Table 4
Comparison of macro-averaged scores for different model versions.
| Classifier | Precision | Recall | F1-score |
|---|---|---|---|
| top-level current model | 0.98 ± 0.01 | 0.99 ± 0.01 | 0.98 ± 0.01 |
| bottom-level current model | 0.58 ± 0.01 | 0.77 ± 0.01 | 0.60 ± 0.01 |
| top-level 152 features model | 0.96 ± 0.01 | 0.99 ± 0.01 | 0.97 ± 0.01 |
| bottom-level 152 features model | 0.57 ± 0.01 | 0.76 ± 0.01 | 0.59 ± 0.01 |
Notes. The table presents the mean and standard deviation of the macro-averaged scores obtained from the 20 predicted testing sets of the new labeled set. This model includes only the 152 features present in the previous model version, but was trained with the same labeled set as the updated model. For comparison, the macro-averaged scores of the previous model from SS21 are also shown, allowing for an assessment of the performances of the different models and labeled sets.
Current usage metrics show cumulative count of Article Views (full-text article views including HTML views, PDF and ePub downloads, according to the available data) and Abstracts Views on Vision4Press platform.
Data correspond to usage on the plateform after 2015. The current usage metrics is available 48-96 hours after online publication and is updated daily on week days.
Initial download of the metrics may take a while.