Toward Transformer Fusions for Chinese Sentiment Intensity Prediction in Valence-Arousal Dimensions

Keyphrases

• Attention Mechanism 33%
• Bidirectional Encoder Representations from Transformers 66%
• Chinese Words 33%
• Downstream Task 33%
• Encoder Architecture 33%
• Fusion Framework 66%
• Fusion Model 33%
• Granularity 33%
• Language Representation 33%
• Lexicon-based Approach 33%
• Neural Network 33%
• Regression-based Method 33%
• Semantic Representation 33%
• Sentiment Intensity Prediction 100%
• Transformer Fusion 100%
• Transformer Neural Network 33%
• Transformer-based 33%
• Valence-arousal 100%
• Word-level 33%

Computer Science

• Attention (Machine Learning) 50%
• Bidirectional Encoder Representations From Transformers 100%
• Experimental Result 50%
• Granularity 50%
• Neural Network 100%
• Representation Language 50%
• Semantic Representation 50%

Psychology

• Neural Network 100%
• Semantic Representation 50%