BTW some people have asked how we're doing our article parsing. We're using the Diffbot article API which has been great for us so far. Recommended if you need similar functionality.