Table 8 5-Fold Cross-Validation results on LIAR dataset (Mean ± Standard Deviation).

From: Graph-augmented transformer ensemble framework for robust and scalable fake news detection in social media ecosystems

Configuration

Accuracy (%)

F1-score (%)

ROC-AUC (%)

GNN Only

90.4 ± 0.5

90.1 ± 0.6

91.2 ± 0.4

BERT Only

92.1 ± 0.4

91.8 ± 0.5

93.2 ± 0.3

RoBERTa Only

93.0 ± 0.5

92.7 ± 0.5

94.0 ± 0.3

BERT + GNN (no ensemble)

94.3 ± 0.4

94.0 ± 0.4

95.1 ± 0.3

Full Model (GETE Ensemble)

96.2 ± 0.3

96.0 ± 0.4

97.1 ± 0.2

Back to article page