Corpus | LLL | HPRD50 | IEPA | AIMed |
---|
(%) | P | R | F | P | R | F | P | R | F | P | R | F |
---|
k-NN | 71.6 | 79.8 | 74.4 | 71.9 | 62.4 | 65.9 |
68.4
| 66.6 | 67.2 |
52.1
| 35.9 | 42.3 |
BEST1G |
74.7
|
82.2
|
76.5
| 72.5 | 72.6 | 71.6 | 67.7 | 71.3 | 69.2 | 50.8 |
40.9
|
45.1
|
U3G | 74.6 | 80.7 | 75.9 | 72.5 | 72.6 | 71.6 | 68.3 |
71.7
|
69.8
| 49.5 | 40.5 | 44.4 |
O2G |
74.7
|
82.2
|
76.5
|
73.0
|
74.3
|
72.6
| 68.1 | 71.3 | 69.5 | 50.2 | 39.8 | 44.3 |
- Precision (P), recall (R), F-score (F) results of our three approaches (BEST1G, U3G, O2G) evaluated by 10-fold document-level cross-validation on four corpora, LLL, HPRD50, IEPA, and AIMed, shown in the second, third, and fourth row. As a baseline, in the first row, we add results when only k-NN is applied, and feature selection using contribution levels of groups consisting of related features was not performed. Precision (P), recall (R), and F-score (F) values are shown by percentage (%). Bold typeface shows best results per corpus in terms of precision, recall, and F-score