PK-BERT: Knowledge Enhanced Pre-trained Models with Prompt for Few-Shot Learning

Computer Science

• Bidirectional Encoder Representations From Transformers 100%
• Pre-Trained Model 100%
• Few-Shot Learning 50%
• Language Modeling 25%
• Explicit Representation 25%
• Representation Language 25%
• Transfer Learning 25%
• Classification Task 25%
• Semantic Knowledge 25%

Social Sciences

• Bidirectional Encoder Representations From Transformers 100%
• Knowledge Graph 50%
• Learning Transfer 50%
• Language Modeling 50%

Psychology

• Learning Transfer 100%
• Semantic Knowledge 100%