2014 team results: precision, recall, and F1 of token-based binary HIPAA-PHI. Top result for each team selected by F1.
| 2006 teams | 2014 teams | ||||||
|---|---|---|---|---|---|---|---|
| Team name/run | P | R | F1 | Team name/run | P | R | F1 |
| Wellner 3 | 98.7 | 97.5 | .981 | Nottingham 1 | .9900 | .9640 | .9768 |
| Szarvas 2 | 99.3 | 96.9 | .980 | Harbin-Grad 2 | .9776 | .9629 | .9702 |
| Aramaki 1 | 99.1 | 94.9 | .970 | Manchester 3 | .9825 | .9567 | .9694 |
| Hara 3 | 96.1 | 93.8 | .949 | Harbin 3 | .9747 | .9408 | .9575 |
| Wrenn 3 | 94.9 | 94.3 | .946 | Kaiser 3 | .9692 | .8999 | .9333 |
| Guo 1 | 93.8 | 88.2 | .909 | Newfoundland 1 | .9207 | .8522 | .8851 |
| Guillen 1 | 92.8 | 70.5 | .801 | LIMSI-CNRS 3 | .9605 | .8048 | .8758 |
| ECNU 1 | .9666 | .7404 | .8385 | ||||
| San Marcos 1 | .9095 | .6045 | .7263 | ||||
| UCSD 1 | .7591 | .5161 | .6144 | ||||