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

Keyphrases

• Zero-shot 100%
• COVID-19 Tweets 100%
• Chat Generative Pre-trained Transformer (ChatGPT) 100%
• Low-income Households 100%
• GPT-3 66%
• Knowledge Distillation 66%
• Supervised Machine Learning 33%
• Self-knowledge 33%
• Vulnerable Communities 33%
• Social Services 33%
• COVID-19 Pandemic Impact 33%
• Social Inequality 33%
• Social Justice 33%
• Impact on Economy 33%
• COVID-19 Data 33%
• Income Effect 33%
• Labeled Data 33%
• Social Media Content Analysis 33%
• Crisis Data 33%
• Semi-supervised Method 33%
• BERT Model 33%
• Unlabeled Data 33%
• Justice Issues 33%
• Self-training 33%

Computer Science

• Bidirectional Encoder Representations From Transformers 100%
• ChatGPT 100%
• Household Income 100%
• Knowledge Distillation 66%
• Generative Pre-Trained Transformer 3 66%
• Machine Learning 33%
• Content Analysis 33%
• Unlabeled Data 33%
• Baseline Model 33%
• Social Medium Content 33%
• Semisupervised Learning 33%