Our research introduces a new approach using LLMs to understand how discussions about health measures on social media influence public health outcomes. We track 'gists' in online conversations and demonstrate how language patterns predict real world health trends.
Our research analyzes how language used by major news channels influences conversations on social media. We compare TV news content with social media posts. The research shows that news presentation can deepen political polarization online.
Our study investigates motivations and barriers regarding willingness to respond to hate speech online. We surveyed hundreds of participants and found that targets of hate respond more often. Younger individuals, women, and regular witnesses hesitate due to fears of retaliation.
Our research introduces CounterQuill, an AI tool that helps users create empathetic responses to hate speech. We evaluate the tool through a step by step process, enabling effective human-AI collaboration for counterspeech writing.
