How can research help us better understand online hatred and manage it effectively?