Identifying COVID-19 Tweets Relevant to Low-Income Households Using Semi-supervised BERT and Zero-shot ChatGPT Models

Keyphrases

• Tweets 100%
• COVID-19 100%
• Zero-shot 100%
• Low-income Households 100%
• ChatGPT Model 100%
• GPT-3 40%
• Knowledge Distillation 40%
• Economic Impact 20%
• Supervised Machine Learning 20%
• Self-knowledge 20%
• Vulnerable Communities 20%
• Social Services 20%
• Social Inequality 20%
• Social Justice 20%
• Income Effect 20%
• Labeled Data 20%
• Social Media Content Analysis 20%
• Semi-supervised Method 20%
• BERT Model 20%
• Unlabeled Data 20%
• Justice Issues 20%
• Self-training 20%
• Pandemic Impact 20%
• Crisis Data 20%
• Chat Generative Pre-trained Transformer (ChatGPT) 20%

Social Sciences

• COVID-19 100%
• ChatGPT 100%
• Zero-Shot Learning 100%
• Bidirectional Encoder Representations From Transformers 100%
• Low Income Household 100%
• Knowledge Distillation 50%
• Alanine Aminotransferase 50%
• Social Services 25%
• COVID 19 Epidemic 25%
• Social Inequality 25%
• Content Analysis 25%
• Justice Issues 25%
• Social Media 25%
• Machine Learning Model 25%
• Crisis 25%
• Program Content 25%
• Semisupervised Learning 25%

Computer Science

• Bidirectional Encoder Representations From Transformers 100%
• ChatGPT 100%
• Household Income 100%
• COVID-19 100%
• Zero-Shot Learning 100%
• Knowledge Distillation 50%
• Generative Pre-Trained Transformer 3 50%
• Content Analysis 25%
• Unlabeled Data 25%
• Baseline Model 25%
• Social Medium Content 25%
• Semisupervised Learning 25%
• COVID 19 Epidemic 25%
• Machine Learning Model 25%
• Self-Training 25%