Table 6 Comparison of sarcasm detection results of various models in 16-shot scenario.
From: A multi-modal sarcasm detection model based on cue learning
Model | Modality | Prompt Learning | ACC (%) | F1 |
|---|---|---|---|---|
P-Tuning v242 | T | \(\checkmark\) | 65.73±2.52 | 65.28±2.48 |
BBT v243 | T | \(\checkmark\) | 64.91±2.04 | 64.26±2.59 |
BERT44 | T | \(\times\) | 62.89±4.78 | 62.56±3.24 |
ViT45 | I | \(\times\) | 58.13±5.53 | 55.35±3.35 |
HFN17 | T+I | \(\times\) | 63.12±3.67 | 62.98±3.45 |
Attr-BERT18 | T+I | \(\times\) | 64.32±4.76 | 63.87±4.09 |
InCrossMGs20 | T+I | \(\times\) | 65.09±4.65 | 65.98±4.23 |
MVK22 | T+I | \(\times\) | 66.23±3.76 | 66.89±3.54 |
SAHFN-BERT4 | T+I | \(\times\) | 67.87±3.13 | 67.96±4.72 |
CLIP-MLP49 | T+I | \(\times\) | 69.62±2.53 | 69.46±1.98 |
Our Model | T+I | \(\checkmark\) | 72.22±2.10 | 72.57±2.3 |
