Table 3.
IMT results on the test dataset
| System | Precision | Recall | F1 score | AUC iP/R |
|---|---|---|---|---|
| m-LR | 58.37 | 55.80 | 57.06 | 34.96 |
| m-SVM | 62.33 | 48.94 | 54.83 | 33.18 |
| b-SVM (*) | 52.56 | 52.45 | 52.50 | 28.45 |
| union(m-SVM,b-SVM) (*) | 53.21 | 59.61 | 56.23 | 35.85 |
| union(m-LR,b-SVM) | 52.51 | 64.67 | 57.96 | 39.06 |
| union(m-LR,m-SVM) | 57.43 | 56.18 | 56.80 | 35.26 |
| intersect(m-LR,b-SVM) (*) | 64.06 | 44.01 | 52.17 | 29.47 |
| intersect(m-LR,m-SVM,b-SVM) (*) | 64.86 | 44.42 | 52.73 | 30.52 |
Results on the IMT test dataset (%). The models were trained on the combined training and development datasets. m-LR – multi-label Logistic Regression; m-SVM – multi-label Support Vector Machines; b-SVM – binary Support Vector Machines. Asterisks (*) denote systems that were submitted to the challenge.