Trust, transparency & control in inferred user interest models

   Abstract