This thesis is motivated by the massively increasing structured data on the Web, and the need for novel methods to exploit these data to their full potential. Building on the remarkable success of Web 2.0 mash-ups, and ...