A Multimodal Music Emotion Classification Method Based on Multifeature Combined Network Classifier

Table 3

Accuracy of different lyrics classification models.

Classification models	Angry	Happy	Relaxed	Sad	Average
Word2vec + CNN	0.584	0.62	0.615	0.671	0.622
Word2vec + LSTM	0.685	0.709	0.647	0.731	0.693
Word2vec + CNN-LSTM	0.721	0.823	0.627	0.742	0.728
Our model	0.752	0.815	0.646	0.756	0.742