Table 5 Performance of DL classifiers with feature selection (results in% acc).
From: Clickbait detection in news headlines using RoBERTa-Large language model and deep embeddings
Model | Embeddings | Accuracy (%) | Precision | Recall | F1-Score |
|---|---|---|---|---|---|
LSTM | Word2Vec | 85 | 87 | 85 | 85 |
FastText | 77 | 79 | 78 | 78 | |
Sentence Embeddings | 94 | 95 | 96 | 94 | |
Bi-LSTM | Word2Vec | 87 | 88 | 87 | 87 |
FastText | 79 | 80 | 80 | 80 | |
Sentence Embeddings | 95 | 96 | 96 | 96 | |
| Â | Word2Vec | 89 | 89 | 89 | 89 |
GRU | FastText | 78 | 79 | 79 | 78 |
| Â | Sentence Embeddings | 94 | 94 | 95 | 96 |
| Â | Word2Vec | 89 | 89 | 89 | 89 |
Bi-GRU | FastText | 84 | 84 | 84 | 83 |
| Â | Sentence Embeddings | 95 | 97 | 97 | 96 |
BERT | Pre-Trained Embeddings | 95 | 95 | 96 | 95 |
T5 | 87 | 88 | 89 | 88 | |
DistilBERT | 91 | 92 | 91 | 92 | |
RoBERTa-Large | 97 | 98 | 97 | 98 |
