Joint Multimodal Aspect-Sentiment Analysis on TWITTER 2017 (test)
66.9PrecisionVLP-MABSA
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| VLP-MABSAModality=Multimodal2022.04 | 66.9 | 69.2 | 68 | |
| JMLModality=Multimodal2022.04 | 66.5 | 65.5 | 66 | |
| BARTModality=Text-based2022.04 | 65.2 | 65.6 | 65.4 | |
| D-GCNModality=Text-based2022.04 | 64.2 | 64.1 | 64.1 | |
| OSCGA-collapseModality=Multimodal2022.04 | 63.5 | 63.5 | 63.5 | |
| OSCGA+TomBERTModality=Multimodal2022.04 | 63.4 | 64 | 63.7 | |
| UMT+TomBERTModality=Multimodal2022.04 | 62.3 | 62.4 | 62.4 | |
| SPANModality=Text-based2022.04 | 59.6 | 61.7 | 60.6 | |
| RpBERT-collapseModality=Multimodal2022.04 | 57 | 55.4 | 56.2 |