Contentious Term

Research on "contentious terms" focuses on identifying and mitigating the harmful effects of biased or offensive language in various digital contexts, including legal documents, social media, and knowledge graphs. Current efforts leverage machine learning models, such as transformer-based architectures like RoBERTa and prototypical networks, to automatically detect and analyze such terms, improving user understanding and flagging potential issues. This work is crucial for enhancing fairness, transparency, and accountability in AI systems and online platforms, ultimately aiming to reduce the spread of misinformation and harmful stereotypes.