media-bias-group / neural-media-bias-detection-using-distant-supervision-with-babe Goto Github PK

Hi,

Thank you so much for your research work.

I am just wondering can you also report the TensorFlow version and any other library versions for training classification models because when I re-runed the scripts on TensorFlow 2.4.2, the results are slightly lower than the original results in the paper. My results are shown as follows:

5-Fold CV Loss: 0.5169394910335541
5-Fold CV Accuracy: 0.753519156488151
5-Fold CV Precision: 0.8360317202906071
5-Fold CV Recall: 0.6097539149888143
5-Fold CV F1 Score: 0.6999760907781679
5-Fold CV Micro F1 Score: 0.753519156488151
5-Fold CV Weighted Macro F1 Score: 0.7462714051484434

Best Regards,
Qin

great work, just curious why distant labeling is used when the data was manually annotated?