Table 3.
Average precision changes with Triage (training) + Triage (testing)
| Chemical names | BASE | IXN | TOPIC | IXN + TOPIC |
|---|---|---|---|---|
| 2-Acetylaminofluorene | 0.6702 | 0.6742 | 0.6969 | 0.6956 |
| Amsacrine | 0.6980 | 0.6956 | 0.6773 | 0.6848 |
| Aniline | 0.7765 | 0.7891 | 0.7887 | 0.8006 |
| Aspartame | 0.4845 | 0.5096 | 0.4687 | 0.4859 |
| Doxorubicin | 0.8610 | 0.8627 | 0.8690 | 0.8689 |
| Indomethacin | 0.9758 | 0.9766 | 0.9748 | 0.9751 |
| Quercetin | 0.9315 | 0.9313 | 0.9310 | 0.9313 |
| Raloxifene | 0.8060 | 0.8107 | 0.8152 | 0.8191 |
| Average performance | 0.7754 | 0.7812 | 0.7777 | 0.7827 |
The Triage dataset is used for training and testing in a leave-one (chemical)-out approach. ‘BASE’ means word features without substance/journal names. ‘IXN’ and ‘TOPIC’ mean semantic and topic features, respectively. ‘BASE’ features are used for all the experiments.