BiLSTM achieves 89% accuracy and 0.89 weighted F1 on 20-class emotion detection, marginally outperforming SVM at 88.11% on a 79,595-sentence dataset.
Semeval-2007 task 14: Affective text
1 Pith paper cite this work. Polarity classification is still indexing.
1
Pith paper citing it
fields
cs.CL 1years
2026 1verdicts
UNVERDICTED 1representative citing papers
citing papers explorer
-
Benchmarking PyCaret AutoML Against BiLSTM for Fine-Grained Emotion Classification: A Comparative Study on 20-Class Emotion Detection
BiLSTM achieves 89% accuracy and 0.89 weighted F1 on 20-class emotion detection, marginally outperforming SVM at 88.11% on a 79,595-sentence dataset.