
Analyzing Hope vs. Hate in LGBTQ+ News
Leveraging AI to understand positive and negative user engagement
This research analyzes over 1.4 million comments across 3,161 YouTube news videos to understand how users engage with LGBTQ+ content in mainstream media.
Key findings:
- Developed a fine-grained hope speech classifier to detect positive, negative, neutral, and irrelevant content
- Provided comprehensive analysis of user interactions with LGBTQ+ news content
- Created valuable insights through consultation with public health experts
- Established a framework for distinguishing between supportive and harmful discourse
Medical Significance: This research offers a vital resource for the LGBTQ+ community by identifying supportive content and harmful rhetoric, with direct implications for mental health outcomes and community wellbeing.