Joint Multimodal Aspect-Sentiment Analysis on TWITTER 2015 (test)
65.1PrecisionVLP-MABSA
Evaluation Results
| Method | Links | |||
|---|---|---|---|---|
| VLP-MABSAModality=Multimodal2022.04 | 65.1 | 68.3 | 66.6 | |
| JMLModality=Multimodal2022.04 | 65 | 63.2 | 64.1 | |
| OSCGA-collapseModality=Multimodal2022.04 | 63.1 | 63.7 | 63.2 | |
| BARTModality=Text-based2022.04 | 62.9 | 65 | 63.9 | |
| OSCGA+TomBERTModality=Multimodal2022.04 | 61.7 | 63.4 | 62.5 | |
| UMT+TomBERTModality=Multimodal2022.04 | 58.4 | 61.3 | 59.8 | |
| D-GCNModality=Text-based2022.04 | 58.3 | 58.8 | 59.4 | |
| SPANModality=Text-based2022.04 | 53.7 | 53.9 | 53.8 | |
| RpBERT-collapseModality=Multimodal2022.04 | 49.3 | 46.9 | 48 |