Anti LGBTQ+

Research on detecting anti-LGBTQ+ sentiment focuses on improving automated systems for identifying homophobic and transphobic hate speech across various languages and online platforms. Current efforts utilize transformer-based models like BERT and its variants, often incorporating multilingual capabilities and training data augmented with contextually relevant social media language to enhance accuracy. This work is crucial for mitigating online harassment and fostering safer digital environments, with implications for both improving social media moderation and advancing natural language processing techniques for bias detection.