From: A Maximum-Entropy approach for accurate document annotation in the biomedical domain
Method | All Terms | Monosemous Terms | Ambiguous Terms | ||||||
---|---|---|---|---|---|---|---|---|---|
P | R | F | P | R | F | P | R | F | |
Exact Matching | 52.3 | 22.1 | 23.92 | 53.73 | 18.87 | 21.61 | 45.48 | 37.01 | 34.82 |
Naive Bayes | 82.06 | 97.45 | 88.79 | 81.95 | 97.48 | 88.72 | 82.57 | 97.29 | 89.09 |
Decision Trees | 95.55 | 79.42 | 86.06 | 95.54 | 79.72 | 86.24 | 95.57 | 78.01 | 85.2 |
MaxEnt | 99.41 | 86.77 | 92.4 | 99.43 | 86.75 | 92.39 | 99.32 | 86.87 | 92.42 |