Team Jigya

All Insights (Default)

Key biases in AI models used for detecting depression on social media uncovered by Northeastern grads

Key biases in AI models used for detecting depression on social media uncovered by Northeastern grads Yuchen Cao and Xiaorui Shen conducted a systematic review of AI models used in studies detecting depression in social media users and found major flaws. Social media platforms like Twitter, Facebook and Reddit offer researchers a trove of user-generated content, increasingly used to train AI tools for detecting signs of depression.