Integrating BERT Embeddings and BiLSTM for Emotion Analysis of Dialogue

Table 1

Comparison of the weighted average of all models.

Models	Precision	Recall	F1-score
BERT-BiLSTM-CNN	0.5900	0.5795	0.5804
BERT-BiGRU-CNN	0.5932	0.5864	0.5841
BERT-BiLSTM-attention	0.5975	0.5972	0.5895
BERT-(BiLSTM + CNN)	0.5902	0.5752	0.5766
BERT-(BiLSTM-attention + CNN)	0.5963	0.5818	0.5862
BiBERT-BiLSTM	0.6020	0.6011	0.5981