E2E-TBSA

Reported on 4 benchmarks across 2 tasks · 1 paper · 4 SOTA

Note: results are matched by exact model name. Different papers may use the same name for different model variants.

Natural Language Processing5 results

Sentiment AnalysisonSemEval 2014 Task 4 Subtask 1+2
F1· 2018-11-13
57.9
best: 83.76 (gpt-3.5 finetuned)
SOTA
A Unified Model for Opinion Target Extraction and Target Sentiment Prediction arXiv:1811.05082
Sentiment AnalysisonSemEval 2014 Task 4 Laptop
F1· 2018-11-13
57.9
best: 79.34 (InstructABSA)
SOTA
A Unified Model for Opinion Target Extraction and Target Sentiment Prediction arXiv:1811.05082
Aspect-Based Sentiment Analysis (ABSA)onSemEval 2014 Task 4 Subtask 1+2
F1· 2018-11-13
57.9
best: 83.76 (gpt-3.5 finetuned)
SOTA
A Unified Model for Opinion Target Extraction and Target Sentiment Prediction arXiv:1811.05082
Aspect-Based Sentiment Analysis (ABSA)onSemEval 2014 Task 4 Laptop
F1· 2018-11-13
57.9
best: 79.34 (InstructABSA)
SOTA
A Unified Model for Opinion Target Extraction and Target Sentiment Prediction arXiv:1811.05082
Sentiment AnalysisonSemEval 2014 Task 4 Subtask 1+2
F1· 2018-11-13
57.9
best: 83.76 (gpt-3.5 finetuned)
A Unified Model for Opinion Target Extraction and Target Sentiment Prediction arXiv:1811.05082